Home

An Integral Web-based Environment for Control - e-Spacio

Contents

1. Wait Wait Wait Connection Connection i Connection r Users management Documentation Collaborative services Client applets files Web Server Database Bookings System Lab Server Figure 8 2 AutomatL bs project network 2 To develop a virtual and remote control laboratory based on standard es tablished in this dissertation 3 To generate all the necessary documentation for the laboratory such as an explanation of the Ejs based GUI tasks protocol and practical guides for students to complete activities independently 4 To keep the plants in an operative state during the execution period of the practical activities 5 To answer students questions regarding the laboratory tasks and or ex periments they are working on 6 To evaluate reports written by students These remote labs were offered to a total of 112 Master students in engineering schools coming from the other universities Students from each university worked 8 2 The AutomatL bs project 165 on three of the nine remote systems available offered by the AutomatL bs project one plant from their university two plants from other locations Unlike the UNED the teaching methodology of the other academic centres focused on face to face hands on activities in the presence of a teacher assistant For this reason some live demonstrations taking pl
2. Desorption E U N E D Bookings Server Server settings E MySQL settings Emersion settings Server values Password ecccccccecce Repit password eccececsecec re J se J o jJ Figure 6 13 Server site of the automatic bookings system The most important configuration parameters of the server are the following 1 Configuration of the supervisor s e mail account of each laboratory The system confirms reservations made by users sending an e mail to the su pervisor Also if the Lab Server is switched off when a new reservation is made an e mail is sent to the Lab Server administrator reporting operation went badly 2 IP address of the server computers connected to the laboratory plants Lab Servers which are suscribed to the bookings system 6 4 Conclusions 147 3 IP address and port of the MySQL database from eMersion This infor mation is necessary since the Bookings Main Server allows access to the system to users on the eMersion list only 4 General parameters of the server as for example the communication port with the Java Interface of every Lab Server updating frequency of the plants database in Lab Servers destination of the log files etc 5 Schedule constraints for each one of the practical experiments 6 4 Conclusions This chapter has described the implementation of the Automatic Bookings and Authentication System that manages access to
3. b Stage 2 HTML Test Page Mozilla Firefox Details of your booking alejes Plant 6 Motor de corriente continua 2 UNED Date 28 10 2008 Press Save button to confirm definitivly your booking e 00 00 01 00 e 01 00 02 00 e 02 00 03 0 e 03 00 04 0 es 04 00 05 00 e 05 00 06 00 Available em Occupied Availability of the plant 6 D Selected Not available Motor de corriente continua 2 UNED Date 28 10 2008 e 06 00 07 00 e 12 00 13 00 e 18 00 19 00 0700 08 e 15 00 14 00 e 1190 2000 e 08 00 09 00 e Diosa e 20 00 21 00 e 09 00 10 00 en 11500 1600 e 21 00 22 00 e 10 00 11 00 en 11600 1700 e 22 00 23 00 11 00 12 00 17 00 18 00 e 23 00 00 00 CJ The plant is turned on sct time e 00 00 01 00 e 06 00 07 00 e 12 00 13 00 es 01 00 02 00 e 13 00 14 00 es 02 00 03 00 e 08 00 09 00 7 14 00 15 00 e 03 00 04 0 09 00 10 00 15 00 16 00 e 04 00 05 00 e 10 00 11 00 16 00 17 00 es 05 00 05 00 e 11 00 12 00 e 17 00 18 00 en Available em Occupied Selected C The plant is turned on Not avalable e 11500 1900 e 19 00 20 00 e 20 00 21 00 e 21 00 22 00 e 22 00 23 00 e 23 00 00 00 sto c Stage 3 HTML Test Page Mozilla Firefox Confirmation sto d Stage 4 My confirmed boo
4. 38 3 7 Publishing the virtual lab of the single tank process as an applet 38 3 8 Stream of information between client and server 40 3 9 Structure of exchanged data packets between client and server 41 3 10 Stack connections of the Internet Protocol Suite 42 3 11 Encapsulation of data descending through the protocol stack 43 3 12 Feedback control scheme o o o 45 3 13 Distributed control architecture using LabVIEW and Ejs 46 3 14 Concurrent tasks in the server side AT 3 15 3 16 3 17 3 18 3 19 3 20 3 21 3 22 3 23 3 24 3 25 3 26 3 27 3 28 3 29 3 30 3 31 3 32 3 33 4 1 4 2 4 3 4 4 4 5 4 6 4 7 4 8 List of Figures Three loops running concurrently in the LabVIEW server 48 Concurrent tasks in the client side o o o o 51 Structure of exchanged data packets between client and server 51 Communication methods for remote experimentation 52 Visual feedback of the remote plant through video images 55 Camera SONY EVI D31 with external video server AXIS 2400 56 WebCamimage view element on Ejs ooo o 59 Simple illustration of the augmented reality concept 60 Command based architecture o o o oo o 63 The JiL Server front panel o o oo 64 Use of two Invoke Node blocks to modify a property and invoke a method on a virtual inst
5. 67 Methods in Ejs to link with a Jil published Vl 75 Description of the main VIs for the Identity checking module 137 Student questionnary results 2006 2007 UNED 163 Parameters obtained from the identification process 201 SSGM for the threetank system o o 202 RGA for the threetank system o leen 202 System parameters wel oat a tue ea Pete eds 220 System parameters ees 226 Dimensions and power requirements a 227 Data acquisition requirements o 227 Calibration of the equipment llle 227 Motor parameters e 231 Steel disk parameters 232 Otherparametefs vu a e A Y OR DW rs 232 List of Figures 1 1 Scheme of the flexible education paradigm 6 2 1 On line experimentation system Global architecture 18 2 2 Experimentation layer General scheme 19 2 3 Server side implementation requirements 21 2 4 elearning platform architecture o o 24 3 1 Simplified model view control paradigm in EjS 33 3 2 Simulation algorithm of any Ejs application 35 3 3 Control of a first order system The single tank process 35 3 4 Closed loop control in the single tank process 36 3 5 Defining the MODEL of the single tank process 37 3 6 Defining the VIEW of the single tank process
6. Drawables Basic Bodies Bg Aegre Figure 4 16 The WebCamlmage appears at the top of the hierarchy of the DrawingPanel container to enrich the view with augmented reality The final graphical user interface of the three tank system s web based labora tory can be seen in Figures 4 17 and 4 18 Figure 4 17 a shows the GUI working in simulation mode We can see that it is similar to the virtual lab developed in Section 4 2 2 however a few changes were made A new button named Connect was placed at the bottom right corner of the control panel By pushing this but ton the user connects with the server side since the connectButton method see Figure 4 13 is invoked Figure 4 17 b shows the GUI working in remote mode when clicking on the Connect button The graphical representation of the process is replaced by video images captured from the IP camera located in the 94 4 Prototypes Developed remote laboratory of the university while the main variables of the interface are updated with data obtained from the JiL server One should highlight the fact that the tabbed panel of the interface has a new element named VIDEO This tab contains a cluster of six buttons which allow to observe the process from different perspectives through video images In particular Figure 4 18 a shows the upper part of tank 1 that corresponds to the perspective obtained when pre
7. Four variable system University of Le n 8 2 The AutomatLQbs project 169 Robot arm University of Alicante Using this laboratory enables to control a robot arm to plan and track trayecto ries see Figure 8 8 The virtual lab allows to work on a 3D representation of the Scorbot ER IX robot arm located in the University of Alicante The simu lation represents the real movement of the robot arm as realistically as possible Students can design simple and complex trayectories and analyze the dynamic behaviour of the response Jara et al 2008 2009b Connect to remote xcecute Simple S Connected Trajectory 1 Ix x a Hardware b Ejs view in remote mode Figure 8 8 Robot arm University of Alicante Ball and beam system Polytechnic University of Valencia This system enables to obtain the position control of a ball rolling on a beam with horizontal movement see Figure 8 9 The beam inclination angle is controlled eJournal Control Language Position 90 95 time seconds reference green output blue and error red U P H D theta rd CONTROL PID 75 80 85 90 95 ction co time seconds O C Position Control Speed Position m 0 34 Speed m sec 0 Moving Rterence O Reference m 0 37 Ang Vel rd s 0 Play Pause Reset Disconn Time sec 98 8 Theta rd 0 23
8. Kumamoto Japan Jara C Esquembre F Candelas F Torres F amp Dormido S 2009a New features of Easy Java Simulations for 3D modeling in Proceedings of the 8th IFAC Symposium on Advances in Control Education ACE Kumamoto Japan Java 2009 Sun microsystems Website http java sun com docs books tutorial ii8n index html JiL 2009 JiL Server Overview Website http www dia uned es hvargas jiloverview html Johansson K H 2000 The quadruple tank process a multivariable labora tory process with an adjustable zero IEEE Transactions on Control Systems Technology 8 3 456 465 188 9 Bibliography Johnson L Levine A amp Smith R 2009 The Horizon Report 2009 Edition Technical report New Media Consortium EDUCAUSE Learning Iniciative http www nmc org publications 2009 horizon report Kazmer M amp Haythornthwaite C 2005 Multiple perspectives on online learn ing Special issue on online learning communities 25 1 7 11 Kumpaty S amp Haeg D 2007 Innovations 2007 World Innovations in Engi neering Education and Research International Network for Engineering Edu cation amp Research iNEER pp 421 429 Kurhila J Miettinen M Nokelainen P amp Tirri H 2004 The Role of the Learning Platform in Student Centered e Learning Technical report Helsinki Institute for Information Technology HIIT http cosco hiit fi Articles
9. L I 500 600 700 800 900 1000 1100 1200 1300 1400 1500 time seconds Figure A 6 Disturbance rejection Three tank system On the other hand Figure A 6 shows the behaviour of the controllers under these sudden changes in the dynamic of the system disturbances The experi ment starts around the normal operating point At time t 600 sec the leak valve of tank 1 is opened at 25 and the leak valve of tank 2 is opened at 1596 at time t 1000 sec In other words a disturbance is added to the system at this moment We observe that controllers are able to keep the liquid levels around the operating point in a reasonable time regardless the disturbances A 2 The DC Motor Modelling and control 207 A 2 The DC Motor Modelling and control A 2 1 Modelling The apparatus is composed of permanent magnets that provide a constant mag netic field whose value is weak due to the rotor s current remains small Under this assumption and neglecting the inductance of the coil the static and dynamic equations that govern the behavior of the engine can be derived Mathematical model The dymanic model of the DC motor is described by a set of variables u t V Voltage applied to the motor i t A Current in the motor w t rad s Rotational speed of the charge Mm t Nm Torque applied to the motor Vs Constant flow generated by the inductor R Q Resistence in the motor inductor K Nm AVs Constant of the motor
10. 3 2 Adapting virtual labs for remote experimentation 53 Listing 3 1 Excerpt of the Java code to connect the Ejs client to server side 1 public boolean connect connected false try javaSocket new Socket onetank dia uned es 2055 create connection in new DatalnputStream javaSocket getInputStream input buffer out new DataOutputStream javaSocket getOutputStream output buffer if javaSocket null If connected connected true connection is ok OMAN DAF WN p o play executing evolution RoR Noe catch java net IOException io System out println Method startTCP Problems connecting to host return connected BPE RRR Nook Ww Listing 3 2 Excerpt of the Java code to disconnect the Ejs client from server side public void disconnect if connected if javaSocket null try in close close input stream out close close output stream javaSocket close close connection javaSocket null in null out null connected false catch java io IOException e System out println Method closeTCP Close socket error OMAN OAT RWwNHrFROHKHAN AUF WBN BS The receiver and sender methods see Listings 3 3 and 3 4 should be launched on independent Java threads just when the connect method is started The sender method is used to report changes in the view that affect the operation of the remote equip
11. Learning Technologies 24 3 620 622 186 9 Bibliography Gillet D amp Fakas G 2001 eMersion A New Paradigm for Web based Training in Engineering Education in Proceedings of the International Conference on Engineering Education Oslo Norway Gillet D Geoffroy F Zeramdini K Nguyen A V Rekik Y amp Piguet Y 2003 The Cockpit An Effective Metaphor for Web based Experimenta tion in Engineering Education International Journal of Engineering Education 19 3 389 397 Gillet D Nguyen A V amp Rekik Y 2005 Collaborative Web Based Exper imentation in Flexible Engineering Education IEEE Transactions on Educa tion 48 696 704 G mez L amp Garcia J 2007 Advances on remote laboratories and e learning experiences Deusto Publicaciones Guzman J Dominguez M Berenguel M Fuertes J Rodriguez F amp Reguera P 2010 Entornos de Experimentaci n para la Ense anza de Con ceptos B sicos de Modelado y Control Revista Iberoamericana de Autom tica e Inform tica Industrial RIAI 7 1 10 22 Guzm n J L str m K J Dormido S H gglund T Berenguel M amp Piguet Y 2008 Interactive learning modules for PID control Lecture Notes IEEE Control Systems Magazine 28 5 118 134 Guzm n J L Berenguel M Rodr guez F Vargas H S nchez J amp Dormido S 2007b Desarrollo de un entorno de experimentaci
12. induce disturbances Guzm n et al 2007a b 2010 M eJournal Control H1 H3 Setpoint Status video Reading video from camera PID control Parameters 100 Kp 19 46 Tij2224 Teo El o pa 4 22 cm yO 0 20 40 60 80 100 120 sec A Mm H cm 6 1 H3 cm 0 0 125 250 375 500 0 0 0 im Kit f AL Q cm 3 sec 248 9 Q2 cm 3 sec o Play Pause Reset Disconn SPIcm 5 Time sec 121 v video LEFT TIME sec 2383871 a Hardware b Ejs view in remote mode Figure 8 5 One tank system University of Almer a Roto magnet system Polytechnic University of Catalonia This plant was conceived and designed to analyze control problems linked to the frequential response of linear systems see Figure 8 6 The process allows to 168 a Hardware Status video Reading video from camera Device PI PL Res Repetitive 8 System Assessment velocidad on o o8 T gt igna velocidad re PI OPLRES Repetitive Play Pause Reset Status Lab Connected to the labserver Disconn Vel revis 9 4 Ref revis 8 0l U voltios 0 06 time s 71 98 Velocidad rev s T 68 0 68 5 69 0 69 5 700 705 710 71 5 tiempo Acci n de control V T 68 0 68 5 69 0 695 700 705 710 71 5 tiempo b
13. pensation via an integrator and the derivative term improves transient response through high frequency compensation via a differentiator To calculate the PID controller parameters in each loop SISO tunning TITO environment V squez amp Morilla 2002 1999 has been used This tool allows to A 1 The Three tank system Modelling and control 205 simulate 2 x 2 systems with descentralized control in both open and closed loop with PID controllers to analyze the performance of the controller design RGA DNA Gain and Phase Margin measurements and to comparate several designs to choose the most optimal PI controllers Tq 0 were calculated by using a TITO environment for the threetank system 1 Ci s 2 3 38 1 ETE Co s 8 54 1 50 895 4 15 The controllers in A 15 were obtained for a Phase Margin 60 in a normal frequency operation range for the system Reference tracking and disturbance rejection Figure A 5 shows the experiment carried out on the threetank system for testing the performance of the controllers in reference tracking both simulation and real testings The PI controllers used for testing are the previously calculated equation A 15 w o o o Qo o O o o o levels mm N N nN m N D eo o o o o T 200 Magl sso TE AT pa aeu AA OA 18 0 L 1 L L L 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 time seconds q1 sim q
14. 39 Vargas H Dormido R Duro N S nchez J Far as G Dormido S Canto M amp Esquembre F 2006a Heatflow Un laboratorio basado en Web us ando Easy Java Simulations y LabVIEW para el entrenamiento en t cnicas de automatizaci n in Proceedings of the XII Latin American Congress on Automatic Control CLCA Salvador de Bah a Brasil Vargas H Far as G Dormido S amp S nchez J 2006b Web based learning resources for automation technicians vocational training illustred with a Heat 9 Bibliography 193 flow and liquid level laboratory in Proceedings of the 7th IFAC Symposium on Advances in Control Education ACE Madrid Spain Vargas H Salzmann C Gillet D amp Dormido S 2009c Remote Experimen tation Mashup in Proceedings of the 8th IFAC Symposium on Advances in Control Education ACE Kumamoto Japan Vargas H S nchez J amp Dormido S 2009b The Spanish University Net work of Web based Laboratories for Control Engineering Education The Au tomatL bs Project in Proceedings of the 10th European Control Conference Budapest Hungary Vargas H S nchez J Dormido S Jara C Candelas F amp Torres F 2010 Docencia en Autom tica Aplicaci n de las TIC a la realizaci n de activi dades pr cticas a trav s de Internet Revista Iberoamericana de Autom tica e Inform tica Industrial RIAI 7 1 35 45 Vargas H
15. A 11 Typical position control system o e 211 xiv List of Figures A 12 PI controller for position control on the DC Motor 211 A 13 Tracking reference in position control DC motor 212 A 14 Tracking reference in speed control DC motor 212 A 15 Step responses at the three sensors when V goes from 3 to 4 volts and a fan fixed flow rate voltage of 3 volts 214 A 16 Tracking reference closing the loop around sensor 1 constant volt age applied to the blower of 3 volts 216 B 1 Three tank hardware in the Automatic Control Lab UNED 220 B 2 Humusoft data acquisition board MF614 and LabVIEW driver 223 B 3 LabVIEW driver for Humusoft data acquisition board MF614 223 B 4 Heatflow hardware in the Automatic Control Lab UNED 226 B 5 Q8 acquisition card used to control the Heatflow System 228 B 6 LabVIEW driver for the Q8 acquisition card 228 B 7 DC Motor hardware in the Automatic Control Laboratory UNED 231 B 8 The NI PCI 6221 Multifunction DAQ 232 B 9 LabVIEW driver DAQmx for NI PCI 6221 233 C 1 Global view of the eMersion environment running the virtual lab oratory corresponding to the three tank system 236 C 2 Accessing to the documentation of the practice 238 C 3 Reduced left and extended right views of eJournal 240 C
16. A university disposes of laboratories Each laboratory has a professor and most likely tutors Students attend courses taught by professors 121 5 2 eMersion A novel approach from EPFL University R Professor ri Professor Teacher Professor 1 Assistant RR Teacher Teacher 1 Assistant Assistant Students Figure 5 3 General abstractions representing the eMersion organization These three key points are the centerpiece in the case of the eMersion design In fact every piece of the environment is built in order to represent each of these concepts Hence the following lines describe how these ideas were introduced in the development of this web based experimentation environment eMersion disposes of a first level of administration see Figure 5 4 A super user or Global administrator manages the system This user can create Spaces Super user Global Administrator Space admin ia Space admin T Space admin TOC O Professors Tutors Professor Tutor Students attending several courses 68680080 Figure 5 4 First level of abstraction in eMersion 122 5 The e Learning Layer which could be seen as an organizational unit of a university as for example a de partment When the Global administrator creates a space a Space administrator is automatically created The internal composition of a space would correspond to a second level of abstraction of eMersion A space administrat
17. DAQ board Process Figure 3 27 Modular view of the JiL Server approach Local control VI to be created by developers acquisition and closed loop control used to set up the connections are provided This library can be easily inte grated into Ejs programs to dialog with the server Similarly on the server side the LabVIEW executable program JiLServer exe operates as a middleware com munication layer between the client and the process Thus developers are only required to create a local control VI that performs the acquisition and closed loop control of the plant 3 3 0 A very simple example of use Now a complete albeit rather naive example is presented to describe the library Consider the flat sequence structure of a generic VI for feedback control of an industrial process as shown in Figure 3 28 In this structure an initial frame labelled Init HW is used for hardware 3 3 A new approach to connect Java and Lab VIEW 69 Acquisition and Closed Loop Control Figure 3 28 Scheme of a generic VI for feedback control purposes initialization code and a final frame Reset HW is used to leave the process in safety conditions when the control operation is finished The central Control Loop wiring diagram of the VI must be designed according to purposes sought control supervision diagnostics identification etc It could for instance implement a timed loop with the control code such as a PID co
18. S nchez J Duro N Dormido R Dormido Canto S Farias G S Dormido Salzmann C amp Gillet D 2008 A Systematic Two Layer Ap proach to Develop Web Based Experimentation Environments for Control En gineering Education Intelligent Automation and Soft Computing 14 505 524 Vargas H S nchez J Salzmann C Esquembre F Gillet D amp Dormido S 20092 Web enabled Remote Scientific Environments Computing in Science and Engineering 11 34 46 Visual NET 2008 Visual Studio Website http www microsoft com spanish msdn latam visualstudio2008 V squez F amp Morilla F 1999 An iterative method for tuning decentralized PID controllers in Proceedings of the 14th World Congress of the IFAC In ternational Federation of Automatic Control Beijing China V squez F amp Morilla F 2002 Tuning decentralized PID controllers for MIMO systems with decoupling in Proceedings of the 15th World Congress of the IFAC International Federation of Automatic Control Barcelona Spain 194 9 Bibliography Williams R 2007 Innovations 2007 World Innovations in Engineering Edu cation and Research International Network for Engineering Education amp Re search NEER pp 279 290 Zutin D G Auer M E Bocanegra J F L pez E R Martins A C B Ortega J A amp Pester A 2008 TCP IP Communication between Server and Client in Multi User Remot
19. The Heatflow system was developed by Quanser Consulting see Figure 4 20 It allows to study concepts of temperature flow control transport lags and identifi cation techniques Quanser 2009 From a practical point of view the graphical representation of the process in Ejs offers some interesting challenges since the temperature flow is not perceptible Blower Heater Sensor 1 Sensor 2 Sensor 3 Vh S1 S2 s3 Figure 4 20 The heatflow apparatus 98 4 Prototypes Developed to the human eye This means that creativity in the development of the user interface is a key issue We will therefore see how Ejs facilitates the work involved by using the predefined graphical elements of the view 4 3 1 System overview The plant consists of a duct containing the following components a heating el ement and a blower located at one end of the structure and three temperature sensors 1 S2 and S3 located along the duct Power delivered to the heater is regulated using an analog signal The fan speed can also be controlled with an analog signal the fan speed is measured using a tachometer Fast settling platinum temperature transducers are used to measure the temperature We have to bear into account that temperature sensors located inside the structure will perceive the effect of the heater according to the distance between the energy suppliers and the sensors Figure 4 21 shows a lateral view of the sys tem illustrating this ide
20. 61 62 65 66 68 70 76 77 78 79 79 81 85 Contents 5 6 4 3 2 Local control of the heatflow system 4 3 3 The virtual and remote laboratory 4 4 Prototype III The DC Motor 4 4 1 System overview ooo ia aa a et 4 4 2 Local control of the DC Motor 4 4 8 The virtual and remote laboratory 4 5 Conclusions eya rs ri eo ncmo m m a we dws The e Learning Layer 5 1 On line learning Pedagogical aspects 5 2 eMersion A novel approach from EPFL 5 2 1 Introduction RR A ee A OR SS 5 2 2 Web technology behind eMersion 5 2 3 eMersion abstractions and concepts 5 2 4 Functional architecture of the eMersion GUI 5 9 Conclusions 2725 a ung tu A ay ener od bee ida a Access Control to Experimentation Resources 6 1 Scheduling access to physical resources 20005 6 1 1 Backgrounds iia ee at PA es 6 2 A simple point to point authentication protocol 6 2 1 User authentication lee 6 2 2 A database for the booking of physical resources 6 2 3 LabVIEW implementation for authentication 6 2 8 1 Integration into the Jib server 6 2 4 User authentication issues in Ejs o 6 3 A flexible scheme for booking and authentication 6 3 1 The automatic bookings system client interface 6 3 2 The a
21. Curso de Doctorado Technical report Department of Computer Science and Automatic Control UNED Madrid Ngolo M Brito L Coito F Gomes L amp Costa A 2009 Architecture for Remote Laboratories based on REST Web Services in Proceedings of the 3rd IEEE International Conference on e Learning in Industrial Electronic ICELIE Porto Portugal Nguyen A V 2006 Web based interaction and collaboration in flexible engi neering education an artifact based approach PhD thesis Ecole Polytech nique Federal de Lausanne Lausanne Switzerland Nguyen A V 2007 Activity theoretical analysis and design model for Web based experimentation n Proceedings of the International Conference on Human Computer Interaction Beijing P R China Nguyen A V Rekik Y amp Gillet D 2006 Iterative Design and Evaluation of a Web Based Experimentation Environment Idea Group Inc pp 286 313 Niederlinski A 1971 A heuristic approach to the design of linear multivariable interacting control systems Automatica 7 691 701 190 9 Bibliography Nitu C I Gramescu B S Comeaga C D P amp Trufasu A O 2005 Op tomechatronic System for Position Detection of a Mobile Mini Robot IEEE Trans Industrial Electronics 52 4 969 973 Open UA 2009 Athabasca University Website http www open au com Oppenheim A Willsky A amp Hamid S 1996 Signals and
22. Ejs view in remote mode Figure 8 6 The Rotoiman system Polytechnic University of Catalunia a Hardware Figure 8 7 Four variable system University of Le n observe the effect of disturbances caused by certain frequencies on the system s output and validate different control techniques as resonant or repetitive control strategies Diaz et al 2009 Costa Castell et al 2010 The four variable system is a multifunctional industrial equipment designed and developed by the Automatic Control Group of the University of Le n Figure 8 7 The plant allows to operate basic and advanced control experiments on the following variables flow pressure temperature cooling and warming circuits and level It also incorporates industrial instrumentation neumatic and electrical actuators for valves with variable frequency in the pumps Guzm n et al 2010 Estado video No video en modo simulacion 2 Aumentada LEVEL PID TEMP PID WEBCAM 40r 20 4 Temp Control Nivel 100 0 20 40 60 80 100 120 s Variador Frecuencia 0 20 40 60 80 100 120 s Nivel 45 590 Variador 63 462 T Proceso 23 085 Val Calenta 0 000 Tiempo 137 000 Val Enfriami 0 000 T Restante 0 000 T Agua Calie 23 940 Conectado l TIEMPO RESTANTE min 4074037 b Ejs view in remote mode
23. Figure 4 2 The three tank system From left to right tanks 7 Ta and T3 are serially connected to each other by pipes voir supplies pumps 1 and 2 with liquid which eventually returns to the system pumps 1 and 2 represent the input flows of tanks T and T gt In this closed system the liquid that enters the reservoir from the tanks returns to the tanks via the pumps However these pumps automatically switch off when the liquid level of T or T5 exceeds a given upper limit In addition to the outflow valve at 75 the system has five more valves Two of them connect two consecutive tanks one for the T to T3 connection through which Q13 flows and the other for the 73 to To connection corresponding to the Q32 flow and can be manually adjusted to close the link between them The other three are leak valves located at the bottom of each tank that can be used to drain the tanks manually Pump flow rates correspond to process input signals the levels of liquid in tanks T and T5 represent the output signals System users can manage all these signals for control purposes The mathematical modelling of the process and the control strategies applied to the system are analyzed in Appendix A Section A 1 The analysis was car ried out to introduce students to multivariable systems control 212 decentralized controller tunning controllers variables interaction and stability However this development is not described in the following sect
24. Status Remote mode whatever cons IET a Hardware b Ejs view in remote mode Figure 8 9 Ball and beam system Polytechnic University of Valencia 170 8 System Assessment with the voltage applied to the motor that is mechanically coupled to its mobile end In this case this system had the server side programmed in C This means the remote connection between Ejs programs Java and C had to be solved Diez et al 2009 Costa Castell et al 2010 The following section presents a brief analysis of the results obtained from this last system evaluation experiment involving the integration of remote labs from other academic institutions around Spain 8 2 3 Analysis of results A collection of statistical graphs summarizes results obtained in the evaluation process Figure 8 10 shows a general view of the level of the students satisfaction during the practical experiences 19 of them answered that they strongly agreed and 69 agreed with the use of the system Moreover other questions about the advantages that students teachers expected when using remote experiments in the educational process were reported The results obtained were that the use of new technologies specially the Internet encourages students to do most of their practical exercises with the help of these resources Learning in relation to traditional methods 8 1 n A 51 Much better Better Equal Less Much less Figure 8 10 S
25. Steps linking a virtual lab with the server side are shown hereafter 3 2 4 Linking a virtual lab to the server side Based on experience acquired throughout of this project the following consid erations are important when Ejs is used to create experimentation interfaces of web based laboratories and LabVIEW is chosen as the application running at the server side 50 3 The Experimentation Layer 1 Java methods for connecting disconnecting sending and receiving data blocks to and from LabVIEW server application must be created 2 Data types in LabVIEW must match Java data types 3 The same GUI should be used to experiment with the laboratory in either simulation or remote mode 4 The graphical user interface must be simple and intuitive The methods mentioned in the first point use TCP sockets to access the com munication layer The package java io net of the Java API is used internally to support the communication Regarding the second point Table 3 1 can be used as a reference for struc turing the data exchanged The selection of the data type for every variable is completed to ensure its minimal payload For example in most cases SGL data type is sufficient to represent floating point information instead of DBL the payload using SGL data is the half than DBL Table 3 1 Casting data types LabVIEW v s Java LabVIEW Data type and payload VENTO RENE Boolean 1 byte true or false boolean Long signed inte
26. The pedagogical aspects to consider when moving from a traditional labo ratory context to a flexible learning model across the Internet The major characteristics that a web experimentation environment should fulfill based on the pedagogical aspects previously mentioned Additionaly the solution adopted to implement a social learning model for UNED control engineering students has been described Some of the main con clusions of this implementation can be summarized as follows A complete description of the web based experimentation environment eMer sion has been presented Each web component that composes the environment has been described separately The reason for that is that in fact an eMersion environment is a set of independent web applications which covers all the most relevant aspects of remote experimentation Chapter 7 will show the integration of the e Learning Layer and the Exper imentation Layer The linking between them will be illustrated by means of a complete remote and virtual control laboratory deployed by eMersion Access Control to Experimentation Resources Overview Sometimes physical resources available in university laboratories do not meet students demands In fact in most cases these have to form workgroups to share the pieces of equipment available in the laboratory However in a flexible learning context through the Internet we need to organize the access to these reso
27. User name and password to access the remote system are sent to the student by e mail once the simulation stage has been accepted by the tutor 2 When user has accessed the bookings system s he must choose a time and date as well as the remote laboratory Figure 6 12 b 3 Afterwards the student must check if the remote plant selected is available on the chosen day Figure 6 12 c 4 The student must select the timetable established to use the remote lab Figure 6 12 d At present students are allowed up to 6 hours per experi ment with the constraints of 2 hours within the same day with a maximum 145 6 3 A flexible scheme for booking and authentication of 4 hours per week 5 The booking system confirms the reservation The user also receives a confirmation email Figure 6 12 e 6 Finally the system shows a bookings list with the registered user Figure 6 12 f Reservations can be removed from the list up to 24 hours before HTML Test Page Mozilla Firefox Enter your username and password Bookings laboratory User estudiante Password eeeeeses0e a Stage 1 HTML Test Page Mozilla Firefox Select date and plant Make a booking TAC 40 7 Sun Tue Wed Thu 0 qm an 2 mm 30 3 2008 En Sat cry of o Motor de corrente continua 2 UNED 8 8 1d n 44 45 46 417 18 25 2 25 Make a booking Eb e voor de comente continua 2 UNED
28. because enhances students motivation their safety and the positive feeling they may experience when conducting remote experimetal sessions On the other hand block c k contains the control parameters susceptible to be modified by distant users The choice of these parameters should be made in accordance with the objectives of the laboratory and the kind of experiments to run Additionally any other kind of information could also be transmitted in 3 2 Adapting virtual labs for remote experimentation 41 X custom header Y measurements X custom header Z video images Y control parameters Lo s k block c k block a State block b Control block Figure 3 9 Structure of exchanged data packets between client and server order to achieve any other specific objectives An exhaustive study about how to modify the size of s k packets changing the compression factor of video images to provide a certain level of quality of service in the network communication is available in Salzmann 2005 To summarize in the field of remote experimentation over the Internet the true challenge is to ensure optimal user s perception of the remote equipment when any user s interaction is carried out Communication between client and server commonly relies on Internet protocols to transport data blocks In this context custom made solutions must be addressed due to the fact that the Internet is a non deterministic best effort network S
29. connection attempts and the information about whether a book ing has been used or not Nowadays only the accesslist table is used while the others are reserved for future usage id timestamp usemame email password starttime endtime valid 1 20090319122105 ohijano oliver_hijano hotmailcom R2FIbXBEcVk 2008 12 13 15 00 00 2008 12 13 15 59 59 false 2 20090319122129 estudiante hyargas bec uned es ZxN dwhRpYw5 z 2008 09 02 15 00 00 2008 09 02 15 59 59 false 3 20090319123224 rvargas hyargas bec uned es om uYTE2NzY 2009 04 01 11 00 00 2009 04 01 11 59 00 true 4 20090319122136 ohijano oliver_hijano hotmailcom R2FIbXBEcVk 2008 12 13 16 00 00 2008 12 13 16 59 59 false 5 20090319122140 chijano oliver_hijano hotmailcom R2FIbXBEcVk 2008 12 13 17 00 00 2008 12 13 17 59 59 false 6 2009031812214 ohijano oliver_hijano hotmailcom R2FtbXBEcVk 2008 12 14 15 00 00 2008 12 14 15 59 59 false Figure 6 4 Bookings registers in the accesslist table of the database Each of the registers in the accesslist table see Figure 6 4 corresponds to the bookings of timeslots made locally by a users administrator for example the teacher assistant of the course In the implementation process we used MySQL server as the database repository MySQL was chosen as it is a widely extended database management system DBMS and mainly because free and open source LabVIEW toolkits can be used to interact with it 1
30. free access to the Target Plant see step 6 and 8 in Figure 6 9 144 6 Access Control to Experimentation Resources w Note The bookings and authentication system architecture presents three interesting advantages from a maintenance point of view It provides a backup of all bookings hosted in a centralized database Thus in the case that a Lab Server with valid bookings is damaged these could be retrieved later by the Lab Server from the Central Server The authentication process being carried out directly with the Lab Server represents a second advantage And thirdly the administrators of a Lab Server can manage the bookings locally In case of problems with the central side bookings can be made manually 6 3 1 The automatic bookings system client interface So far the global architecture of the automatic bookings and authentication system has been described The detailed steps to follow in order to make a reservation are now presented Figure 6 12 shows the user interface of the Applet for bookings whereby students perform their reservations in any experimentation laboratory First of all the process to make a booking implies the student requests a reservation and secondly that the response of the bookings system indicates date and time assigned to use for the remote plant Steps to follow to book are described in detail below 1 Student should be identified by a login and password registered in the database Figure 6 12 a
31. n de om Caracter sticas de la respuesta temporal INTRODUCCI N 1 SISTEMA DE PRIMER ORDEN 1 1 Caracteristicas Si un sistema de primer orden como el de la Figura 1 se excita con una entrad m t gt 0 h aJ i ESO 1 OBJETIVOS DE LA PR CTICA 0 imf1 e 1 gt 0 0 t Los objetivos de est son lo e e Estudio de orden 1y T 1 a una entrada escal n de c Practice guide d Appendix information Figure 5 10 HTML Documentation of the DC Motor experiment way This information should be classified according to its role in the comple tation of these experimental activities Figure 5 10 shows the documentation in eMersion of the DC Motor prototype The prototypes developed in the previous chapter have been fully documented The documents have a homogeneous structure and the content has been written bearing certain practical considerations in mind from a pedagogical point of view The most relevant aspects to take into account when completing tasks can be summarized as follows t Activities Protocol A first document defines a set of tasks or activities that a student should carry out in order to be evaluated effectively by the teaching staff This protocol is divided into two parts PRE Labs activities and Labs activities PRE labs are based on the use of the experimentation console in simulation mode Thus the teaching team is ensured that the student has a previous knowledge of the system before working with
32. nchez H Vargas S Dormido 2007 Web based learning resources for vocational training on Control and measurements systems The Au toTECH Project European Control Conference ECC Kos Greece S Dormido J S nchez F Esquembre H Vargas S Dormido Canto R Dormido N Duro G Far as Ma A Canto 2007 The development of web based virtual laboratories using Easy Java Simulations Ejs Inter national Conference on Remote Engineering and Virtual Instrumentation Rev07 Oporto Portugal H Vargas R Dormido N Duro J S nchez G Far as S Dormido M Canto and F Esquembre 2006 Heatflow Un laboratorio basado en Web usando Easy Java Simulations y LabVIEW para el entrenamiento en t cnicas de automatizaci n XII Latin American Congress on Automatic Control CLCA Salvador de Bah a Brasil G Far as F Esquembre J S nchez S Dormido H Vargas S Dormido Canto R Dormido N Duro and M Canto 2006 Desarrollo de labo ratorios virtuales interactivos y remotos utilizando Easy Java Simulations y Modelos Simulink XII Latin American Congress on Automatic Control CLCA Salvador de Bah a Brasil 14 16 17 18 19 1 Introduction Objectives and Structure H Vargas J S nchez R Dormido G Far as and S Dormido 2006 De sarrollo de laboratorios virtuales y remotos usando Easy Java Simulations and LabVIEW El sistema Heatflow como un caso de e
33. programmer to create a complete remote graphical user interface that accesses an existing LabVIEW VI through the Internet However to non experts in Java graphical programming may be difficult to create this interface To overcome this 3 3 A new approach to connect Java and LabVIEW 71 limitation Easy Java Simulations was granted the ability to create JiL enabled applets using the Jil library in a transparent way As seen in previous sections Ejs allows users to define the model and the view of a simulation in a simple and efficient way The model must provide an algorithmic description of the industrial process didactical setup or physical phenomenon under study This includes a A list indicating its state parameters input and output variables together with their initial values and b Equations that describe how these variables relate to each other evolve with time or change depending user interaction The view must provide a graphical representation of the program output and an interface for user interaction Ejs provides a simplified program structure custom model tools such as an ODEs editor and drag and drop view elements that allow authors to work at a high level of abstraction thus speeding up the creation process Authors input the qualified information on the simulation that only a human can provide and the program takes care of all computer related aspects of creating a finished independent Java applet or a
34. sound or graphic technology capabilities Multimedia demands incredible computing power and only recently in the past 5 years at least affordable computers of this kind are becoming widespread Although one of the most relevant features is the simplicity of the language creating a graphical simulation in Java is not a straightforward task Conceiv ing relatively complex web based applications requires advanced knowledge in object oriented programming and other features of Java Esquembre 2005 For this reason the following subsection presents Easy Java Simulations Ejs in short the software tool used to create the virtual and remote control laborato ries developed in this dissertation and the main reasons why this tool was chosen as the development tool 3 1 1 Easy Java Simulations as a development tool Easy Java Simulations is a freeware open source tool developed in Java specially designed to create interactive dynamic simulations Esquembre 2005 Christian amp Esquembre 2007 Ejs 2009a b Ejs was originally designed to be used by students for interactive learning purposes under the supervision of educators with a low programming level However the user needs to know the analytical model of the 32 3 The Experimentation Layer process and the design of the graphical interface in detail The architecture of Ejs derives from the model view control MVC paradigm whose philosophy is that interactive simulations mus
35. y C Autoplay Comment b Control logic Figure 4 4 Ejs Model of the three tank virtual lab Evolution ure 4 4 b In this case the page shows the switching between manual or automatic control Custom This section is composed of five pages that define a set of customized Java methods ODEs Interface ControlSignal saveFragments and set Language The first page contains the definition of methods that describe the dynamics of the process by ordinary differential equations see Figure 4 5 These functions are called from the page Dynamics of the section Evolution The second and third page define the discrete PID controller and associated methods used to modify control parameters Kp Tj and T4 on the fly The fourth page contains methods used to save the values 4 2 Prototype I The Three tank system 83 ODEs Interface ControlSignal saveFragments setLanguage public double dhif double hi double h3 double A13 double Aperi if hl 0 hi 0 if h1 gt 60 hl 60 if h3 0 h3 0 if h3 gt 60 h3 60 double 013 azl A13 0 5 Sn signo h1 h3 Math sqrt 2 g Math abs hl1 h3 double Qlleak Aper1 0 5 5n coef_leak Math sqrt 2 g h1 return 01 013 Q1leak A public double dh2 double h2 double h3 double A32 double A20 double Aper2 if h2 lt 0 h2 0 if h2 gt 60 h2 60 if h3 lt 0 h3 0 q Comment Figure 4 5 E
36. 2 2 Design of the PID controllers A 3 The Heatflow system Modelling and control A 3 1 A 3 2 Modelling miii e a E Controller design 0 o leen Hardware Description B 1 The Three tank system datasheet o o B 1 1 B 1 2 HTA WATE sa e a pum sed ev es DAQ MF614 Humusoft Multifunction I O B 2 The Heatflow system datasheet 2004 B 2 1 B 2 2 Hardware 3x40 29 da e a a Gea X3 DAQ Q8 Hardware in the Loop H I L Board B 3 The DC Motor datasheet B 3 1 B 3 2 Hard Ware as meum ES Vade A iu d e c s The NI PCI 6221 Multifunction DAQ eMersion Management C 1 How to work with eMersion C 1 1 C 1 2 C 1 3 Introduction sic say end io aras ala Gr General description 0000002 ee eee Interface and eMersion functionalities C 1 3 1 Navegation bar 2 2000 C 1 3 2 Experimentation console C 1 3 3 Working with eJournal 209 213 214 215 219 220 220 222 226 226 228 231 231 232 3 1 3 2 3 3 3 4 6 1 8 1 Al A 2 A 3 B 1 B 2 B 3 B 4 B 5 B 6 B 7 B 8 List of Tables Casting data types LabVIEW v s Java 50 API Java of the Jil class methods to control the connection 67 API Java of the Jil class setter and getter methods
37. 26 Part of the JiL Server wiring diagram that obtains the list of a VI controls and indicators using two Invoke Node blocks When a LabVIEW virtual instrument is published through the JiL server the server s first action is a programmatic reading of all VI controls and indica tors names using the method Control Value Get All in two Invoke Nodes blocks see Figure 3 26 It does this by calling the control value get all method in two invoke node blocks one to get the indicators names and the other for 66 3 The Experimentation Layer the controls names Once the JIL server receives this information it awaits an incoming connection from a Java applet Once the connection is established the server continues to listen for incoming commands from the applet Anytime the JIL server receives a command it is parsed and processed ac cording to its type If the command is a request for the controls and indicators available the JIL server returns the name list as it was extracted when the two invoke blocks obtained it When the user issues a control command s he can start stop or reinitialize the VI via invoke nodes with run VI abort VI or reinitialize all to default methods The server can receive a state com mand which is either to read an indicator or to write a control Either way the corresponding method performs this action in an invoke node block The rest of the time when information is not received the JIL server
38. 79 4 2 Prototype The Three tank system Figure 4 1 corresponds to the DTS200 three tank system manufactured by Amira GmbH Amira 2009 Figure 4 1 The DTS200 three tank system by Amira The three tank system is suggested as a benchmark system used for different purposes It can be used as a test system for fault detection and identification as well as for reconfigurable control Heiming amp Lunze 1999 It was positively welcomed for its interesting properties in both control education and research The system exhibits the typical characteristics of a constrained hybrid system Mignone 2002 and has proven useful to serve as a test bed for algorithms related to state estimation and control or identification of hybrid systems For these reasons the three tank system is used to show the results of different control strategies and as an educational tool in teaching advanced control techniques 4 2 1 System overview The three tank plant consists of three cylinders Ti 75 and 73 with the same cross section A These cylinders are connected serially to each other by cross section S pipes Figure 4 2 shows the plant s full structure The right hand side of tank T has a single outflow valve through which Q o flows that has also a circular cross section Sn The liquid flowing out of the system is collected in a reservoir located under the three tanks This reser 80 4 Prototypes Developed Pump 1 Pump 2 Q Q Q
39. Base64Encoder passNoEncrypted String passEncripted encoder processString return passEncrypted Noo P WNEH 2 To use the new Java method public int authenticate username pass Encrypted of the class Jil class in the jil jar library introduced in Section 3 3 4 When this method is invoked from a client the Java application re ports the user s credentials to the server side that is the username and the password previously encrypted as seen in Listing 6 1 The method returns an integer value indicating the connection attempting result If the value is 1 this means the username and password do not match with any register in the database If the result is 2 then this implies the username and pass word are recorded in the database but the user is attempting to connect outside the timeslot And finally if the result is 3 user s credentials and timeslot are valid and the user is granted access to the remote plant The above logic should be programmed in the client side Java application in order to provide this authentication service to the graphical user interface The term Base64 refers to a specific MIME content transfer encoding It is also used as a generic term for any similar encoding scheme that encodes binary data by treating it numerically and translating it into a base 64 representation 6 3 A flexible scheme for booking and authentication 141 6 3 A flexible scheme for booking and authentication Prev
40. C 46 96 U Heater volt 3 52 Temp S3 C 45 69 Time sec 495 Status Lab Remote mode with labserver Gi connected I TIME REMAINING min 10878 gt b Remote mode with video and the augmented reality option Figure 4 25 The web based laboratory of the heatflow system in remote mode 4 4 Prototype III The DC Motor 103 Video images can then be visualized in the background of the 3D represen tation resulting in a very pleasant augmented reality effect see Figure 4 25 b Students perception is increased as they can actually observe the temperature flow effect inside the structure The results of this study were published in Vargas et al 2006a b 4 4 Prototype lll The DC Motor The DC motor is one of the classical experiments in automatic control labora tories It allows to study the dynamic behaviour with regards to the speed and position of a motor fed by a direct current source The DC motor is a common actuator in control systems and requires therefore although this is a simple and straightforward example special attention given its contribution in the field of education Figure 4 26 shows the didactical equipment used and its hardware components Results were published in Vargas et al 2008 LENTOS Figure 4 26 The Direct Current Motor didactical equipment used in the laboratory 4 4 1 System overview The didactical equipment is an electromechanical process intended
41. Command Parser remains in sleeping mode leaving the processor free for other duties Similarly the Sender sends the measurements acquired by the Control Loop to the client side when required by a command Analyzing the server side of Figure 3 23 it is clear that a separation line can be drawn between the two communication tasks gray boxes and the Control Loop red box which is the task connected to an industrial process or didactical 64 3 The Experimentation Layer setup In this scenario the Control Loop is a LabVIEW VI able to supervise diagnose or control actions in principle with no TCP IP communication capa bilities To provide a convenient TCP IP wrapping to the Control Loop the JiL server is introduced in the scheme This middleware plays the role of both the Command Parser and Sender blocks of Figure 3 23 and communicates the Control Loop VI with a remote Java client in an effective manner To publish a VI using the JiL server the author first starts the JiL server program and uses its control panel the File menu of Figure 3 24 to select the local VI to be published Pressing the Start button finishes the publishing task Every control and indicator of the VI becomes immediately accessible to any applet that uses the provided Java library file see next subsection Jil Server 1 0 File Options Help m Is there any user connected Start Connection report SUE EI ONES S SEO eduroam dhcp 10 211 13 23
42. Input B 2 The Heatflow system datasheet 227 Table B 3 Dimensions and power requirements Table B 4 Data acquisition requirements Data Acquisition Requirements Analog inputs 0 5V DC 50 x 15 x 10 Analog outputs 0 5V DC Calibration Occasionally it is necessary to calibrate the sensor measurements To do that a measurement of ambient temperature and a DC Voltmeter are required 1 Power up the system 2 Open the service panel as shown 3 Measure S1 and adjust potentiometer P1 such that the voltage is propor tional to the room temperature 4 Measure 2 and adjust potentiometer P2 such that the voltage is propor tional to the room temperature 5 Measure 3 and adjust potentiometer P3 such that the voltage is propor tional to the room temperature Table B 5 shows the voltages that should be obtained at various temperatures Table B 5 Calibration of the equipment 228 B Hardware Description B 2 2 DAQ 8 Hardware in the Loop H I L Board Figure B 5 shows the acquisition card used to control the Heatflow System The Q8 is an innovative H I L control board with an extensive range of input and output supports A wide variety of devices with analog and digital sensors as well as quadrature encoders are easily connected to the Q8 This single board solution is ideal for use in control systems and complex measurement applications wu Qo 40 a Q8 Hardware in the Loop H I L Board
43. It would be very difficult to enumerate the knowledge teach ing and experience he has taught me He always provided me with sound advice support encouragement guidance and many other things which have changed my way of seeing and dealing with my personal and professional life Thanks Sebasti n for giving me the opportunity to be one more in your research team I have felt a really lucky man during all this time I 1 II 3 Contents PRELIMINARIES Introduction Objectives and Structure 11 Introduction 4 i49 oc ouod OY a ee 12 ObjectiVes xc ee de be Se ada ee le RSS res 1 9 Outlines xd REESE ee ska xx uic 1 4 Publications Awards and Projects a Environment Global Architecture 2 1 Planning the structure of the system o o 2 2 A systematic two layer approach o o o o 2 2 1 Layer 1 The experimentation layer 2 2 1 1 The client Requirements and specifications 2 2 1 2 The server Requeriments and specifications 2 2 2 Layer 2 The e learning layer 0 ooo 2 3 Cl ncluslornss a A E Et A els Da ANA IMPLEMENTATION The Experimentation Layer 3 1 Development of virtual laboratories 17 18 19 19 20 21 23 25 27 29 i Contents 3 1 1 Easy Java Simulations as a development tool 3 1 2 Structure of a generic virtual control lab in Ejs 3 2 Adapting virtual labs for remote experimen
44. Methods of the API setValue String name boolean value Sets a BOOL variable to LabVIEW setValue String name int value Sets a 132 variable to LabVIEW setValue String name float value Sets a SGL variable to LabVIEW setValue String name double value Sets a DBL variable to LabVIEW setValue String name String value Sets a STR variable to LabVIEW boolean getBoolean String name Gets the boolean value from LabVIEW int getInt String name Gets the integer value from LabVIEW float getFloat String name Gets the float value from LabVIEW double getDouble String name Gets the double value from LabVIEW String getString String name Gets the string value from LabVIEW Figure 3 27 summarizes the JiL Server approach within the overall infrastruc ture The framework relies on the use of two software tools particularly useful to develop the experimentation layer LabVIEW and Ejs The approach is based on the creation of generic communication modules both on the client and the server sides On the client side the Java library jil jar was created By this API the TCP protocol is hidden to users and instead simpler Java classes methods 68 3 The Experimentation Layer USER client BROWSER applet Java created by Ejs applet jil jar must be imported into Ejs program Internet JiL Server Communication layer Middleware layer REPAS VI Server functions VI Server Real Time LabVIEW VI
45. Mozilla Firefox E Objective a Control de nivel en el sistema de tres tanques Navigation Current Task 34 Awareness lt r a ii Tasks in progress F1 Mx ag ral pa zs fl E Li Gi BEI Zir E Password http lab dia uned es 8080 Home page for threetank Mozilla Firefox E ETE joj x eJournal Control Language H1 H2 H3 Setpoint oe H2 H3 Setpol T T T T Default M Li Eb Active Journal Trash 400 4 E 300 Folders 200 ry inbox an zl Sin o Active Folder Filter by Type Filte 0 100 200 300 40 50 seconds Fragments Pump1 Pump2 Copy Move Delete Rename Import Expor T T T 100 T Name Author 80 4 T madridm Estudiante Status video No video in simulation mode O gp 4 CONTROL PID VIDEO i op T portugalm ES A13 100 A32 100 A2 100 20 RUE HEN CR iln oue T HectorS3rem m Estudiante Y O 0 i i L T HectorS3simm Estudiante Aleak196 0 Aleak296 0 Aleak396 0 0 100 200 300 400 500 n HectorHoySIM2 m Estudiante Q Q seconds r f Estudiante Pump1 20 Pump2 43 Fps 20 H1 mm 300 09 H2 mm 200 03 HectorHoyRem m studian Pump 36 19 57 Pump 36 43 25 F HectorHoySim m Estudiante m 5 A SP1 mm 300 SP2Imm 200 I HectorHoy m Estudiante 2 Time sec 585 H3 mm 250 03 re
46. PID controller Simulink model 88 Discrete PID controller with antiwindup protection programmed in LabVIEW lt a dd AA 89 Linkage of the Ejs and LabVIEW variables 91 Method to be launched when connect button is pushed 91 The _external step method is called in every iteration of the Evolution during the connection 92 The _external synchronize method is called when user mod ifies the value in pumpl manually sss 92 The WebCamImage appears at the top of the hierarchy of the DrawingPanel container to enrich the view with augmented reality 93 The three tank system virtual and remote control laboratory 94 The virtual and remote control lab of the three tank system 95 Connection scheme via web browser between Ejs and LabVIEW The Ejs application controls the threetankReal vi through the JiL SOLVERS att des Sake Mae ge Rob nei Re Rees alas be ee ees 96 The heatflow apparatus c 97 Variables of the apparatus 2222s 98 Local control of the heatflow real system Front panel 99 Local control of the heatflow real system Block diagram 100 The web based laboratory of the heatflow system in virtual mode 101 The web based laboratory of the heatflow system in remote mode 102 The Direct Current Motor didactical equipment used in the labo TAO A ed ee a Ad ee ee RUN 103 The DC motor scheme o e e 104 L
47. Salzmann 2005 the author works using this option to implement algorithms capable of adapting the quality of the images received by modifying a compression factor by the client according to changes in the quality of service in the network communication 3 2 Adapting virtual labs for remote experimentation 57 3 2 5 1 Development of the Java library webcam jar In order to read the video streaming published by the video server of an IP camera a Java library called webcam jar was developed This API allows Java programmers to use the classes and methods of the library to get video images from any IP camera connected to the Internet The software provides a first level of interface with the camera allowing to write high level software hiding the low level programming details such as the structure of the received packets the decoding of the HTTP headers or the sockets implementation The main class of the API called Video class includes a constructor that allows to create an instance of a Video object with the following input arguments a string indicating the URL where the camera supplies the video stream a boolean indicating what the reading format either MPEG true or JPEG false and finally an integer to include a delay in milliseconds in the reading of the images A detailed description of this class and its methods is provided below public Video String url boolean isMPEG int delay When an object of this class is cre
48. Systems 2nd Edi tion Prentice Hall Signal Processing Series OU 2009 The Open University Website http www open ac uk OUA 2009 Open Universities Australia Website https www open edu au wps portal Pastor R Mart n C S nchez J amp Dormido S 2005 Development of an XML based lab for remote control experiments on a servo motor International Journal of Electrical Engineering Education 42 2 173 184 PHP 2009 Website http www php org Quanser 2009 Website http www quanser com React 2009 Real Time Coordination and Distributed Interaction Systems Group EPFL Website http lawww epfl ch pageb5358 html Riva G 2001 Learning and Teaching on the World Wide Web Academic Press pp 131 148 Rosen M A 2007 Innovations 2007 World Innovations in Engineering Edu cation and Research International Network for Engineering Education amp Re search NEER pp 1 11 Salzmann C 2005 Real Time Interaction over the Internet Quality of Service and End to End Adaptation for Remote Experimentation PhD thesis Ecole Polytechnique Federal de Lausanne Lausanne Switzerland Salzmann C amp Gillet D 2002 Real time interaction over the Internet in Proceedings of the 15th IFAC World Congress Barcelona Spain 9 Bibliography 191 Salzmann C Gillet D 2007 Challenges in Remote Laboratory Sustain ability in Proceedings International Confere
49. They can use tools such as Matlab Simulink or Scilab to perform these activities 9 2 Future work 179 currently This means we can close the instruction cycle the exper imentation phase the analysis process and the design by using just the web browser avoiding the use of standalone applications or the purchasing of licenses To develop a web module to eMersion that includes awareness tools Students could then track their own progress and they could compare the progress they have made with other classmates This develop ment would also allow to balance the workload of the system that is students could start their work earlier compared to other students avoiding the high demand of students at the end of the course Bibliography Adamo F Attivissimo F Cavone G amp Giaquinto N 2007 SCADA HMI Systems in Advanced Educational Courses IEEE Trans Instrumentation and Measurement 56 1 4 10 AJAX 2009 Ajax Website http www php org Amira 2009 Amira GmbH Website http www amira de Anderson P 2007 What is Web 2 0 Ideas technologies and impli cations for education Technical report JISC Technology and Standards Watch http citeseerx ist psu edu viewdoc download doi 10 1 1 108 9995 amp rep repi amp type pdf Anderson T amp Elloumi F 2004 Theory and practice of online learning Athabasca University Copyright Creative Commons Online book Apache Logging 2009 A
50. a magnetic brake The engine is fed through a unity gain amplifier capable of delivering the power required to the motor A tachometer provides a voltage proportional to the angular velocity of the load A potentiometer led through a reducer device provides a voltage propor tional to its angular position A similar potentiometer is also available to perform a eventual change of the reference directly in the equipment Tables B 6 B 7 and B 8 show the technical features of the servo mechanism Table B 6 Motor parameters Parameter ombo value 0 030 N A Inertia tachymeter motor 24x107 kgm 59 19 382367 Nmo rad 232 B Hardware Description Table B 7 Steel disk parameters small medium large Masse kg Inertia Ja kgm 74x10 342x1076 1440x10 Table B 8 Other parameters Parameter ome B 3 2 The NI PCI 6221 Multifunction DAS The National Instruments PCI 6221 is a low cost multifunction M Series data ac quisition DAQ board optimized for cost sensitive applications M Series devices are compatible with the following versions or later of NI application software LabVIEW LabWindows CVI or Measurement Studio versions 7 x and Lab VIEW SignalExpress 2 x Figure B 8 The NI PCI 6221 Multifunction DAQ Low cost M Series devices incorporate advanced features such as the NI STC 2 system controller NI PGIA 2 programmable amplifier and NI MCal calibration technology to increase p
51. and use of the physical resources of the remote laboratory This development allows for a rational use of resources an essential issue to address in most universities where the number of didactical setups to attend the students demand is insufficient The global architecture of the system has been thought in such a way that any Lab Server from different locations can use this service Chapter 8 will present two pilot experiences carried out with the on line experimentation system devel oped in this dissertation Both use the automatic bookings system The first one uses it with the physical systems located in the Department of Computer Science and Automatic Control of the UNED and the second one uses the resources of different universities around Spain This dissertation considers the automatic booking and authentication system as part of the e Learning Layer This is because the following chapter will show how the bookings system was completely integrated into eMersion as an External Toolkit that belongs to the own experimentation environment Integration of Layers Overview In this chapter the integration of the Experimentation and e Learning Layers is addressed Low level linkages among different pieces are required to ensure the appropriate functioning of the full environment The eMersion environment plays a key role in this part of the dissertation since this application could be considered as a wrapper of the other components Th
52. approach to multivariable controller design in which the matrix transfer function G s is a square plant 2 x 2 to be controlled with a diagonal controller see Figure A 4 204 A Modelling and Control h1_setpoint h2_setpoint Figure A 4 Descentralized control strategy G s is the three tank system to be controlled C s is the controller to be designed and is composed of two elements C1 and C2 Ci controls the level in tank i The variables hi_setpoint ei ui and hi are the setpoint to be reached the error signal the control signal and the liquid level in tank i respectively This strategy works well if G s is close to diagonal because the plant to be controlled is essentially a collection of independent subplants and we can design each element in the controller C s independently as indicated in Equation A 13 C s diag C1 s C2 s A 13 A classical and popular technique for designing C1 and C2 is via proportional integral derivative PID control A standard PID controller also called a three term controller has a transfer function C s given by C s Kp 1 x Tas A 14 where Kp is the proportional gain T the integral time and Tq the derivative time For the three term functionalities the proportional term provides an overall control action proportional to the error signal through the all pass gain factor the integral term reduces steady state errors through low frequency com
53. appropriate learning model the concepts mentioned in the case of third methodology were adopted The solution relies on the use of eMersion Gillet amp Fakas 2001 Gillet et al 2005 Nguyen 2006 a web based application which somehow implements a social learning model focusing on the on line teaching of students in subjects with a high technical content The following section presents the advantages of this software tool in the context of remote experimentation 5 2 eMersion A novel approach from EPFL 5 2 1 Introduction The implementation of the e Learning Layer is based on eMersion a web based experimentation environment created by the research group ReAct Real Time Coordination and Sustainable Interaction System Group React 2009 from the Automatic Control Laboratory of EPFL Switzerland Remembering the words of the developers in Gillet et al 2003 to explain how the eMersion environment was conceived the authors state the following Carrying out web based experimentation is a matter of observing and acting on a virtual model or on a real equipment using convenient visualization and control devices Hence a student enrolled im such activities can be seen as a pilot siting im the cockpit of an exploration vehicle to complete a mission In a web based experimentation frame work the given mission is tipically a laboratory assignment and the cockpit is a computer To take advantage of this similarity a general cockpit me
54. are initially filled out manually by the administrators of the laboratories However Section 6 3 will show an automatic mechanism used to populate them transforming this simple authentication protocol into a custom automatic bookings system for re mote experimentation 6 2 A simple point to point authentication protocol 135 A simple database aimed at hosting the bookings was created The database contains five tables accesslist accesslisthistory labstatus labstatushistory and loginresult Figure 6 3 shows the structure of the tables in the database tid int timestamp timestamp tid int timestamp timestamp status varchar timestamp timestamp username varchar tusername varchar temail varchar temail varchar password varchar password varchar startime datetime starttime datetime endtime datetime endtime datetime valid varchar tused varchar id int labstatushistory timestamp timestamp sessionID varchar tid int result varchar timestamp timestamp status varchar eee A loginresult enum true false Figure 6 3 Database to host bookings The most important table is the accesslist because as it hosts bookings This table acts as a middleware layer between the user connection attempts and phys ical resources The other tables can be used to record statistical information about the bookings the status of the lab ONLINE OFFLINE RUNNING or UNKNOWN
55. available nowadays With regards to this aspect the present thesis provides a structured framework to develop remote experimentation systems using three soft ware tools Easy Java Simulations Ejs 2009a LabVIEW LabVIEW 2009a and eMersion eMersion 2009 Since the UNED is a distance learning university the introduction of this kind of educational methodology is highly beneficial for students who must conjugate 8 1 Introduction Objectives and Structure work and academic life In this context capturing students perception of the sys tem developed is paramount so as to improve it and include new functions This dissertation also analyses the remote experimentation environment developed and provides a series of guidelines and suggestions on how to use the system and for future works 1 2 Objectives The first objective of this paper is to describe the background needed to im plement web based environments to carry out hands on laboratories through the Internet in the field of engineering education The overall aim is to suggest a two layered framework to create remote and virtual control laboratories and the web based tools required to publish them on the Internet The second objective is to provide a new approach to create virtual and remote laboratories for control education The approach suggested is based on the use of two software tools specifically chosen for this purpose Easy Java Simulations and LabVIEW the idea being
56. available only The reason behind the designing of an automatic bookings system is to manage user access in order to better use the physical resources The sixth and final objective is to evaluate the developed remote experimen tation environment from the student s point of view Two pilot experiments were carried out in order to analyze the positive and negative impact of this new way of completing the practical activities on the students 1 3 Outlines This thesis has been structured as follows Chapter 2 The typical application scenario of a remote experimentation system for pedagogical purposes is presented and discussed This chapter also introduces the approach to follow giving some work guidelines in order to develop hands on web based laboratories Chapter 3 The methodology used for the implementation of virtual and remote control laboratories is presented in this dissertation The chapter describes the JiL Server approach a structured framework that enables to create the virtual and remote labs mentioned in this thesis by using both Easy Java Simulations and NI LabVIEW Chapter 4 This chapter shows how to build virtual and remote control labo 10 1 Introduction Objectives and Structure ratories using the approach proposed in Chapter 3 Additional information on hardware modelling and control of the labora tory setups used in this work is provided in Appendices A and B Chapter 5 Various teaching models
57. b Terminal board Q8 Figure B 5 Q8 acquisition card used to control the Heatflow System qx Hardware In The Loop LU aA am E d Configuration Tasks Utility Ar Br Br Br pan Analog Encoder Inputs Digital General Figure B 6 LabVIEW driver for the Q8 acquisition card The Q8 is a versatile and powerful measurement and control board with an extensive range of input and output supports The card can be accessed from LabVIEW programs by installing the driver see Figure B 6 Every Q8 is supplied complete with ribbon cables and a Terminal Board with standardized connectors simplifying your hardware integration The Q8 has 8 single ended analog inputs with 14 bit resolution sign extended to 16 bits in hardware With A D conversion times of 2 4 us channel simultaneous sam pling and sampling frequencies of up to 350 kHz for 2 channels the Q8 can handle even your most demanding performance requirements All 8 channels can be sampled simultaneously at 100 kHz B 2 The Heatflow system datasheet 229 The Q8 is equipped with 8 analog outputs with software programmable volt age ranges and simultaneous update capability with an 8 usec settling time over full scale 20 V Aside from having 8 x 24 bit single ended quadrature encoder inputs using TTL CMOS the Q8 uniquely supports simultaneous sampling of all eight encoder channels 32 lines of individually programmable digital I O with Totem pole out p
58. be carried out so as to clarify teaching methodology learning resources and objectives This in turn will determine both technical and organizational aspects Over the last years the Department of Computer Science and Automatic Con trol of the UNED has been placing a great deal of effort into the analysis design development and exploitation of virtual and remote laboratories for automatic control teaching and learning Aranda et al 1998 S nchez 2000 S nchez et al 2002 Dormido 2004 S nchez et al 2004 Pastor et al 2005 Dormido et al 2005a S nchez et al 2005 Duro et al 2008 However despite all this work there is no appropriate environment as of yet that provides remote experimentation services through Internet to students within a holistic and structured framework This dissertation tries to go one step further in this direction by implementing a com plete web based integral environment to perform remote practical experiences for automatic control courses On the other hand a mix of web based technologies and software agents Salz mann amp Gillet 2008 2007 is commonly used to develop remote experimentation systems designed for pedagogical purposes Incidentally most recorded devel opments aimed at creating remote experimentation systems are custom made solutions This means their conception selection of software tools and global system architecture are not simple tasks because of the wide variety of software tools
59. camera can be obtained from the input buffer by using this method The method returns an object of the java awt Image class Image objects are easily displayed on Java swing components Summarizing Java programmers can import this library to capture the stream ing video of any IP camera connected to the network With regards to the present essay the library was developed in such away that its inclusion in the develop ment process of a new virtual and remote laboratory can be achieved with a minimum programming effort Following this philosophy the library webcam jar was integrated into Ejs as a new view element Then section presents this new component and how it can be used by Ejs users 3 25 2 The WebCamimage view element in Ejs In order to simplify the use of the webcam jar library a new view element of Ejs called WebCamImage was developed The novelty of this approach lies in that the element was implemented based on the characteristics of an existing graphical component of Ejs the Interactivelmage Thus the WebCamImage element inherits all methods and properties of this last object and can therefore be placed on an Ejs container with coordinate axis such as the DrawingPanels or the PlottingPanels in order to render the images from the camera Figure 3 21 shows how to use the WebCamImage element in an Ejs appli SSwing is a widget toolkit for Java It is part of Sun Microsystems Java Foundation Classes JFC an API able
60. do not give us all the elements needed to provide remote experimentation services in a pedagogical context complementary web based resources are needed to manage the learning process of students Thus the so called e learning layer implies the development of the functionalities required to support the teaching and learning of students through the Internet Vargas et al 2008 2 2 1 Layer 1 The experimentation layer The experimentation layer wraps the design methodology and building of the graphical user interface of the client side and the server application which links to the real plant in the lab The development process of this layer is addressed by means of the well known client and server structure Figure 2 2 illustrates the general scheme of the experimentation layer based on this communication architecture From a software engineering point of view the development process begins with the analysis of requirement and specifications of implementation The following subsections indicate some recommendations to follow on how to develop this layer camera points to plant e g didactical setup in the laboratory Public IP Private IP 2 eth usb interface NA Client Internet Lab Server IP network Figure 2 2 Experimentation layer General scheme 20 2 Environment Global Architecture 2 2 1 1 The client Requirements and specifications When we talk about the graphical user interface on the client
61. ex perience involving more students and university groups The AutomatL bs project presented in Chapter 8 is the most recent research result and it demonstrates its benefits over the last three academic years in the universities that took part in the project Results from the previous evaluation will let us debug and improve the framework in different directions First the number and variety of physical processes will be increased by enrolling new uni versities to the AutomatL bs project Also we will try to let students carry on with their practical experiences using other devices mobile phones PDAs and user interfaces e mail web forms HTML Javascript thin interfaces etc 9 2 Future work Finally some ideas about possible extensions of this work are the following 1 The notion of smart objects discussed in Thompson 2005 Salzmann amp Gillet 2008 Johnson et al 2009 is pursued in the following two points This idea can be applied in remote experimentation topics by increasing the intelligence of the server side applications i e by extending the capabilities of JiL Server 178 9 Conclusions and Future Research To include new functionalities to the JiL Server application in order to improve the maintenance of laboratory equipments In this sense administrators could benefit from implementing automatic fault noti fication mechanisms via email or mobile messages statistic tracking of connections state notifi
62. generic module without real time constraints given that com munication functions are inherently non deterministic by existing delays in the network and the closed loop control module whereby real time execution must be guaranteed In this case a local control VI could be created using some of 4 4 Prototype III The DC Motor 105 the technological alternatives that National Instruments provides to LabVIEW users to develop applications with real time restrictions for example NEDAQ with dedicated processor to real time tasks whereby the local control VI can be embedded into card itself or any other more robust solutions Figures 4 28 and 4 29 show the application motorReal vi that locally controls the DC motor The sampling period used in the control loop is 20 ms Input Rate of 50 Hz in Timing amp Triggering Parameters of the Front Panel Figure 4 28 Local control of the DC motor real system Front panel To avoid overflow in the communication channel due to the delivering of data packets to the client the JiL server is configured to send data every 100 ms that is the control task runs at a rate of 20 ms but the JiL server retrieves and sends the client details of the state of the motor every 5 steps of evolution in the real time control task Although some data are lost this sending rate is high enough to fluently observe the dynamic behaviour of the process in the client interface see Figure 4 30 INIT
63. hiit 2004 10 pdf LabSQL 2009 Jeffrey Travis Studios Website http jeffreytravis com lost labsql html LabVIEW 2009a Website http www ni com labview LabVIEW 2009b Command based Communication Using Simple TCP IP Mes saging NI Website http zone ni com devzone cda tut p id 3098 Leva A amp Donida F 2008 Web enabled laboratory on the implementation of industrial controllers International Journal of Electrical Engineering Edu cation 45 1 72 91 Lim D 2006 A Laboratory Course in Real Time Software for the Control of Dynamic Systems IEEE Transactions on Education 49 3 346 354 London External 2009 University of London Website http www londonexternal ac uk M Series 2009 NI DAQ M Series Website http sine ni com nips cds view p lang en nid 14114 9 Bibliography 189 Maharg P 2004 Virtual communities on the Web Transactional learning and teaching Wolf Legal Publishers Mart n C 2007 Object Oriented Modelling of Virtual Laboratories for Con trol Education PhD thesis Universidad Nacional de Educaci n a Distancia Madrid Espa a Matlab 2009 Mathworks Website http www mathworks com Mignone D 2002 Estimation and Control of Hybrid Systems Technical report Automatic Control Laboratory ETH Z rich Modelica 2009 Modelica Association Website http www modelica org Morilla F 2005 Introducci n al control multivariable
64. resources needed to use remote labs The system users clients can access experimentation services available from anywhere providing there is an Internet connection The above mentioned describes in a few words the remote experimentation context in engineering education However developing a complete environment able to provide such experimentation services is not an easy task For this reason the next section presents a systematic approach to follow to develop a system with these characteristics wired clients Common IP Core S Network gt Remote Experimentation Services Remote Lab University i Main Server wireless clients res R Server iN 23 pesi P Lab Server didactical setup clients from home didactical setup Figure 2 1 On line experimentation system Global architecture 2 2 A systematic two layer approach 19 2 2 Asystematic two layer approach Although the development responsibilities can be clasified in multiple ways in this dissertation we have divided the problem into two levels or layers The first layer is the so called experimentation layer that includes all the software and hardware components needed to develop the experimental applications that is the web based virtual and remote laboratories As web based labs alone
65. side we refer to the set of controls indicators graphical elements and visualization components which all constitute the application the users will use to complete their exper iments In this sense the design and building process of these interfaces must follow certain basic rules to simplify and facilitate their understanding A set of characteristics that should be used as design options for the client side interface are described hereafter The software used for developing should be multiplatform Java shows the expected characteristics for designing this kind of applications as it provides various features as the platform independence and the simple accesibility from any web browser without additional software only the runtime engine JVM is needed This means the user only needs a web browser with Java support to access the web based lab The communication protocol used to communicate with the server side should be low level protocols to stream data through the Internet TCP Transfer Control Protocol or UDP User Datagram Protocol They al low for a better control of data packet transmisions in networks based on IP protocol such as the Internet The graphical user interface must prove simple and intuitive It should be user friendly and it should also have been tested successfully in different environments Access to the laboratory in either virtual or remote version should be en abled using the same g
66. that it will facilitate the creation of web based labs to new developers and or educators who wish to include this new learning paradigm in their curricula The third objective of this dissertation is to create a set of virtual and remote control laboratories by following the framework developed step by step Three didactical setups located in the Automatic Control Laboratory of the UNED will be used to develop the three web based labs a threetank system a temperature control system and a DC motor The developed prototypes will demonstrate the validity of the approach described in this work The fourth objective is to provide and implement the features that a flexible learning environment should offer to sustain an on line community of web based labs The aim is to describe the additional web based resources needed to support 1 3 Outlines 9 the learning process effectively The collaborative web based application eMer sion specifically designed to organize and enable engineering practical activities on the Internet will be used for this purpose The fifth objective is to develop a web based automatic bookings system to ar range user access of physical resources available in remote laboratories As is the case in most academic institutions the number of didactical setups available in laboratories is insufficient and does not match the demand of students For cost reasons it is possible in some cases to have one or two pilot plants
67. the real process On the other hand Labs tasks are based on the use of the experimentation console in remote mode Remote access to the system is enabled by the teaching staff once the work in simulation mode has been evaluated satisfactorily 5 2 eMersion A novel approach from EPFL 127 ts Practical guide for the experimentation console A second document describes each graphical element of the console Thus students can use this information as a starter s guide w Modelling and control of the process under study A third docu ment provides theoretical and practical information about the process under study such as the mathematical model of the system the basic theory of control and identification techniques or any additional information t Appendices It contains any additional information useful for students For example general information common to several modules datasheet of the hardware used reference tables etc Integration of external web applications eMersion also offers the possibility of integrating third party web based applica tions to the environment Thus the functionality of the system can be increased according to needs For example during an on line experimentation session stu dents can collect data registers that will be analyzed with specific software tools such as Matlab Simulink Scilab SysQuake etc eMersion allows to incorporate toolkits to access remotely through their URL Figure 5 11 show
68. the format of variables to exchange both client and server know beforehand what variables will be sent and received during con nection Thus when run method is invoked a private method called from Receiver thread of the Jil class reads all the values from the JiL server into the internal list and places them in a buffer from which the getValue method retrieves their values Similarly to modify the value of one or more VI controls the suitable setValue methods must first be called to set the new values in a buffer This buffer is read by a private method called from Sender thread of the Jil class which when all values have been set transmits the new values to the JiL Server Table 3 2 API Java of the Jil class methods to control the connection VI Control Methods of the API Jil String serverAddress Constructor used to create a JiL object void connect Creates and initializes the TCP IP connection with the JiL server void disconnect Closes the TCP link with the JiL server boolean openVI String modelPath String initCmd Opens a VI specified by modelPath and reports the exchange variables by initCmd void closeVI Closes a refnum associated with an openned VI void run Starts the VI A java io IOException is thrown if an I O error occurs void stop Stops the VI A java io IOException is thrown if an I O error occurs Table 3 3 API Java of the Jil class setter and getter methods Setter and Getter
69. the second one is used to write the Java code needed to get the PI control action and the switching between manual and automatic control modes The right side of the ODE is defined by a custom method as shown in Figure 3 5 f 38 3 The Experimentation Layer Once the model has been defined the composition of the view is carried out Figures 3 6 a and 3 6 b show both the tree like structure of the view in Ejs and the final look as a standalone application of the single tank virtual lab Introduction Model View B Single tank control DOS Bl Graphs Tree of Elements El Elements for the view Tank level e Simulation View EA El Frame Containers Interface Menu pd Ead DrawingPanel ii TheTank e om Piperromtank PipeOutputFlow Wi TankBasin a Fda setPoint AeA T sp PipeTovaive Drawables f vave Basic Bodies Fields y PipeFromvalve 100 150 200 250 300 350 pipelnputFlow e SNL T N tole LOrirele T LCText a from f a IZ N tovatve T SPText Other A tromse Control 30 r A A O Controls Play Pause Reset Manual 8 Automatic O Panelup LE ro B 2eBr R 0025 r 00015 sp 0082 u 2638 B Dialog de PlottingPanel P r y v Level w ES SetPoint a View in Ejs b The final standalone application Figure 3 6 Defining the VIEW of the s
70. time with almost any existing software and hardware via DLLs Active X NET TCP IP Ethernet Bluetooth CAN OPC IEEE 1394 RS232 485 GPIB etc as well as with other scientific packages such as Matlab Simulink or Mathematica However once a LabVIEW VI has been de veloped implementing a bidirectional data exchange with a Java applet requires editing the existing VI wiring diagram to include new TCP IP communication blocks into the diagram It also requires implementing the TCP IP based com munication library calls in the Java client This procedure is time consuming and requires both knowledge of Java and an understanding of how TCP IP com munication works But also maintainability and scalability of legacy code is compromised since the author of the original code is not always available to carry out these modifications Web enabled environments for remote diagnostic control or experimentation are commonly based on client server architectures where TCP IP links are used to exchange data and commands between both sides Figure 3 23 depicts the design pattern known as command based architecture LabVIEW 2009b that is the base of the JiL approach Command based architecture is focused on high performance applications where both the minimization of the network traffic and the ability to manipulate different data types using a naming convention to the variables represent the fundamental objective The communication protocol cov ers the fol
71. to the Bookings Main Server step 3 in Figure 6 9 3Greenwich Mean Time GMT is the world wide reference time to calculate countries time zone around the world 6 3 A flexible scheme for booking and authentication 143 e The Bookings Main Server calculates the time lag and amends the timeslot f The Bookings Main Server reports the new register to the Java Inter face see step 4 in Figure 6 9 g The Java Interface receives the register and inserts it in the Lab Server DB step 5 in Figure 6 9 h The Bookings Main Server tells the client that the new reservation has been done see number 1 in Figure 6 9 Authentication Stage A state diagram of the authentication process is de picted in Figure 6 11 b Figure 6 11 State diagram of the authentication process A detailed description of the state flow during authentication is as follows a The Applet for experimentation starts the process by sending user credentials see step 6 in Figure 6 9 b The Identity Checking Module receives the keys and checks whether user exists in local DB Should this be the case it then checks whether the connection attempt is between the starttime and the endtime of the timeslot reserved see step 7 in Figure 6 9 c The Identity Checking Module sends the result of the checking to the Applet for experimentation see step 6 in Figure 6 9 d If the checking result is ok then the Applet for experimentation gets
72. 0 stu dents enrolled in the Physics degree program at the UNED for which the process control laboratory is a compulsory subject The evaluation questionnaires re turned at the end of the term showed the opinions of the audience who rated the different statements of the survey as strongly agree agree neutral disagree or strongly disagree These answers are summarized in Table 8 1 The analysis of the evaluation results showed that the pilot experience was a very promising achievement The experience also helped to extract some in formation about the positive and negative aspects of a remote experimentation system for teaching in the field of engineering Dormido et al 2008 8 2 The AutomatL bs project 163 Table 8 1 Student questionnary results 2006 2007 UNED Sub scale ASA Y A YN D SD NA Learning value 55 36 3 eZ O 0 0 Value added 2 5 25 15 0 0 0 Design and usability 38 8 55 5 5 7 0 0 0 Technology function 53 3 13 3 133 0 0 20 1 SA Strongly Agree D Disagree A Agree SD Strongly Disagree N Neutral N A None of the previous One should highlight that the UNED is an important distance learning insti tution in the world Experiments should be carried out once and again improved and encouraged These should be seriously taken into consideration so as to prepare programs adapted to the European Space of Higher Education 8 2 The AutomatL bs project The second pilot experience consisted of expanding the first s
73. 0 uned es 2371 disconnected at 10 11 2009 11 44 09 eduroam dhcp 10 211 13 230 uned es 3405 connected at 10 11 2009 12 16 55 Openned reference of hooke v6 vi at 10 11 2009 12 16 56 eduroam dhcp 10 211 13 230 uned es 3405 disconnected at 10 11 2009 12 16 56 eduroam dhcp 10 211 13 230 uned es 3854 connected at 10 11 2009 12 53 59 Openned reference of snell v3 vi at 10 11 2009 12 53 59 Application absolute path 4 C Documents and Settings Usuario Escritorio JILiapps snellisnell_v3 vi Figure 3 24 The JiL Server front panel When a VI is published through the JiL server the server performs an auto matic scan of all the VI controls and indicators initializes the network input port and waits for an incoming connection from a Java applet When the connection is established the server then listens for incoming commands which it parses and serves as requested More details about this process can be found in JiL 2009 There are three types of commands request control and state commands Request commands require information about the controls and indicators avail able which the server returns in the form of a list Control commands are used to open run stop or close the VI These commands perform the same function as if a user interacts with the VI locally Finally state commands are used to read the value of an indicator or set the value of a control in the VI 3 3 A new approach to connect Java and Lab VIEW 65 3 3 3 Details of t
74. 1 m v current folder Please check this fragment instead of the previous one Best H ctor Terminado Figure C 6 Sending fragments by the dialog window Send The list of fragments is displayed in a table located at the lower part of the eJournal interface see Figure C 7 The displayed files depend on the selected filter and the current work folder It is also possible to add comple mentary information to the fragments to make them more comprehensibles For instance in this space users can find information about date time of creating of fragments author fragment size etc Even it is possible to dis play them in a cronological order By dragging the mouse over a fragment a short description about its source is shown This information includes the creation time work module task size etc 244 C eMersion Management http lab dia uned es 8080 eJournal Folder View Mozilla Firefox DER Fragments O Copy Move Delete Rename Import Export Share Send Assign Finalize Submit Hote Name Author Task State Creation Annotation Carlos m Dormido SS 30 June Text file Anton m Dormido zj 25 June Text file HeatFlowGraph gif Dormido x 25 June GIF image aldoThreeTank m Estudiante S 24 June Text file ThreeTankGraph gif Dormido 18 May GIF image registroOneTanksss m Estudiante 8 May Text file registroOneTankeee m Estudiante 8 May Text file Pepe m Dormido 27 Ap
75. 2 sim Y q2 real q1 real 0 L J L L L 1 L L L 800 1000 1200 1400 1600 1800 2000 2200 2400 2600 2800 time seconds Figure A 5 Tracking references Three tank system 206 A Modelling and Control The experiment starts with the tanks in a zone close to the normal operating point h1 300 mm h2 200 mm h3 250 mm Then the reference for the tank 1 is risen until 350 mm at t 1000 sec and after that the reference for the tank 2 is placed in 250 mm at t 1800 sec The plots in Figure A 5 show the behaviour of the system under changes in the references in both simulation and real We can see that the liquid level for the tank 1 and tank 2 track perfectly to the setpoint Furthermore the control signals do not reach the pumps saturation for these jumps in the references Nevertheless the controllers implement antiwindup techniques in order to counteract this effect when saturation levels are reached 320 T T T T hisim EA h2sim h3sim 280 F htreal E 260 h3real h2real o ee o 240 4 gt 220r 4 200 Pal AAA pear FE an AL E aama 1 80 1 L 1 1 L L I 1 500 600 700 800 900 1000 1100 1200 1300 1400 1500 time seconds 60 T T T T T qi sim a tn q2 sim a EE o q1 real 2 pH E 40F E 3 a o go J 20 4 L L L L L L L
76. 34 Costas L Fari a J amp Rodr guez J 2009 A Configurable Framework for the Education of Digital Electronic Control Systems in Proceedings of the 3rd IEEE International Conference on e Learning in Industrial Electronic ICELIE Porto Portugal DAQ PCI6221 2009 NI PCI 6221 Multifunction DAQ Website http sine ni com nips cds view p lang en nid 14132 Diaz L Ramos G Vargas H amp Costa R 2009 A Virtual Remote Lab oratory to illustrate the Internal Model Principle for periodical signals in Proceedings of the 8th IFAC Symposium on Advances in Control Education ACE Kumamoto Japan Diez J L Vall s M amp Valera A 2009 A ball and beam system virtual and remote laboratory based in Ejs and C in Proceedings of the 8th IFAC Symposium on Advances in Control Education ACE Kumamoto Japan 184 9 Bibliography Dormido Canto R Vargas H Duro N S nchez J Dormido Canto S Farias G Esquembre F amp Dormido S 2008 Development of a Web Based Con trol Laboratory for Automation Technicians The Three Tank System IEEE Transactions on Education 51 35 44 Dormido S 2004 Control learning present and future Annual Reviews in Control 28 115 136 Dormido S Canto S D Canto R D amp S nchez J 2005a The Role of Interactivity in Control Learning The International Journal of Engineering Education Special Is
77. 36 6 Access Control to Experimentation Resources 6 2 3 LabVIEW implementation for authentication In order to offer a wider vision of the LabVIEW implementation for authentication purposes Figure 6 5 shows the information flow between client and server along the authentication process NETWORK Local management of the bookings 0 AS M A 3 enquiries i r 3 LabVIEW SC Identity checking module EN j a ty g F CAN Physical resource Teacher administrator Middleware interface Lab Server Figure 6 5 Global vision of the first stage of the booking process The sequence of actions numbered in Figure 6 5 is briefly described hereafter 1 An administrator from the lab server creates a new entry in the database for a student 2 Once the new entry has been created the student can attempt to access the remote plant with his her user credentials previously sent by the teacher administrator These credentials are intercepted by a new LabVIEW mod ule named Identity checking 3 The Identity checking module verifies user credentials with the data base and reports the result of the access attempt to the client The client can receive three possible answers A message indicating that credentials do not exist or are wrong A message indicating that credentials are valid but the user is attempt ing to connect out of the established timeslot A message indicating that cred
78. 4 18 b depicts how the augmented reality concept was in cluded Here the virtual representation of each tank is kept visible overlapping the video images of the remote process The arrows located on the side of the tanks 1 and 2 are also visible Users can change the setpoint value for tanks 1 and 2 interactively by dragging and dropping these elements The upper menu of the graphical user interface includes an additional element named eJournal El three tanks system eJournal Control Language H1 H2 H3 Setpoint 200 300 seconds Pump1 Pump2 Status video Reading video from camera z Augmented gt gt CONTROL PID Choice a zoom view 300 400 3 H2 mm 1994 Top Tank1 i Buttom Tank1 Q1 cm3 sec 80 Q2 cm3 sec 43 47 Top Tank2 X Buttom Tank2 SP1 mm 300 SP2 mm 200 Time sec 4309 H3 mm Status Lab Remote mode with labserver 8 connected TIME REMAINING min 432520 66 a Example of the video options E three tanks system eJournal Control H1 H2 H3 Setpoint T T 100 200 300 400 500 seconds Pump1 Pump2 Status video Reading video from camera e Augmented gt CONTROL o VIDEO jo 100 200 300 400 500 PID control parameters seconds kpt 3 38 Kp2 855 oe E Tit 5555 Ti2 50 9 Q1 cm3 sec 11 41 Q2 cm3 sec 39 4 Ta1 0 Ta2 0 SP1 mm 300 SP2 mm Time sec 5648 H3 mm 67 Status Lab Remote mode with labs
79. 4 The trash in the eJournal ee 242 C 5 Export fragment registroRealQ1 m to the local hard disk 243 C 6 Sending fragments by the dialog window Send 243 C 7 Fields description of the fragments table 244 C 8 Adding annotations to a fragment o o e 245 Part I PRELIMINARIES Introduction Objectives and Structure Overview This chapter introduces the main objectives of this thesis work Firstly it pro vides a review of the use of Internet based technologies in new teaching learning paradigms The main challenges faced when trying to apply these new paradigms in engineering subjects are addressed In particular the paradigm of flexible ed ucation in a remote experimentation context is first introduced and then the applicability of these ideas in the development of a complete remote experimen tation environment through Internet is presented The objectives and structure of the thesis will then be described These sections provide insights on the main goals of the dissertation and a brief description of each chapter The final section indicates the list of sources and research projects used to support the work presented 1 1 1 Introduction Objectives and Structure Introduction Nowadays Internet is the first tool that comes to mind whenever a student teacher or person in general wishes to find out more about a specific subject The use of computers in the t
80. A complete description of the interface can be obtained from the icon Navegation bar that provides the documentation C 1 3 3 Working with eJournal The eJournal is the container of the work carried out for students Collaboration and interaction features are implemented in this space In this shared workspace students are going to store or retrieve their documents which can be files contain ing information of any kind as for instance data registers image of the evolution signals scope the state of the system notes about experimental results work re ports etc 240 C eMersion Management The communication between the experimentation console and the eJournal is performed through the sub menu called eJournal that is located on the left top part of the console From here it is possible to save either a graphical image of the controlled and manipulated variables of the system or a data register in a text file with extension m to be loaded in the Matlab workspace When users use these options these data files called data fragments are automatically stored in the eJournal Then these data fragments can be used by students to generate their practical reports to be evaluated by the teaching staff dia uned es 8080 Mozilla HA E dia uned es 8080 eJournal Folder View Mozilla Firefox eJournal current Folder Inbox eJownal Name Author State Dett z g SIE madrid m 8 9 Acti
81. Exception e 1 System out println Method sender Error sending data PEP P PRP a Nou wne The data are also rendered to the client by an Ejs view The interface contains the same graphical elements of the virtual lab but now the dynamic behaviour of the elements is updated using the measurements obtained from the server when working in remote mode The previous lines provide guidelines and a background to convert a virtual lab created by Ejs into a remote virtual lab However the development of ap plications using these approaches implies implementing a specific software on the server side LabVIEW and on the client side Ejs and therefore new applica tions for remote experimentation will need to implement all the tasks described on both sides Nevertheless the thesis has addressed one way to take advantage of the legacy code concept and present a new framework for developing virtual and remote labs Section 3 3 will present the new approach that covers this objective 3 2 Adapting virtual labs for remote experimentation 55 3 2 5 Visual feedback from the remote plant A remote visualization system is an essential element in any tele operated en vironment since it allows the users to feel and be aware of the consequences of their actions during a remote working session As a result of this users are more motivated and confident in the use of the system In order to include this module in the construction of the Ejs view tw
82. FLOW LOCAL CONTROL VI til sec tdi sec BlowerManual ge cS OTT 50 52 54 56 58 60 62 64 66 68 70 ti2 sec td2 sec lo kp2 9 0 1 gz setpoint 12 ar VA e 25 30 35 40 45 50 55 60 65 70 75 80 dt sec joi ti3 sec td3 sec y kp3 P 4 P E setpoint T3 34 ALII 4042 44 46 48 50 52 54 56 58 60 62 64 66 68 70 THDGDUDGEDUEDURUEUNUL 35 27 Time 1 25 30 35 40 45 50 55 60 65 70 75 80 Figure 4 22 Local control of the heatflow real system Front panel Making the VI available on the Internet means that only the JiL server is required to run and move the heatflowReal vi file to the apps folder One should mention that no specific communication facilities have to be included in the VI 4 3 3 The virtual and remote laboratory Figures 4 24 and 4 25 show the views of the simulation created with Ejs The first one see Figure 4 24 a depicts the interface of the laboratory when working in simulation mode that is using the mathematical model programmed in js with its ODEs editor The 3D graphical representation of the process was built by faithfully reproducing the shape of the apparatus rur amp Ec fo Figure 4 23 Local control of the heatflow real system Block diagram 4 3 Prototype II The Heatflow system 101 The tridimentional representation of the apparatus provides more realism when working in simulation mode thus al
83. Firefox M List of the current folder s fragment in Trash Restore Delete definitively Name Author Creation Annotation O Report DC motor remoteMode pdf Vargas 6 February O Report DC motor simulationMode doc Vargas 6 February Terminado Figure C 4 The trash in the eJournal Rename This option allows to rename a data fragment Import This function allows to upload to eJournal a data fragment from the local hard disk It is possible to attach notes to the uploaded file Export By using this option students can download export from eJournal point of view fragments from eJournal to their local hard disk C 1 How to work with eMersion 243 http lab dia uned es 8080 Export f DER Export fragment You can click directly on registroRealQ1 m to open or right click and choose Save Link As to download it to your local disk Terminado Figure C 5 Export fragment registroRealQ1 m to the local hard disk Send By this option students can send messages with questions to the teacher assistants Fragments from the workspace can be attached to the messages Figure C 6 shows the dialog window to send a message http lab dia uned es 8080 Send email Mozilla Firefox From To assitant and or others Subject Attach DoR hvargas bec uned es fmorilla dia uned es Y sdormido dia uned es User Estudiante Group Estudiantes registroRealQ
84. IALIZE HARDWARE CLOSED LOOP CONTROL RESET HARDWARE Analog 1D DBL NChan 15amp NumOfWarmuplters ConvLateErrorsToWarnings Analog DBL E E Analog DBL WaitForNextSampClkWaitMode 1Chan 1Samp j E Keeping up so 3 1Chan 1Samp ldo nothing Figure 4 29 Local control of the DC motor real system Block diagram 4 4 Prototype III The DC Motor 107 time yA AAA gt Oms 20ms 40ms 60ms 80ms 100ms data from control task Data 1 Data 2 Data 3 Data4 Data 5 Data6 states vector v lost data ES a shared memory space JiL server flushing data to client Figure 4 30 Operations of the JiL server when sending data to a client The A M amp Speed Position area of the Front Panel contains two boolean buttons positionControl and m a When pushing the first button users can choose between working on either position or on the speed control of the disk The second button allows to switch between manual or automatic control mode The box PID Parameters of the GUI allows to configure the value of the controller parameters Kp T and T and the actuator saturation limits wmax and umin The Setpoint amp umanual options allow the users to interactively change the setpoint value and at the same time modify the voltage applied to the motor in manual mode In the Ke
85. Ja Nms rad Intertia of the steel disk charge Jt Im Nms rad Inertia of the motor brake ft fm Nms rad Friction coefficients of the motor brake The numerical values of the parameters can be obtained in Appendix B The voltage supplied to the motor u t must be compensated by voltage in the re sistence of the inductor Ri t and the induced voltage generated by the rotational motion u t u t Ri t uj t A 16 208 A Modelling and Control The induced voltage is proportional to the angular speed w t of the motor u t Ke c t A 17 where Ke is a constant associated to the motor and the constant flow generated by the inductor The torque M t produced by the engine is proportional to the current passing through it Mm t Km i t A 18 where Km denotes a constant On the other hand a simple energy balance demontrates that constants Ke and Km are identical that is Ko Kn K A 19 According to the Newton motion law the electromagnetic torque supplied to the motor is given on one hand accelerating the engine the inertia and the driven magnetic brake and on the other hand by compensating for the viscous friction of the engine and the magnetic brake the dry friction is ignored Jolt Mm t f w t A 20 where J is the rotational mass inertia J Jm J Jr and f is the total friction f fmt ff For the model based on the experimental setup located in the laborato
86. Server of the current state of the Lab Server and other parameters that the central server needs For example a new reservation will be effectivelly made when the Lab Server is turned on and working perfectly 142 6 Access Control to Experimentation Resources On the other hand information about the time zone of the Lab Server in relation to the GMT is also reported by this module to the central side This information is paramount when a Lab Server is located outside the Main Server time zone Thus when a new booking for a Lab Server is made the central side server accounts for the time zone of the Lab Server reported by the Java Interface to calculate the temporal lag between both and insert the correct timeslot in the database Summarizing the full process is divided into two stages in order to use a physical resource of the laboratory and is described as follows Booking Stage Figure 6 10 shows a state diagram of the booking process Applet for bookings Java Interface g Figure 6 10 State diagram of the booking process A description of the state flow during a reservation is as follows a The Applet for bookings requests a new reservation step 1 in Figure 6 9 b The Bookings Main Server takes the request and saves it in a local DB step 2 in Figure 6 9 c The Bookings Main Server requests the Java Interface for its time zone step 3 in Figure 6 9 d The Java Interface provides its time zone
87. The plant s schematic represen tation is meant to be as self explanatory as possible Students can easily visualize the system s components and see how the liquid levels in the tanks change accord ing to different valve and pump parameters The plant s behaviour is simulated using the software model defined in Figure 4 4 a and Figure 4 5 based on the equations listed in Appendix A see Section A 1 1 Graphical view Plotting view FB three tanks system tmm x Control Language H1 H2 H3 Setpoint seconds Pump1 Pump2 Pumpi 19 01 Pump 45 78 e e Q1 cm3 sec 19 01 Q2 cm3 sec 45 78 E Pause pie Reset SP1 mm 300 SP2 mm 200 Time sec 4 H3 mm Estado Lab Modo Simulacion Control panel Indicators panel Figure 4 7 GUI of the three tank system virtual lab Students can play pause and reset the simulation using three push buttons from the Control panel They can also modify the percentage of opening of the connecting A13 A32 A2 and draining valves Aleak1 Aleak2 Aleak3 If the system is in manual mode the student can control the flow sup plied by the pumps with two sliders located in the lower part of the applet s main window Pumpl and Pump2 Moreover two small arrows placed on the sides of tanks Ti and T3 allow to change the setpoints when the system is in automatic mode Students can modify these setpoint values interactively by dragging the 4 2 Prototype I The Thre
88. Universidad Nacional de Educaci n a Distancia Departamento de Inform tica y Autom tica Doctoral Dissertation An Integral Web based Environment for Control Engineering Education H ctor Vargas Oyarz n Electronic Engineer from Universidad de La Frontera Chile Submitted at E T S de Ingenier a Inform tica Universidad Nacional de Educaci n a Distancia MADRID 2010 Universidad Nacional de Educaci n a Distancia Departamento de Inform tica y Autom tica Doctoral Dissertation An Integral Web based Environment for Control Engineering Education H ctor Vargas Oyarz n Electronic Engineer from Universidad de La Frontera Chile Submitted at E T S de Ingenier a Inform tica Universidad Nacional de Educaci n a Distancia MADRID 2010 Department Faculty Dissertation Title Author Academic Degree Advisors Informatica y Automatica E T S de Ingenieria Inform tica An Integral Web based environment for Control Engineering Education H ctor Vargas Oyarz n Electronic Engineer Universidad de La Frontera Chile Dr Sebasti n Dormido Bencomo Dr Jos S nchez Moreno In memory of my dear grandmother Doraliza Alvarez Acknowledgements First of all I want to express my deep gratitude to the Spanish Ministry of Ed ucation and Science for granting me the scholarship that allowed me to develop this thesis Secondly I would like to thank all those who have contributed to the dev
89. Using Interactive Tools Computer Applications in Engineering Ed ucation 13 1 84 91 S nchez J Dormido S Pastor R amp Morilla F 2004 A Java Matlab Based Environment for Remote Control System Laboratories Illustrated With an Inverted Pendulum IEEE Trans on Education 47 3 321 329 192 9 Bibliography S nchez J Morilla F Dormido S Aranda J amp Ruip rez P 2002 Virtual and Remote Control Labs Using Java A Qualitative Approach IEEE Control Systems Magazine 22 2 8 20 SysQuake 2009 Calerga Website http www calerga com Thompson C W 2005 Smart Devices and Soft Controllers IEEE Computer Society 9 1 82 85 Travis J 2000 Internet Applications in Lab VIEW Prentice Hall UDIMA 2009 Distance University of Madrid Website http www udima es UNAD 2009 Open University of Colombia Website http www unad edu co UNED 2009 National University for Distance Education Website http www uned es UOC 2009 Open University of Catalonia Website http iip uoc edu Urqu a A Martin C amp Dormido S 2005 Design of SPICELib A Mod elica library for modeling and analysis of electric circuits Mathematical and Computer Modelling of Dynamical Systems 11 1 43 60 Valera A Diez J L Vall s M amp Albertos P 2005 Virtual and Remote Control Laboratory Development IEEE Control Systems Magazine 25 1 35
90. Value E Ticks Format Background Minimum 0 Closest Foreground 64 128 192 Maximum 100 Interactivity Font Dialog BOLD 11 Enabled manualControl Format pumpllabel Tooltip Orientation On Preas On Drag On Release external synchronize Figure 4 15 The _external synchronize method is called when user modifies the value in pump1 manually 4 2 Prototype I The Three tank system 93 Figure 4 15 shows the properties window of the slider in the control panel associated with pump 1 The _external synchronize method is called in the On Release property of the slider so that when the mouse is released from this element the method is triggered Step 3 Finally visual feedback is included by using the WebCamImage ele ment of Ejs Figure 4 16 shows the tree of elements used to compose the view of the application The WebCamImage is placed in first position in the hierarchy for a view with augmented reality Tree of Elements k Elements for the view gt Swing E Simulation View MainFrame Containers Interface Menu gt Y MenuBar oO i LJ Console EB spiitPanel sl de 9 LJ PanelPrincipal F1 J DrawingPanel Camlmage go Tank1 a Tank2 a ranks m TankSumi Pump12 Control 3D N setpointt D Setpoint2 e Valve13 T setpoint1
91. a The modelling and control of this process is addressed in Appendix A 3 and no additional information is given here This chapter will only focus on how to create web based control labs using the software tools and approaches suggested in this dissertation X 7 Vh Jur Ambient temperature Figure 4 21 Variables of the apparatus The setup of the server side applications is described hereafter Again this section focuses mainly on the development of the local control VI since issues of communication have already been tackled in the three tank prototype JiL server and in the previous chapter 4 3 Prototype II The Heatflow system 99 4 3 2 Local control of the heatflow system The LabVIEW VI developed is shown in Figures 4 22 and 4 23 A sequence structure was used to program the local control with hardware in the loop see Figure 4 23 This VI controls the hardware of the thermal process based on the structure described in Figure 3 28 Section 3 3 5 The application was developed so that users can switch between manual or automatic mode When working in automatic mode the user can choose between one of the three sensors to close the loop It is also possible to tune the parameters of the PID controller on the fly or change the temperature setpoint in the sensor used to close the loop As was the case for the three tank prototype a resetting code is executed when the variable labelled as stop is pushed see Figure 4 22 HEAT
92. a register it also called fragments for later analysis Ga The third icon opens a window that contains links to external Web tools Currently the only link allows the access to the automatic bookings system to make time slot reservations for a later use of the remote systems ri Through this icon all the exercises and theory documentation are accesible By clicking on it an auxiliar window containing a link to the documentation is popped up see Figure C 2 http lab dia uned es 8080 Documentation DER Documentation Protocolo de Tareas Task Protocol http lab dia uned es 8080 Depto Informatica y Automatica UNED Mozilla Firefox Protocolo de tareas Motor de corriente continua Interfaz DI adminiab INTRODUCCION En este documento se enumeran las distintas tareas que constituyen la practica que se va a realizar con el motor de corriente continua 1 TAREAS EN MODO SIMULACI N Y REMOTO Tarea 1 Estudio de las caracter sticas del motor e identificaci n del modelo Realizar las experiencias de control manual propuestas en el apartado 6 2 del gui n de pr cticas secci n de tareas en modo simulaci n y remoto experimentos Tarea 2 Control de velocidad Figure C 2 Accessing to the documentation of the practice This documentation is structured as follows A first link called Protocol con tains the set of activities that a student will must carry out to complete a module it is t
93. aboration mechanisms could help reinforce this idea as these interaction resources would encourage collabo ration between peers and tutors and allow them to become active agents Discretionary of Collaboration Collaboration and interaction should not be mandatory The system should allow students to switch between a single and a collaborative working mode in a transparent way From a handling point of view the previous quotation describes this environ ment as a space in which the students can find all the means necessary to carry out on line laboratory assignments in a flexible learning context The graphical user interface of eMersion allows to surf across the experimentation environment such 5 2 eMersion A novel approach from EPFL 119 as a pilot would do in an aircraft cockpit Hands on activities are conducted on line using either simulators virtual labs or remote connections to real laboratory equipment remote labs Typical experimental sessions are mediated by teach ers and or tutors emulating hands on activities carried out in the face to face learning modality Furthermore student student interaction and collaboration is also encouraged by means of a shared workspace where they actively create their own contextual meaning rather than passively acquiring knowledge structures created by other actors in play Nguyen et al 2006 5 2 2 Web technology behind eMersion The selection of an appropriate web technology to dev
94. ace during classroom sessions on the use of the remote experimentation interface were completed by teaching assistants Then students were required to use the system to get a fluent han dling of the interfaces Finally and after several sessions they could complement their work by using the web based experimentation system of AutomatL bs UNED students were first offered the possibility of accessing the systems avail able at the UNED the DC motor the heatflow system and the threetank system and could then complete their work remotely through the Internet During these experimental sessions students could save data measurements and parameters which they used later to write their final reports These reports were placed by students in the eJournal space for a post evaluation Teaching assistants from each university taking part were in charge of completing this task 8 2 1 Access to AutomatL bs Figure 8 3 shows the website whereby students access to the experimentation system It provides information on the whole system the processes available the network s virtual and remote laboratories etc The purpose of this information is to introduce students to the virtual experimentation before working with the real processes As mentioned before teachers and students from other universities took part in this second evaluation of the environment The main aims of this study were a To enable students to access other practical experimen
95. acts as a gateway and input filter to the system An access management module This software module contains all the in formation related to users timeslot bookings and other parametrizable information A database manager is used to handle the users reservations and physical resources Each register in this database corresponds to a booking scheduled for a specific time and date An instrumentation module This module incorporates all the hardware needed to connect the physical system with the server amplifiers power modules data acquisition cards DAQ signal conditioning etc A remote visualization module This module allows the users to examine what is happening with the physical system during its remote manipulation from a client In order to provide this feature video cameras are used The system must be able to transmit a sense of realism in order to encourage its usage and increase motivation whilst completing tasks 2 2 A systematic two layer approach 23 An audio feedback module This module is optional It would allow to further increase the sense of realism For example in a level control remote lab we could transmit the ambient sound that is produced by manipulating pumps or just by listening to the water filling the tank Each module in the server has its counterpart in the client side For example to read the video streaming captured by the server the client application must implement a new m
96. ained are c 2EnT Tpeak V 1 E TE pe AE 4 30 GTpeak Let 127 Ki A 31 ll In order to set the gains K and K a damping ratio 0 4 and a time peak 12 can be used to compute them Thus the PI control parameters are Kp 0 1437 V deg and K 0 0679 V deg sec Reference tracking A simple experiment to show the performance of the PI controller designed for the control around sensor 1 has been carried out see Figure A 16 Closed loop control around Sensor 1 T T T T T T T1sim 5 BEBE 00 00 as Tiref T1real E 110 120 130 140 150 160 seconds 45 L T T T T J Vhsim 4r Vhreal g 9 3 57 7 3 4 2 5 I M 1 L 1 1 110 120 130 140 150 160 seconds Figure A 16 Tracking reference closing the loop around sensor 1 constant voltage applied to the blower of 3 volts A 3 The Heatflow system Modelling and control 217 The experiment starts with the system stabilized around the operating point T 43 C and U 3 volts see Figure A 16 At 110 sec the temperature reference Tires is changed to 53 C We can observe that the behaviour of the system both in simulation and remote mode are similar for the set of control parameters obtained during the designed process The system reaches the steady state in almost 20 sec without saturation of the control signal which is accord in
97. ained results Then another user or even a teacher assistant can answer the ques tion adding new annotations to the fragment Figure C 8 shows the dialog window to attach comments to a fragment http lab dia uned es 8080 Annotations Mozilla Firefox Annotations for ThreeTankGraph gif Author Dormido 18 May GIF image It is possible to add an annotation comments to this fragment Post new annotation Close Terminado Figure C 8 Adding annotations to a fragment
98. alization or interaction purposes just as in any other Ejs generated view obtaining direct access to the VI controls and indicators Authors need to take into account however that variables connected to VI controls should be considered as write only and variables connected to VI indicators as read only variables 3 4 Conclusions The methodology proposed in this chapter provides a structured framework for new developments of on line experimentation systems The move from former applications mentioned whereby all components in the server and client were mixed within a single program towards a module based architecture has allowed to reduce the amount of effort to create new virtual and remote laboratories The command based architecture is a particularly appropriate solution to orga nize low level programming details to communicate embedded applications to a remote host via networking protocols such as TCP or UDP Furthermore this architecture presents a well defined and expandable structure allowing the addi tion of new commands to enhance the system s intelligence On the other hand Ejs and LabVIEW evidently have all the features needed to enable the development of on line experiments following a client server ap proach The following chapter presents some examples of web based remote and virtual control laboratories created applying the approaches described here Prototypes Developed Overview The various prototypes presente
99. alzmann et al 2000 The following section provides some key issues on communication protocols used in the experimental applications developed along this work and how they have been applied seeking the best Internet practices 3 2 1 1 Communication protocols As described above data packets travelling across a communication network al low a source host to reach a destination host The data transportation is carried out using network protocols Network protocols define rules and conventions to enable communication among network devices That includes the mechanisms used to identify devices and connect them to each other as well as formatting 2 Quality of service QoS in short refers to the ability of a network to provide better service to a traffic network selected over various technologies and techniques Elements of network per formance within the scope of QoS often include availability uptime bandwidth throughput latency delay and error rate 42 3 The Experimentation Layer Client a Internet k i Internet Figure 3 10 Stack connections of the Internet Protocol Suite rules that specify how data are packaged into messages In the particular case of remote experimentation Internet Protocol Suite is suited to encapsulate and deliver data between client and server using either the TCP Transmission Con trol Protocol or the UDP User Datagram Protocol protocol as transportatio
100. and present data without program 234 B Hardware Description ming In addition to LabVIEW SignalExpress M Series data acquisition devices are compatible with the following versions or later of NI application software LabVIEW 7 x LabWindows CVI 7 x or Measurement Studio 7 x or LabVIEW with the LabVIEW Real Time Module 7 1 M Series data acquisition devices are also compatible with Visual Studio NET C C and Visual Basic 6 More resources and information about both DAQ board and driver software can be found in DAQ PCI6221 2009 eMersion Management This appendix provides information about the management of the eMersion en vironment from a student handling point of view This documentation presents a summary of the available options for this platform i e the options that allow a new user to start working with the remote experimentation environment and to discover other new features by himself Note This information is accesible to the students performing the hands on laboratories in the UNED from the eMersion environment 236 C eMersion Management C How fo work with eMersion C 1 1 Introduction This document aims to serve to students as a guide for the management and handling of eMersion Through the steps described hereafter the student will be able to carry out and complete satisfactorily a remote experimentation session http lab dia uned es 8080 Supervision Welcome Estudiante Estudiante
101. ards and Projects 11 pr cticas a trav s de Internet Revista Iberoamericana de Autom tica e Inform tica Industrial RIAI ISSN 1697 7912 Vol 7 Nr 1 pp 35 45 Enero 2010 2 H Vargas J S nchez Ch Salzmann F Esquembre D Gillet S Dormido 2009 Web enabled Remote Scientific Environments Computing in Sci ence and Engineering vol 11 no 3 pp 34 46 3 H Vargas J S nchez N Duro R Dormido S Dormido Canto G Farias S Dormido F Esquembre Ch Salzmann D Gillet 2008 A systematic two layer approach to develop Web based experimentation environments for control engineering education Intelligent Automation and Soft Comput ing vol 14 no 4 pp 505 524 4 R Dormido H Vargas N Duro J S nchez S Dormido Canto G Far as F Esquembre S Dormido 2008 Development of a Collaborative Web Based Control Laboratory for Automation Technicians The Three Tank System IEEE Trans on Education vol 51 pp 35 44 5 N Duro R Dormido H Vargas S Dormido Canto J S nchez G Farias S Dormido 2008 An Integrated virtual and remote control lab the three tank system as a case study Computing in Science and Engineering vol 10 pp 50 59 6 J L Guzm n H Vargas J S nchez M Berenguel S Dormido F Ro dr guez 2007 Education Research in Engineering Studies Interactivity Virtual and Remote Labs ISBN 1 60021 829 6 Book Chapter Di
102. ared so as to contribute actively In other words all par ticipants should feel that they are members of a group that they are not working in isolation and without support especially in the case of courses involving high tech contents The social learning methodology provides slightly more than the blended learn ing method In particular in addition to the necessity of being part of a group it provides the opportunity to meet with a teacher In fact the presence of the role of teacher or tutor enhances students motivation the same as in a traditional class with a leader and an formal objective In Anderson amp Elloumi 2004 one can find a general model of how teacher B el student student student content 1 y 1 1 Search amp retrieval Student KNOWLEDGE Content Tutorials A Communication Asynchronous lt gt CONTENT gt Simulations Pee aoa synchronous content INTERFACE content amp games P PP Virtual labs e Books A 1 student teacher teacher content H s OTHER TEACHERS Figure 5 1 Interaction model illustrating the two main modes of on line learning 5 2 eMersion A novel approach from EPFL 117 and student interact with one another and with the contents of a course Figure 5 1 shows such interaction model illustrated with communication links between the different model components As a summary and in order to introduce UNED control engineering stu dents an
103. ase was explored As a result a full automatic booking and authentication system was implemented 6 2 A simple point to point authentication protocol 133 6 2 Asimple point to point authentication protocol 6 2 1 User authentication Figure 6 1 depicts the general idea of the p2p authentication scheme implemented The protocol considers the sending of authorization credentials from the user to the lab server The server provides a simple authentication service in which credentials are checked in the local database that contains a list of authorized users Then the server returns the result of the authentication attempt to the user and grants or denies access to the plant pp gt Lab Server User Figure 6 1 Point to point authentication protocol between a user and a remote server providing authentication services Figure 6 2 illustrates the authentication process From a client interface point of view the process suggested consists of a programming module to send a user name and an encrypted password from the client to the server and obtain an answer from the server indicating the result of the access attempt The server is therefore responsible for checking whether the username and password exist in the bookings database and must also verify whether the user is trying to connect during the timeslot booked Although the implementation of the authentication programming modules in the client and server side can be achieve
104. asks run in parallel to the control task but in different threads and with lower execution priorities Finally methods that share data among tasks have to be performed but must take the determinism of the control task into account 3 2 Adapting virtual labs for remote experimentation 47 LabVIEW is especially suitable here since it provides simple multi thread programming using Timed Loop blocks These blocks allow to the programmers to include multiple threads in a single virtual instrument VI and running these threads at different sampling periods and with different priorities Figure 3 15 shows a LabVIEW block diagram that corresponds to the ar chitecture of the server side in Figure 3 13 The three loops in the diagram run concurrently to perform the main tasks control communication and video grab bing Figure 3 14 indicates instructions executed for each of the task Task 1 Control critical time priority Recover control parameters from the communication task Data acquisition and closed loop control Transmit the system state to the communication task Task 2 Video normal priority Get images from the camera and give them to the communication task Task 3 Communication normal priority Receive control data from clients and write them to the control task Read the system state from control task and the images from video task Concatenate state images and send them to the client Figure 3 14 Conc
105. assword are valid Returns result together with starttime and endtime Check Time Slot Check T a EAS 7 EN If username and password exist it verifies whether the current time of the server is between the starttime and the endtime of the booking Result Send Result to Client ree Gets the authentication result message from the previous block and sends it to the client application through the TCP channel Microsoft s ActiveX Data Objects ADO for accessing data sources 138 6 Access Control to Experimentation Resources 6 2 3 1 Integration into the JiL server The LabVIEW module previously described has been integrated into the JiL server as a new command for client to use By adding this new feature the JiL server application explained in Chapter 3 looks as shown in Figures 6 7 and 6 8 gt JiL Server 1 0 File Options Help m Is there any user connected Stop Edit General Options Connection Parameters General Parameters TCP Port 2055 S Open Front Panel Smtp server smip uned es Package Size 1 E C Start JIL at launch Sender s email adminlab dia uned es Sending Rate ms 100 2 Recipients MEF ES A v se AutomatL bs Authentication Data Source Name motor Updating Delay sec 20 Check DB Connection to DB 9 a Edit General Options window gt Jil Server 1 0 Fie Options Help Is there any user connected Start Edit General Options Connection Paramete
106. ated a Java thread is ready to be started Once launched the set of public methods that the Video class provides can be used to control the connection with the camera These methods are public start This is the main method of a Video class When the start method is invoked on a Video object previously created the run method of a Java thread is executed The run method of Video class opens a socket connection to the remote IP camera and it starts the reading of the video stream according to the chosen format public closeStream This method stops reading the video stream and closes the socket The Moving Picture Experts Group commonly referred to as simply MPEG is a working group of ISO IEC charged with the development of video and audio encoding standards In computing JPEG is a commonly used method of compression for photographic images The degree of compression can be adjusted allowing a selectable tradeoff between storage size and image quality JPEG typically achieves 10 to 1 compression ratio with little perceivable loss in the image quality 98 3 The Experimentation Layer objects created at low level The run method of the thread is over when this method is executed public boolean isConnectedVideo Enquires if the link to the camera is up or down Commonly this boolean indicator is used when the getImage is invoked to get the images public Image getImage The images captured from the IP
107. atisfaction enquiry about the practical activities Figure 8 11 a and Figure 8 11 b give comparative information about how learning improved when using these new technological methods compared to the traditional ones In most cases students disagreement around 9 is due to the fact that they did not have the chance to experiment directly with the real equipment 8 2 The AutomatL bs project 171 Quality of the remote laboratory 10 c m Y 38 Very Good Good Acceptable Bad Very Bad a Quality of the virtual laboratory Very Good Good Acceptable Bad Very Bad Figure 8 11 Regarding the overall system A solution to this problem is to follow an educational methodology based on the blended learning concept That is to first carry out face to face classes where students can interact and experiment in situ with the real plant and to enable them afterwards to have access to the experimentation environment to complete their practical exercises at a distance With regards to the quality of virtual and remote laboratories most students expressed a positive opinion about its development in terms of user functionality see Figures 8 12 a and 8 12 b Negative opinions were a consequence of the bad quality of the Internet con nections established Ejs presents great advantages compared to other software alternatives to develop the client side interfaces for remote experimentation The high
108. atories with educational purposes From eMersion it is possible to get a portfolio with all the materials that a student needs to complete an on line session of practical assignments The portfolio is composed by user s manual theory assignments task protocols booking system and the eJournal eJournal is a shared space to facilitate the communication and collaboration among students and instructors during the learning process AutomatL bs puts at the disposal of the students this valuable and innovative tool in order to complete the experimental sessions in control engineering across internet Heatflow system Control of the air temperature inside a rectangular duct i di amp x DC motor Position and velocity control of a Direct Current s Motor eELab Position and velocity control of a Direct Current Motor Simple tank Liquid level control of a simple tank under EMERSION EXPERIMENTATION ENVIRONMENT MOTOR C C disturbances Figure 8 3 AutomatL Obs home page 8 2 2 Remote systems available In addition to the prototypes presented in this thesis each university participant provided a physical plant A brief description of the remote systems available in the AutomatL bs project are presented hereafter DC motor Miguel Hern ndez University of Elche The Direct Current Motor is one of the classical experiments in automatic control laboratories It allows to study the dynamic behaviour of a mo
109. ay suado 8ajej ayy suado sjuauwbeyejep spues 158 7 Integration of Layers space Figure 7 5 eJournal option Thus students can use the results obtained images of the system s evolution or data registers in the experimentation ses sions to ellaborate their reports for the final assessment The user interface of the booking system has been completely integrated into eMersion This means the students can make their reservations from the exper imentation environment in order to complete remote experiments 7 5 Conclusions The linkage and interoperability of all the pieces of the remote experimentation system of this dissertation has been shown This mixture of web components working at the same time both synchronously and asynchronously was the big challenge to reach during the development of this thesis The software tools Ejs LabVIEW and eMersion used provide considerable benefits compared to other alternatives Also and thanks to this work the con nection between these three software tools have correctly been implemented and tested Ejs applications can be integrated into eMersion in a simple and straight forward way through specific built in methods The eMersion management sys tem is what allows the linkage between the web components that constitute a remote laboratory Finally the next chapter presents some results obtained from testing the sys tem Two pilot experiences were carried out to test the system and c
110. based experimentation environment used in this dissertation to implement the e Learning Layer has now been ex posed The following sections describe the graphical user interface of eMersion that is the look and feel of the web environment and the functionality of each of their components 5 2 4 Functional architecture of the eMersion GUI The eMersion GUI is composed of a set of independent web applications that make up the entire environment Figure 5 7 shows each component of the eMersion GUI Of all these modules the eJournal is highlighted as it implements a big part of the interaction and collaboration services provided by eMersion A student enters the environment via a web browser Once authenticated a set of pop up windows are automatically opened to display the entire experimentation system In Figure 5 7 each window represents a web module developing a specific task into a client web browser 124 5 The e Learning Layer Web browser Navigation bar objectives and state T r Experimentation Complementary Additional External eJournal Console Information Applications Collaboration and n Struts Interaction space E En LabVIEW VI DAQ and Control C vil Server Figure 5 7 Functional structure of eMersion A brief description of each component of the collaborative web based experi mentation environment eMersion is detailed as follows Navigation bar Figure 5 8 shows the look
111. been addressed the reason being that didactical setups used do not have dynamics that require UDP based solutions Furthermore TCP is widely accepted in most communication networks and less filters in firewalls 44 3 The Experimentation Layer are found for this protocol The following sections describe in detail the structure of the terminals in a remote experimentation scheme i e client and the server applications and how they use network protocols to exchange data 3 2 2 Building the server side Figure 3 9 shows the content of data blocks exchanged during a peer to peer connection between client and server Packets outgoing from the server contain data representing a semantic content different according to their origin The generation of such information and its sending implies completing a series of tasks in the server machine concurrently In the particular case of remote labs the three fundamental tasks to complete are 1 The control task It performs the real time closed loop control of the real plant It produces the states of the plant represented by the measurements in s k blocks see Figure 3 9 a At the same time control parameters in c k blocks see Figure 3 9 b are used by the control task to modify the behaviour of the system 2 The video task It retrieves the video streaming captured by a camera focusing on the real process Video images in s k blocks are generated by this task 3 The communicati
112. cal sciences studies Williams 2007 In addition to the textual multimedia information and the other resources that are required to provide the theoretical aspects in an on line course hands on laboratories should also be included This is particularly true in control engineering education a field inherently interdisciplinary where mathematics play a fundamental role and where progress is achieved through a mix of maths modelling computation and experimentation Fleming 1989 Dormido 2004 str m 2006 In this given context students should be able to achieve the following 1 1 Introduction Understand the underlying scientific model of the phenomenon studied Get acquainted with the limits of this model ie how does the model accurately reflect the real behaviour and to what extent it remains a basic approximation Learn how to manipulate the parameters of the model to fine tune the behaviour of the real system The implementation of an effective web based educational environment for any engineering topic should cover these three aspects of the technical education conceptual interpretative and operational These should provide the student with an opportunity to become an active player in the learning process Dormido et al 2005a Flexible education in engineering The potential of web based experimental applications as pedagogical support tools in the learning teaching of control engineering has been presen
113. cation etc To add new communication protocols to the JiL Server approach in such a way that it allows web clients developed with other technolo gies control the devices remotely This would allow to make a simple mashup of these applications with other web enviroments Ngolo et al 2009 Vargas et al 2009c 2 It would be very interesting to modify the current remote labs prototypes in which the structure of the controller is already predefined towards a more flexible structure where students can define the controller form This idea is pursued in the next point Some examples supporting this working line can be found in Farias et al 2009 To extend the virtual and remote control laboratory prototypes devel oped throughout this dissertation by including new control algorithms and even by offering students the opportunity to test their own imple mentations Thus these hands on laboratories through the Internet could be applied to courses in higher degrees Master or PhD pro grams eMersion capabilities can be enhanced by means of integrating external web applications Two new modules could be developed and included in order to improve students motivation in the completation of their remote working experience These ideas were taken from Gillet et al 2005 Salzmann et al 2008 To include a new web module into eMersion that would allow students to carry out the analysis and design of controllers on line
114. chnical and organizational From a pedagogical point of view flexible education means to provide extended access to learning resources for all the students As a consequence students have more liberty as to how to organize their learning activities and thus the participation autonomy and collaboration between the actors involved in the laboratory teachers tutors and students are enhanced Campus University On line Experimentation Services Figure 1 1 Scheme of the flexible education paradigm Figure 1 1 illustrates the idea of the extended accessibility A student could follow out his her laboratory tasks anytime anywhere for instance from the cam pus from home or from any site connected to the Internet From a technical point of view flexible learning means an adequate exploita tion of information communication devices and infrastructures available espe cially the Internet and the services derived from its usage Finally from an organizational perspective the flexible learning paradigm trusts in renewed cur 1 1 Introduction ricula and the relationships and results generated through research work and collaboration Increased contact between researchers from different academia in stitutions will be achieved thanks to network infrastructures The flexible learning paradigm undoubtedly defines and clarifies a set of key aspects to be considered However an analysis of the pedagogical aspects involved should
115. confident Thanks again Pilar for the nice con versations and moments that you have given me Thanks also to Professor Francisco Esquembre the Easy Java Simulations developer for his support in the use of Ejs his guidance at the beginning of the research and for his valuable help in reviewing my work Thanks also to Professor Matilde Santos Pe as from the Complutense Uni versity of Madrid for the trust she placed in me before coming to Spain and her kindness and goodwill during my first days here Thanks to Professor Denis Gillet leader of the developing team of the eMer sion project for accepting me in his group to enhance my research work and for his support during my stay at the Ecole Polytechnique Federale de Lausanne EPFL Similarly I would like to express my special gratitude to Dr Christophe Salzmann whose help support and work in remote experimentation topics have hugely contributed to the development of this dissertation I would like to thank my family my parents H ctor and Mirta who have taught me everything I should know to face life I also would like to thank my brother and sister Jaime and Patricia for being there always Especially I would like to thank my dear wife Claudia She has encouraged me and pushed to start this adventure far away from our families Claudia thank you so much for the present you have given me Finally I feel deeply thankful to my director of thesis Professor Sebasti n Dormido Bencomo
116. cription of the contents of the web form in Figure 7 4 is given below Protocol name The name of the tasks protocol created previously to the current module setting Protocol objective A short description about the objectives of the mod ule This description will appear on the navigation bar of eMersion Protocol description file An HTML or PDF file containing an eMer sion s guide This is the first document to be consulted by users Toolkit Integration of external web applications A course s administrator 7 3 Linking all Work modules in eMersion 155 can provide as many URLs as external toolkits s he wishes For example to integrate the Automatic Bookings System the URL of the client interface of the bookings application must be included see Figure 7 4 Choosing the module s logo A custom logo can be used for the module The image file could be located in the local hard disk or any URL on the Internet If no logo is provided the eMersion logo is used by default A language module The course Administrator can choose between three different languages English French or Spanish Experimentation window size pixels The size in pixels of the pop up window containing the experimentation console Ejs applet Experimentation URL The URL where the experimentation console is reachable on the Internet Servers hosting the applets are not restricted to a specific location Documenta
117. d analyzed the UNED pilot experience and the AutomatL bs project The set of remote labs integrated into the remote experimentation system of the AutomatL bs project has been also described Student s perception was analyzed during the academic years 2006 2007 UNED pilot experience and 2007 2008 AutomatL bs project In both tests the re sults show a high degree of satisfaction which demonstrates that the approach presented was correctly developed The information extracted from these studies will be used to improve the quality of the web based experimentation system New functionalities or the im provement of existing ones must be achieved according to the results obtained in these evaluation processes 9 1 Conclusions and Future Research Conclusions General conclusion Virtual and remote experimentation for engineering education can be considered as a mature technology However the process of transforming a classical control experiment into an interactive web based laboratory is not an easy task This dissertation provides a systematic approach experimentation and e learning lay ers for the fast development of remote laboratories using three tools Easy Java Simulations Lab VIEW and eMersion The approach suggested eases the devel opment of online experimentation environments and it also provides an effective scheme to switch between simulation and tele operation of real systems a key feature for hands on learning activ
118. d in this chapter summarize the concepts imple mentation details and aproximations exposed in Chapter 3 Following the guidelines indicated for the experimentation layer three virtual and remote control laboratories were created Each one was specially chosen to satisfy certain security and stability requirements when real equipments are re motely handled The model view control paradigm is first used to design the virtual version model based of every laboratory In this particular case Ejs allows the creation of powerful simulations in Java language applying this framework Once the virtual version of the laboratory has been obtained the sequence of steps needed to in clude remote experimentation into the virtual lab is as follows The main goal of this chapter is to show how to apply this methodology men tioned in previous chapter and serve as a reference guide in the development of new on line experimentation laboratories 78 4 Prototypes Developed 4 1 Case studies A short introduction The following sections describe how three examples of virtual and remote labora tories in the scope of automatic control have been developed the three tank sys tem the heatflow system and the DC motor These three environments are part of the current infrastructure of laboratories that the Department of Computer Science and Automatic Control of the Spanish University for Distance Education UNED provides to students for remote experimentatio
119. d using any type technology the system was developed using LabVIEW on the server side and Java on the client side to match the process developed in previous chapters As we can see the authentication process on the server side generates some 134 6 Access Control to Experimentation Resources Applet client Client sends username and encrypted password to the server Client i INTERNET i Server Y Does user exist in DB verification in database User does not appear on database access list Has the timeslot been booked The user exists but needs current time is within time slot to check timeslot booked yes Successful access to the real equipment in the lab Didactical setup in the laboratory Figure 6 2 Algorithm of the point to point authentication protocol various issues T he following two sections describe both the structure of the book ing database and the LabVIEW program that implements the logic of checking The integration of this programming module into the Jil server will then be de scribed and finally the authentication issues on the client side will be addressed using Ejs 6 2 2 Adatabase for the booking of physical resources A database connected to a real plant located in each server manages user access to physical resources These databases contain the booking registers for the re spective plants As previously mentioned in Section 6 1 1 databases
120. degree of acceptance was because of the quality of the applications in par ticular the simulations which represent almost faithfully the real aspect of the 172 8 System Assessment Most important learning resources 0 18 pg Documentation Questions to teacher Simulation Connection to plant Others Satisfaction degree of students pA 7 i cnt 69 Strongly Agree Agree Neutral Disagree Strongly Disagree b Figure 8 12 Quality of the web based laboratories didactical setups located in the university laboratories This is a very important aspect since students get a first insight of the systems through the work they complete in virtual mode and therefore the simulations behaviour should match real counterparts as much as possible Facility of using the system Mio 0 o 62 Strongly Agree Agree Neutral Disagree Strongly Disagree Figure 8 13 The most important learning resource 8 3 Conclusions 173 Finally Figure 8 13 shows how questions and the documentation needed for practical exercises are essential for the student s performance In fact students felt more comfortable when they were accompanied by a teacher in this case in a virtual way by using the web resources provided by eMersion 8 3 Conclusions In this chapter two pilot experiences carried out with the remote experimenta tion system developed in this dissertation have been presented an
121. del with these characteristics should be fit to provide a non parametric description by means of transfer functions of the system Figure A 3 Based on the above mentioned the value of the parameters calculated for the non parametric description of the process have been obtained by analyzing the last step change of the overall time window 3401 330 E 320 E E B 320 300 310 280 300 1 05 11 1 15 1 2 1 25 1 3 1 35 1 4 1 45 15 2 24 22 23 24 25 2 6 seconds x 10 seconds x10 a H1 step in u1 with u2 constant b H1 step in u2 with ul constant 230 220 E 210 amp 200 190 1 05 14 1 15 1 2 1 25 1 3 1 35 1 4 1 45 15 130 2 24 22 23 24 25 2 6 seconds x10 seconds x10 c H2 step in ul with u2 constant d H2 step in u2 with u1 constant 50 50 g 40 g 4 8 30r 8 30r LJ LJ 3 20 5 20L i 1 05 14 1 15 1 2 1 25 1 3 1 35 1 4 1 45 15 10 2 24 22 23 24 25 2 6 seconds x 10 seconds x10 e Step in u1 with u2 constant f Step in u2 with ul constant Figure A 2 Temporal window to calculate of the model parameters From data in Figure A 2 it is possible to get the parameters of the first order transfer functions that relate the action of each of the pumps with the liquid levels inside the tanks T and 75 i e the representation in blocks diagram of the process such as is shown in Figure A 3 A 1 The Three tank s
122. e Hence by studing the degree of interaction among variables we can decide both the variables pairing and the control strategy To analyze both pairing and interaction variables it is neccesary to get the steady state gain matrix SSGM and the relative gain array RGA Morilla 2005 V squez amp Morilla 2002 For the system under study the calculus of the SSGM is carried out by appling limit s gt 0 to the transfer functions which relate the input flows Q1 and Q2 with the liquid levels H1 and H2 see equations A 8 and A 9 Thus the SSGM obtained is Table A 2 SSGM for the threetank system SSG Mthreetank Q1 Q2 H1 2 4 0 925 H2 1 167 0 95 To get the RGA the following relation for n x n systems same number of input and output variables is used Bristol 1966 rga ssgm inv ssgm A 10 By replacing the SSGM in the equation A 10 we get the RGA Table A 3 RGA for the threetank system RG Athreetank Q1 Q2 H1 1 9 0 9 H2 0 9 1 9 According the pairing criteria based on RGA each output should be paired with the input that presents the relative gain closest to unity avoiding any pairing with negative relative gain Thus by observing this matrix the best pairing to A 1 The Three tank system Modelling and control 203 get the less interaction degree between control loops would be controlling H1 with Q1 and H2 with Q2 On the other hand it is necessary to verify whether pairing causes unstability i
123. e integration of Ejs applications into eMersion is first introduced Then the in clusion of the automatic bookings and authentication system as a new external application in the environment is shown Finally the last section shows the integration of the virtual and remote con trol laboratory of the three tank system described in Chapter 4 into eMersion A summary of the most important steps to follow to obtain the final result is provided 150 7 Integration of Layers 7 1 Linking layers So far the experimentation and e learning layers have been presented separately to maintain a certain order before reaching the final result Thus this chapter is intended to describe the final linkage between both layers in order to get the system working as a whole As mentioned in previous chapters web based experimentation for educa tional purposes should provide the opportunity for students to generate on the fly data registers of their experiments in both simulation and remote mode for post analysis In the particular case of this work these data fragments are stored in the eJournal workspace Since the output of a component of the experimenta tion layer as the Ejs console is going to be the input for a component of the e learning layer such as the eJournal the linkage between both should be solved About this subject two things have been addressed in this disseration 1 The definition of the type of data fragments generated from the e
124. e software Overvoltage 16 V max Analog Outputs Channels 4 channels 12 bit Output Range 10 V Output current 10 mA max Digital I O Input lines 8 TTL compatible B 1 The Three tank system datasheet 225 Output lines 8 TTL compatible Timer Counter a Timer chip 9513 Timer resolution 50 nsec Encoder Inputs Input channels 4 single ended or differential Inputs A B Index Input frequency max 2 5 MHz General Power consumption 100 mA to 5 V 50 mA to 12 and 12 V Operating temperature 0 to 50 C 32 to 140 F Connector 2 x DB 37 a Interface PCI 226 B Hardware Description B 2 The Heatflow system datasheet B 2 1 Hardware Figure B 4 Heatflow hardware in the Automatic Control Lab UNED The heat flow experiment HFE system consist of a duct equipped with a heater and a blower at one end and three temperature sensors located along the duct The power delivered to the heater is controlled using an analog signal The fan speed can also be controlled using an analog signal Fast settling platinum temperature transducers are used to measure the temperature Fan speed is measured using a tachometer and can be used to design speed controllers Table B 2 System parameters Heater maximum power V pu P 49 Blower maximum Aem at max speed 8 409 CFM Maximum blower speed 5 V Input Minimum blower speed 1 V
125. e Internet remote lab http lab dia uned es 8080 Pllgina HTML para motor3 Mozilla Firefox DER eJournal Control Language Position 175 seconds U P I D Status video Reading video from camera v Augmented PID control parameters e d wei weg Position 202 5 Speedpisec 51 6 O Speed Control Position Control Refl 2025 Uvots gt u 5 Time sec 182 7 Information Remote mode with labse El connected TIME LEFT sec 684025 Figure 5 9 The experimentation console of the DC Motor experiment Figure 5 9 shows an example of an experimentation console When working in remote mode the Ejs applet connects to the real equipment located in the laboratory through JiL server such as depicted in Figure 5 7 Complementary information Through this module students have direct access to the theory and practical contents needed to carry out an experimentation session satisfactorily This in formation is available in text files in any electronic format pdf or doc or as HTML web pages located anywhere on the Internet The documentation provided to students must be designed so that it can be followed in an easy and intuitive 126 5 The e Learning Layer Interfaz Motor de corriente continua INTRODUCCION INTRODUCCION En este motor de 1 TAREAS EN MODO SIMULACION Y REMOTO 1 e identificaci n del modelo tas en el apartado 6 2 del gui n de pr cticas secci
126. e Labs Applications International Journal of Online Engineering 4 3 42 45 APPENDICES Modelling and Control This appendix describes the modelling and control strategies applied in the de velopment of the virtual and remote control laboratories presented in Chapter 4 of this thesis First one the mathematical model of the system is introduced Second one simple PI controllers have been obtained for each process based on these models Finally tracking references and performance tests to demostrate the reliability of controller designs are presented Note The information provided in the next sections have been used for devel oping the simulations and local control VIs of each didactical setup 198 A A Modelling and Control The Three tank system Modelling and control A 1 1 Modelling The model obtained through physical relationships of the variables is first de scribed Mass balance equations and Torricelli s law are here applied Next we complement the modelling of the process obtaining a non parametric representa tion of the system transfer functions through identification techniques Mathematical model The mathematical model of the process is described by the following balance equations dhi S Q1 Q13 Quieak A 1 dh B Q13 Q32 Q3leak A 2 dh a Q Qa Q20 Qaleak A 3 where t represents time 1 ha and ha correspond to the liquid level in each tank A represents
127. e access etc One on the main reasons that encourages the use of Lab VIEW is its simplicity since users with a low knowledge of programming can develop complex programs impossible to write using traditional programming languages Blume 2007 Because of the hard constraints in its sampling period reading from the sen sors execution of the control code and writing to the actuators the execution of the control task must take place as regularly as possible With LabVIEW such regularity is achieved by using threads and DAQ board interruptions However to manage a local version of the control task through the Internet implies some changes For example the LabVIEW code that handles the communication with the remote client and with the video camera that provides the visual feedback has to be included Leva amp Donida 2008 Figure 3 13 shows a general scheme of this idea LabVIEW Server Ejs Client 3 E Graphical Communication Task p n User Normal Priority Loop Interface Communication methods N enom 4 ey Decode video APA M oritical data and render Loop Priority Loop Local Simulation Figure 3 13 Distributed control architecture using LabVIEW and Ejs The three boxes in the server side represent each of the tasks required Since the control task must keep the constraints in the cycle time the highest execution priority must be assigned to this block in order to guarantee its determinism The communication and video t
128. e around the normal operating point In this case the experiences have been done around the following working point in steady state Q1 19 56 Q2 43 24 H1 300 mm and H2 200 mm 0 5000 10000 15000 0 05 1 15 2 25 seconds seconds x10 a H1 step in ul with u2 constant b H1 step in u2 with ul constant 260 240 E 220 720 200 pr 200 1 fi 180 0 5000 10000 15000 o 0 5 1 15 2 25 seconds seconds x10 c H2 step in ul with u2 constant d H2 step in u2 with ul constant 100 100 8 50 8 50 5 1 5 5000 10000 15000 0 5 1 15 2 25 seconds seconds x 10 e Step in u1 with u2 constant f Step in u2 with ul constant Figure A 1 Open loop experiences for identification 200 A Modelling and Control Figure A 1 depicts six graphics The plots corresponding to the first column show the temporal response of the variables to control levels in tank 1 and tank 2 when an input step is applied to the first manipulated variable pump 1 while the second one pump 2 is kept constant In the same way the three graphics of the second column show the evolution of these variables when the input step is applied to the second manipulated variable pump 2 while the first one pump 1 remains constant It is possible to appreciate that the system presents a similar response to a first order system with delay and therefore a mo
129. e following chapters describe how each layer is implemented as well as the complete process used to reach the final product which is the re mote experimentation environment from the Automatic Control Laboratory of the Department of Computer Science and Automatic Control of UNED Part II IMPLEMENTATION The Experimentation Layer Overview This chapter presents an in depth analysis of the so called Experimentation Layer that covers the methodology used in the design and creation of web based control laboratories for remote experimentation The approach considered to create web enabled virtual laboratories will be ex plained A particular emphasis is placed on the creation of virtual labs apply in the teaching learning of automatic control Technical considerations to take into account when adding tele operation func tions to virtual labs are also indicated Thus the final approach presented here after considers the possibility of working both in virtual and remote mode using the same graphical user interface in the client side Finally a new approach to develop novel virtual and remote control laboratories using Easy Java Simulations and LabVIEW is presented This framework should help developers in the creation of web enabled environments oriented to remote diagnostics maintenance and experimentation 30 3 1 3 The Experimentation Layer Development of virtual laboratories It is well known that to create a dynamic
130. e tank system 85 arrows up and down Furthermore they can also tune the control parameters for optimal performance by checking the PID tab whereby a second window lets users adjust the values of the controller parameters The Plotting view area displays signal scopes of the controlled and manip ulated variables It is located at the top right hand part of the interface The upper scope plots the levels of the two tanks T1 and T5 the controlled variables and their setpoints The lower scope displays changes occuring in control vari ables that is the voltages applied to the pumps Finally the Indicators panel located on the right lower part contains a set of numerical indicators displaying the instantaneous value of the main variables of the system such as the simulation time the tanks level in millimeters and the setting of the pumps and setpoints Through a menu bar located at the top of the main window users can switch between manual automatic control mode by clicking in the Control menu and also by changing the language of the interface with the Language menu Steps to take in order to include remote experimentation options to the virtual lab are described hereafter Section 4 2 4 specifically describes modifi cations made to this template the virtual lab in order to add these features to the application 4 2 3 Local control of the three tank system in LabVIEW The first step in the development of the server side component
131. ed set of methods makes facilitates the connection and the exchange of data with JiL server For example the _external connect method calls the openVI String modelPath String initCmd method from jil Jil class in ternally using the information provided by the developer when s he introduces the URL of the remote target VI and associates variables as shown in Figures 3 32 and 3 33 Similarly the _external disconnect method calls the closeVI method from jil Jil class internally to close the VI and disconnect from the server side On the other hand once the connection has been established and the VI has been executed calling the external runVI method the _external step method must be used in the Ejs evolution to start the data exchange This last method contains the sequence of get methods necessary to know about the state of the remote VI i e the value of indicators The _external synchronize 76 3 The Experimentation Layer is used to update the current value of every control in the remote VI that have been connected to the external variables page on an internal level it calls the set methods This means the exchange data between Ejs and LabVIEW is performed in an asynchronous way given that Ejs is receiving the state of the process continously while a sending is produced just when user interacts with any element of the view Authors can now use the Ejs model variables in the construction of their view either for visu
132. elop ment of this research especially my co director of thesis Professor Jos S nchez Moreno for his constant support help and guidance He has taught me his excel lent work method which has enabled me to be where I am at present At the same time I would like to thank the whole of the research team of the Department of Computer Science and Automatic Control of the UNED for its great and valuable support in particular to Professors Natividad Duro Raquel Dormido Canto Se basti n Dormido Canto Mar a Antonia Canto Joaqu n Aranda Alfonso Urqu a Carla Mart n Roc o Munoz and Jos Manuel D az I would also like to thank Professor Fernando Morilla Garc a for his useful pieces of advice and ideas stem ming from his experience and his constant support in the development of remote practical experiments carried out with students Thanks once again to all of them My thanks also to other fellow students who like me tried to advance in their respective fields of investigation Miguel Angel Rubio Dictino Chaos Vic torino Sanz David Moreno Oscar Cambra Mar a Guinaldo Ernesto Fabregas Luis Cubillos Alejandro Moreno and Jes s Chac n At this point I would like to thank my dear friend Gonzalo Far as Thank you Gonzalo for the help you gave me at the beginning of my work and for our exciting talks on the topic of research Thanks also to Pilar Riego the secretary of the department During this time she became my friend and
133. elop web based environ ments is of great importance In fact many alternatives are available for such a purpose PHP 2009 Visual NET 2008 AJAX 2009 However choosing which the most convenient is not an easy task Although this dissertation does not deal with the low level development of eMersion but actually just uses it some infor mation regarding the web technology underlying this web based experimentation environment is provided hereafter Jakarta struts framework eMersion was built using the framework Struts used to build web applications Struts is an open source framework that helps develop web applications based on the Model View Controller MVC design paradigm Cavaness 2005 The Model represents the business logic or database code the View represents the page design code and the Controller represents the navigational code Struts puts Java servlets Java Server Pages JSP custom tags and messaging resources together into a unified framework and saves the developer the time of coding a complete MVC model Nguyen 2006 Figure 5 2 shows the request answer mechanism of Struts User actions are intercepted by a servlet controller that processes the requests If the request implies dynamic processing in the server as http struts apache org http java sun com products servlet reference index html 3http java sun com products jsp 120 5 The e Learning Layer Web Container http request Web browser use
134. ements The direct application of the augmented reality concept and its benefits in remote experimentation will be shown in the prototypes of web based control labs presented in the following chapter 3 3 A new approach to connect Java and Lab VIEW 61 3 3 Anew approach to connect Java and LabVIEW 3 3 1 Motivations The deployment of web based control environments is beneficial both for final users and developers process control engineers and teaching staff However developers face the extra work required to transform an existing local system into a web based environment Developers know how to manage the hardware and the software in a local control system well but new problems arise when making the system accessible through the Internet Specifically an interactive graphical user interface for the system must be set up such that can be deployed through the Internet in the form of a Java applet Java applets are the simplest way to integrate interactive user interfaces for remote supervision into web based learning management systems Examples of this use of Java are eMersion Gillet et al 2003 PIDstop Eikaas et al 2006 and Java based SCADA systems Fan et al 2005 Let s consider the case of a virtual instrument VI program created with LabVIEW Publishing a LabVIEW VI on the Internet is a long familiar click and share feature of this software Shirer 2001 Nitu et al 2005 Adamo et al 2007 However there is no simple mecha
135. entials and timeslot are fine 6 2 A simple point to point authentication protocol 137 4 If the ok message received the client application can start the data ex change with the real plant located in the laboratory Otherwise the client interface will inform the user why access was denied via a warning text message The built in implementation of the LabVIEW Indentity checking module can be seen in Figure 6 6 The LabSQL toolkit was used to enquire the database from LabVIEW LabSQL is a collection of open source VIs developed by Jeffrey Travis Studios LabSQL 2009 that use the ADO object collection in LabVIEW for users to connect to almost any database perform SQL queries manipulate records etc Execution flow 1 Receiving user s credentials 2 Checking user s credentials with database 3 Checking whether timeslot is correct 4 Sending authentication result to client TCP reference gt rre feel EN TCP reference gt Figure 6 6 LabVIEW dentity cheking module Extract of the block diagram Table 6 1 presents a descriptive summary of the internal tasks carried out by each VI of the LabVIEW Identity checking module Table 6 1 Description of the main Vls for the Identity checking module Block Description READ Read User and Pass Gets username and password encrypted in Base64 of the client from the TCP channel Check Check User and Pass Queries the database to verify whether username and p
136. ep Real Time amp STOP area a circular shaped led indicates whether real time constraints are kept in every iteration of the loop and a variable named stop handles the stopping of the VI target in a controlled way Finally three waveform chart components display the temporal evolution of the system The first one plots the position in degrees of the motor the second one shows its behaviour in speed in degrees sec and the third one depicts the control signal applied to the motor in volts The following sections describe the final virtual and remote control laboratory used by the students to carry out the experimental sessions across the Internet The interface created with Ejs allows for the handling of the VI through the predefined view graphical elements of Ejs that is all the functionality of the VI is replicated in the client application by Ejs 108 4 Prototypes Developed 4 4 3 The virtual and remote laboratory Figures 4 31 and 4 32 show the four main views of the web based laboratory of the DC motor created with Ejs One can see that the position in the interface of the graphical element is again quite similar to the previous prototypes men tioned This means the usability of the applications is enhanced Students can observe a similar structure in every laboratory thus learning time to understand and assimilate interface options available when they move from one laboratory prototype to another is therefore reduced Visua
137. erformance and accuracy To learn more about M Series technologies device specifications and information on recommended cables and accessories please refer to the data sheet and specifications M Series 2009 B 3 The DC Motor datasheet 233 Technical features Two 16 bit analog outputs 833 kS s 10 digital I O lines 32 bit counters digital triggering Correlated DIO 2 clocked lines 1 MHz 37 pin D Sub connector to reduce connectivty costs by 80 Select high speed M Series for 5X faster sampling rates or high accuracy M Series for 4X resolution NI DAQmx driver software and NI LabVIEW SignalExpress interactive data logging software Driver Software M Series devices work with multiple operating systems using three driver software options including NI DAQmx NI DAQmx Base and the Measurement Hardware DDK M Series devices are not compatible with the Traditional NI DAQ Legacy driver of LabVIEW Figure B 9 shows the NI DAQmx driver used to control the NI PCI6221 board from LabVIEW x DAQmx Data Acquisition DAQmx Te 1 lt T Channel Node Timing Node Triggering Node Read Node Write Node h e g E X Aa o s x DAQ Assist Real Time Dev Config Task Config Ctrl Advanced Figure B 9 LabVIEW driver DAQmx for NI PCI 6221 Application Software Every M Series data acquisition device includes a copy of NI LabVIEW Signal Express so you can quickly acquire analyze
138. erver Bi connected di TIME REMAINING min 43251844 b Remote mode using augmented reality Figure 4 18 The virtual and remote control lab of the three tank system 96 4 Prototypes Developed These options available in the menu allow users to save for example a snap shot of the Plotting view area as an image file gif or the evolution of the main state variables as a binary or text file dat or txt in the local hard disk We have finished describing the first prototype of a web based control labo ratory A brief summary of the steps to follow in order to develop it and deploy it later in the Internet is provided hereafter Figure 4 19 is used as a reference to specify the most important issues to take into consideration threetankReal vi Web Server Web browser points to i T Ts is hostname threetank dia uned es HTML document containing a Ejs applet downloaded from Web server Figure 4 19 Connection scheme via web browser between Ejs and LabVIEW The Ejs application controls the threetankReal vi through the JiL server These steps can be enumerated as follows 1 LabVIEW server side apps JiL server local control VI The JiL server and the local VI are running in the server and waiting for client connections The server can be reached from any host in the Internet 2 Ejs client side application A Java applet must be generated from the standalone application created wi
139. es implementation con straints when a constant sampling period is often assumed str m amp H gglund 2006 The notion of real time control is based on the above implementation constraints where computer operations rely on a deterministic time occurrence setpoint sp Control a algorithm time triggered Figure 3 12 Feedback control scheme The development of this type of applications in control engineering educa tion and especially in virtual and remote experimentation across the Internet is commonly achieved using LabVIEW software Salzmann et al 2000 3 2 3 Using LabVIEW for remote experimentation LabVIEW is a graphical programming language developed by National Instru ments in 1986 LabVIEW 2009a LabVIEW was originally developed for data ac quisition and instrumentation control However the lastest versions of LabVIEW allow users to use it for many other purposes process control industrial au tomation modelling and simulation digital signal processing remote operation real time programming etc Programs developed with G the graphical language of LabVIEW are named VI Virtual Instruments due to their instrumentation 46 3 The Experimentation Layer related origins A VI is made up connecting multiple blocks from the existing libraries of blocks meant for many purposes vision I O hardware mathemat ical calculations simulation Internet protocols Travis 2000 process control databas
140. es in the lab Colaborative web services instant messaging discussing forums shared data etc f Applet Java lab 1 Applet Java lab 2 Applet Java lab 3 Applet Java lab N Figure 2 4 e learning platform architecture At the end of the development process the experimentation and e learning layers are integrated in order to get the final web application to publish the virtual and remote labs Integration means performing certain links between 2 3 Conclusions 25 web modules coming from both layers for example to save the data collected from an experimentation Applet experimentation layer into a shared web space belonging to the e learning layer The data stored in this space could be retrieved later for analysis or even be read again from the experimental Applet 2 3 Conclusions Remote and virtual control labs alone do not provide all the benefits expected Students need to organize their work and be in contact with other users stu dents teachers tutors This is the reason why this chapter has presented a set of requirements and specifications to apply in order to develop a remote ex perimentation environment for engineering education The approach used was divided into two layers to provide a certain structure to the development process The planning of the global architecture of the system is a mandatory step prior to starting developing a web based experimentation environment for educa tional purposes Th
141. esigned a EX WU CEU to ease the learning process However the network of remote laboratories in control o einn engineering AutomatL bs is not only the sum of its parts but it must be appreciated as a laboratory with a uniform structure regardless where the location of the physical processes is To work in AutomatL bs all that a student needs is a Java enabled browser and to be registered into the system So AutomatL bs must not be considered just the simple juxtaposition of the different remote laboratories of each university taking part into the project All the laboratories share a common working scheme and the learning materials have been carefully designed to allow students make the practical experiences in an autonomous way at their own pace The project is coordinated by Professor Sebastian Dormido Bencomo of the Department of Computer Science and Automatic Control of the Spanish University of Distance Learning UNED For Further Available Remote Systems information on the project click on the Project link for detailed instructions on the user environment click on the Connection link AutomatLGbs Demo Do not forget to check the access Requirements 426 before running the demo Three tank system Multivariable control of a three coupled tank system P ds P dubbed What is eMersion eMersion is a web based collaborative experimentation environment Focused to the management and publication of virtual and remote labor
142. ess to the laboratory s real resources The following chapter describes this external web based component integrated in the eMersion environment with more details Collaboration and interaction module The eJournal This component plays a key role in the development and deployment of eMersion as a web based experimentation environment The eJournal is the module that enhances interaction and collaborative work in eMersion In this workspace students can save retrieve share and even exchange documents between each other or among workgroups Figure 5 12 shows the eJournal GUI The objects susceptible to be importable exportable or exchangeable in the eJournal are called Fragments Fragments can be files of any semantic content and students can use them to prepare their session reports http lab dia uned es 8080 eJournal Folder View Mozilla Firefox DER eJournal eJournal a _ sub space Default A Active Journal Folder Folders gre l sub space Al v since ME Ld hd Active Folder Filter by Type Filter by Date Refresh Mew Rename Delete Zip Fragments Copy Move Delete Rename Import Export Share Send Assign Finalize Submit Note Name Author Task State Creation Annotation Report DC motor remoteMode pdf Vargas Today remote Fragments Report DC motor simulationMode doc Vargas Today simulation sub space registroMotor testl m Vargas J Today Text file GraphSpeed gif Vargas Today GIF
143. f didactical plants with fast dynamics and real time constraints One should highligth the high degree of interactivity that these applications developed with Ejs provide compared to previously recorded applications Students can understand all the intuitive and strategical aspects of the models suggested via simple interactions with the interface sliders buttons checkboxes etc or even directly over the view of the process itself virtual representation of the process The e Learning Layer Overview So far the various aspects related to the building of virtual and remote experi mentation labs have been analized and described through three practical exam ples However these experimental applications do not fully cover all the current requirements of the conventional face to face experimental sessions In a conventional laboratory students can interact with others ask questions to the teacher take notes about the work carried out save data registers of their ex periments compare and contrast results with other class mates etc This chapter provides a detailed analysis of how this problem was addressed in this thesis The pedagogical aspects to take into account in remote experimentation are men tioned first In this sense a global vision of the features that a flexible learning environment should provide to effectively support a community of practices in a distributed scenario is presented Finally the approach chosen to addres
144. f2real_lab 0 a refreal2 ind EJ Comment Page comment Figure 4 12 Linkage of the Ejs and LabVIEW variables completed in a similar way as described in Section 3 3 6 the starting or stopping of the target VI Labview AXIS saveFragments setLanguage Video ODEs f Interface ControlSignal public void connectButton if _external isConnected _external connect if _external isConnected external runVI _play else _external disconnect reset Figure 4 13 Method to be launched when connect button is pushed 91 of controls and indicators of the target VI Finally the variables linkage is Step 2 Built in methods in Ejs are used to program the logic of the connection with the VI that controls the three tank system published by the JiL server These methods belong to an Ejs internal Java class so called _external An instance of this class is automatically generated once the External Variable Page has been created The Java code presented in Figure 4 13 shows how these methods were used to link the three tank system s virtual lab to the JiL server At low level the methods to open and or close a connection with the JiL server are used and simultaneously the methods to control 92 4 Prototypes Developed In Figure 4 13 the custom method connectButton initializes the con nection with the server When connection is estab
145. face APIs Salzmann 2005 TCP and UDP APIs are available to user applications to communicate data Figure 3 11 shows how data are encapsulated from the moment it is generated at the application level until leaving the host at link level The header of a packet depends on the protocol used TCP or UDP TCP is a connection oriented service with transmission control i e the desti nation confirms the data received and it can request a retransmission of any lost data On the other hand UDP is a connectionless protocol and therefore there is no checking whether packets have arrived safely to their destination User applications access network protocols using Data Application network APIs TCP UDP TCP UDP Transport header data i IP data Internet IP header Frame Frame data Fiene Link header footer Figure 3 11 Encapsulation of data descending through the protocol stack From a remote experimentation point of view we have to remember that the behaviour of the Internet is not predictable since we have many connections occurring at the same time and therefore transmission delays may occur along the transmission path In this sense UDP protocol is more suitable than TCP because it does not need the mechanisms used by TCP to establish and maintain connection with peers Thus UDP is a more lightweight protocol that does not suffer the drawbacks of TCP networks so much Nevertheless in this dissertation TCP based solutions have
146. g gested methodology step by step in order to develop web based laboratories designed for pedagogical purposes As mentioned at the beginning of the chap ter these prototypes are part of the current infrastructure of laboratories that the Department of Computer Science and Automatic Control of the UNED provides students for remote experimentation For this reason the three prototypes were fully tested using performance and usability criteria 4 5 Conclusions The analysis design and building of three real prototypes of virtual and re mote control labs to perform practical experiments through the Internet have been described in detail The benefits of using the JiL server approach to cre ate web based control labs using Java for the client side and LabVIEW for the server side have been shown successfully The developing time needed in the case of a new remote experimentation environment using both software technologies has been reduced The creation of a generic communication module independent from the control task makes the system more reliable and easy to maintain From a client perspective each example is a novelty in terms of its devel opment The three tank MIMO system presents several interface configuration options in virtual mode The heatflow system has a very pleasing visualization effect that shows the graphic design possibilities that Ejs provides Finally the DC motor system describes how to deal with the remote control o
147. g to the performance requirements It is also possible to see that practically there is not delay between the change of the reference and the control action The effect of delays can be better appreciated when performing control around the sensor 3 because its location is farther from heater and hence a small lag between the control action and its effect over the system can be observed Hardware Description This appendix provides practical information about the hardware of the didactical setups presented in Chapter 4 of this dissertation Furthermore the technical features of the DAQ boards used to sense actuate on the mechanisms from a Lab VIEW program will be described In this context three different DAQ boards have been used to control the plants in the lab the DAQ MF614 from HU MUSOFT s r o the DAQ Q8 from Quanser Consulting and DAQ NIPCI6221 from National Instruments While the last one can be accessed by using the LabVIEW native driver library DAQmx the second one uses the propietary LabVIEW driver software supplied by the provider In the particular case of the DAQ MF614 the LabVIEW driver have been written from the scratch in C code In this way the access to the hardware has been standarized to use LabVIEW for controlling every process 220 B Hardware Description B 1 The Three tank system datasheet B 1 1 Hardware Figure B 1 Three tank hardware in the Automatic Control Lab UNED The three tank system c
148. ger 4 bytes 2 147 483 648 to 2 147 483 647 int Single precision 4 bytes 1 40e 45 to 3 40e 38 or float Double precision 8 bytes 4 94e 324d to 1 79e 308d or double String 1 byte per character simple text message string Byte signed integer 1 byte 128 to 127 byte With regards to the third point a simple way to switch between simulation and remote mode must be included Thus when the interface is running in sim ulation mode the variables associated to the system state are updated according to the evolution of a mathematical model of the process written with the Ejs built in ODE editor However in remote mode these variables are true mea surements Finally regarding the fourth point an interface with too many sliders but tons graphical elements scopes and auxiliary windows incorrectly distributed 3 2 Adapting virtual labs for remote experimentation 51 would puzzle users and thus decrease their motivation Gillet et al 2003 As in the server side concurrent tasks executing common actions must be pro grammed to add tele operation features to a virtual lab created with Ejs Figure 3 16 describes these tasks and the set of instructions executed within each of them These three tasks should be executed on independent Java threads Task 1 Receiver Get the data packets from server side with the system state Decode the streams of information video measurements others etc Task 2 Sender A
149. gistroThreeTank m Dormido Status Lab Simulation mode no connected 4 gt Terminado Terminado Figure C 1 Global view of the eMersion environment running the virtual laboratory corresponding to the three tank system C 1 2 General description The experimentation environment meets in its interface all the management fea tures to support and sustain remote and virtual labs for engineering education Thus eMersion aims to provide a Web interface which allows students to perform their practical experiences as well as they would do in traditional f2f face to face laboratory sessions The environment wrappes this conceptual structure taking into account that to use remote and virtual labs for engineering education is necessary to dispose of C 1 How to work with eMersion 237 The documentation associated to the experimentation module explanations of the theory to solve the problems brief description of the hardware to be remotely manipulated description of the graphical user interface used to carry out the remote experiences in simulation and remote mode and finally a description of the experiments to perform The client interface A Java applet created by Ejs to interact with the process By means of this GUI is possible to perform the hands on activ ities using a mathematical model of the process virtual lab or accessing remotely to the real plant in the lab through Internet remote lab The ma
150. gnals for the pumps in the range 10 10 V Outputs of the signal adaption unit Liquid levels of cylinders 10 V 9 V 0 em 8 61 V 60 cm 2 control signals for the pumps in the range 0 10 V Controller Option 200 02 A D D A card and executable controller program for a PC with PCI slot running with WINDOWS for WIN98SE ME 2000 XP The controller program provides a simple menu control sampling period default 50 ms adjustable in the range 1 54 ms Inputs 3 sensor signals with 10 10 V at 12 Bit A D converter Outputs 2 control signals with 10 10 V at 12 Bit D A converter and amplifier Program C source Option 200 03 The controller program from Option 200 02 in C source with library functions for graphical outputs controlling the sampling period as well as controlling the PC adapter card requires the Borland C Builder V6 Electrical disturbance module Option 200 05 The three sensor signals can be disrupted by switches or scaled by potentiometers The control signals for the pumps can be scaled via potentiometers B 1 2 DAQ MF614 Humusoft Multifunction I O The MF 614 multifunction I O card is designed to connect PC compatible com puters to real electrical external signals The MF 614 contains a 100 kHz throughput 12 bit A D converter with sample hold circuit 4 software selectable B 1 The Three tank system datasheet 223 input ranges and 8 channel input
151. he JiL Server implementation The JiL server was developed in LabVIEW making use of its VI Server fea ture The VI Server is a collection of functions that allows to access objects and functionalities of a virtual instrument programmatically Travis 2000 Any VI exposes properties and methods These can be accessed through block diagram functions A property is a characteristic of a virtual instrument that may be either read and or written depending on the property An example of a virtual instrument class property is Execution State which indicates the execution state of a VI bad idle and running are possible values of this property A method is an action that is performed on a virtual instrument For example the Run VI method is used to run programmatically a VI in the same way as pressing the run button in the VI s user interface manually Figure 3 25 shows a template of how LabVIEW Invoke Node blocks are used to execute methods and set or get property values AR Virtuallnstrument RC Property f Close Vl Reference Open VI Reference Invoke Node d R Virtuallnstrument B H i j Figure 3 25 Use of two Invoke Node blocks to modify a property and invoke a method on a virtual instrument n VI Fi putput Indicators Ctrl Yal Get All Flat p n beer 55 Flat Controls o is putput Controls 53 wv 85 Ctrl Val Get All Flat a rp Controls Figure 3
152. he first document to be consulted at starting to work The second link called Interface provides a functional description for each element of the tele C 1 How to work with eMersion 239 manipulation console A third link called Guide contains the theory to afford the development of the required tasks system description mathematical model guidelines to fullfil the sequence of activities Finally two additional links are also available as appendixes The first of them contains an brief summary about first and second order system analysis Appendix 1 the second one provides a short description of the Matlab tool ident for system identification Appendiz 2 md UN The fifth and sixth icons refresh update the navegation bar and exist of the system respectively The Current task area contains a link named Access Protocol This link allows to get access to a HTML document describing how to use the eMersion environ ment It also shows a indicator of the current task Finally the Complements area is reserved to include new web components or plug ins to the experimenta tion environment Currently only the news plugin is available This functionality is used as an agenda to report news about the laboratory C 1 3 2 Experimentation console As it was above mentioned the experimentation console is the Java applet created by Ejs This application allows the student to carry out their practical activities in virtual and remote mode
153. he metadata information generated in LabVIEW as an array of clusters Each element of the array contains all the necessary information to describe a variable in LabVIEW For instance one variable could be described by its name and its data type 1 lt xml version 1 0 encoding iso 8859 1 gt 2 lt MetadataControls gt w l ocoo 1o0m 29 lt MetadataControls gt Listing 3 6 XML variables descriptor lt Variable gt lt Name gt intin lt Name gt lt Datatype gt int lt Datatype gt lt Variable gt lt Variable gt lt Name gt boolin lt Name gt lt Datatype gt boolean lt Datatype gt lt Variable gt lt Variable gt lt Name gt stringin lt Name gt lt Datatype gt string lt Datatype gt lt Variable gt lt Variable gt lt Name gt doublein lt Name gt lt Datatype gt double lt Datatype gt lt Variable gt lt Variable gt lt Name gt stop lt Name gt lt Datatype gt boolean lt Datatype gt lt Variable gt The JiL server uses an XML document see Listing 3 6 to describe and to send the metadata information extracted from a VI to Ejs The tags of the document define the properties of each element of the array of clusters variables allowing the possibility of adding more properties according to the requirements To access 3 3 A new approach to connect Java and Lab VIEW 73 the variables and their properties in Ejs the JDOM library for Java is used to parse the XML doc
154. he need of previous knowledge about object oriented graphical programming 4 New features to develop interactive simulations are continuously being added to Ejs Dormido et al 2005b Jara et al 2009a 3 1 2 Structure of a generic virtual control lab in Ejs Virtual labs in the field of control engineering education are used by teachers and trainers to reinforce the comprehension and understanding of the control theory 34 3 The Experimentation Layer They enable students to appreciate the dynamic behaviour of a physical system when either a control algorithm or a parameter is changed There are some guidelines to be considered in the design of virtual control labs with pedagogical perspectives The complexity of the application should be considered carefully since users using the simulation generally students from the university will not have the on site support of a teacher Some interactive elements rich in visual contents should be included in the design of the view of the labs to explain ideas concepts and methods present in the automatic control theory A set of practical activities to complete with the virtual lab should be pre pared Students will then be able to follow the sequence of tasks often with an increasing difficulty until reaching a complete insight concepts taught A simulation in Ejs will be achieved when all the information provided by developers in the description of the mode
155. heck the students perception regarding the use of the environment Part III ASSESSMENT System Assessment Overview Capturing the students perception of their learning experiences is a key issue to evaluate the quality and value of the experimentation environment as a teaching tool Two real experiences were carried out to test the remote experimentation system developed in this dissertation The UNED pilot experience and the Au tomatLObs project During the UNED pilot experience students from this institution were encout aged to use the new web based remote experimentation system to perform their practical activities Some preliminary results of this experience were extracted and analyzed in this chapter In the AutomatL bs project a more exhaustive test of the system was performed The experience consisted in integrating virtual and remote control labs from other Spanish universities to produce the first Spanish network of web based control labs This study was designed to increase the quality and robustness of the net work of virtual and remote laboratories aimed at a higher number of students and teachers with different teaching and learning concerns In this sense a set of graphics will show the main results of this evaluation 162 8 1 8 System Assessment A first experience from the UNED In order to evaluate the first version of the experimentation system a pilot ex perience was carried out during the academ
156. ibed in this chapter follow a common structure so as to comply with these requirements The three tank system disposes of a boolean variable called stop to finish the application so that when this variable is pushed in the LabVIEW application the execution gets out from the main while loop and the reseting code is then executed before the end 90 4 Prototypes Developed of the application see RESET HW sequence in Figure 4 9 The reseting code could just consist in zeroing the pumps and finally reseting the hardware or any additional code in order to move the real process to a known and safe state The three tank system also has a hardware safety system that automatically zeroes the pumps when the liquid exceeds a certain level Once the VI is ready to run in local mode the application is published by the JiL server Developer is only required to place the main VI of the local control application into the apps folder of the JiL server Thus the publication procedure ends here and no additional work is needed Summarizing the JiL server helps developers to fulfill the safety requirements above mentioned since Y If the connection with the client is lost the JiL server automatically looks for a boolean variable named stop in the local control VI to finish the application in a safe way controllability Y Extra reliability and maintainability were gained by creating a generic com munication module independent from the contro
157. ic year 2006 2007 Figure 8 1 shows the website which students used to access the remote experimentation system uned m tica y Autom tica Laboratorios virtuales y remotos Laboratorio Sistema 1 Sistema 2 Sistema 3 Departamento de Inform tica y Autom tica wj die el a Heatflow Tres Tanques Informaci n gt Presentaci n Presentaci n Presentaci n Laboratorio Virtual ene Bienvenido al laboratorio virtual y remoto del Departamento de Inform tica y Autom tica de la UNED Estas p ginas contienen la informaci n preliminar necesaria para que el estudiante que est matriculado en la asignatura de Autom tica 1 de la especialidad de F sica Industrial en la carrera de Ciencias F sicas de la UNED pueda acceder a nuestro entorno de simulaci n y IESENE Deus iare pera eater ol core de praia de berto que son necesarias para poder superar la asignatura Sistemas disponibles Agradecimientos Requerimientos Acceso a Easy Java Simulations Recursos en la red EL mim ARA Figure 8 1 UNED pilot experience home page Students worked on the three prototypes of virtual and remote laboratories located at the Department of Computer Science and Automatic Control the three tank system liquid level control the heat flow system temperature con trol and the DC motor speed and position control The evaluation was completed during the summer term of 2007 with 5
158. id init start connection and run the VI jil Jil vi new jil Jil example uned es 8080 try vi connect vi open apps example vi initCmd vi run catch Exception e System out println Error when connecting to the VI try setting the value of the VI controls vi setValue boolin con new Boolean true vi setValue intin con new Integer 1 vi setValue doublein con new Double 0 5 vi setValue stringin con Hello world catch Exception e System out println Error when setting VI controls OMA O Ct 0 I0n OcO c0 1o0 amp wWN N y A a v N 23 24 try reading the value of the VI indicators 25 System out println bool vi getBoolean boolout ind 26 System out println int vi getInt intout ind 27 System out println double vi getDouble doubleout ind 28 System out println String vi getString stringout ind 29 catch Exception e 30 System out println Error when readingVI controls 31 32 try stop the VI and close the connection 33 vi stop 34 vi close 35 vi disconnect 36 catch Exception e 37 System out println Error when closing the connection wo 00 end of the init method po oo e ja 42 end of class 3 3 6 Integrating the API Java in Ejs The previous section provides all the information required by an expert Java
159. igure A 14 Tracking reference in speed control DC motor On the other hand Figure A 14 displays the closed loop response of the system when controlling speed The tracking of reference in speed control is a bit simpler than position control because there is not presence of an integrator in the control A 3 The Heatflow system Modelling and control 213 scheme The control parameters has been calculated to get an over dumped response under step changes in the setpoint The experiment starts with the system running at 50 sec in steady state At 60 sec the speed reference value is changed to 140 sec As previously mentioned there is not overshoot in the response over dumped response and the new setpoint is reached in around 3 sec It also can seen that control signal is not saturated for this range of changes in the references A 3 The Heatflow system Modelling and control A complete thermodynamic model of the system is difficult to derive In principle the following equation applies Tn PV Vp Ta 5 A 23 where Tn Temperature at the sensor n V gt Voltage applied to the blower V Voltage applied to the heater Ta gt Ambient temperature X gt Distance of the sensor n from the heater Next sections are aimed to study the dynamic behaviour of the system both in open and close loop very simple first order model will be obtained and then a feedback controller will be developed In subsequen
160. iles The applications developed with Ejs can be easily integrated into the eMer sion environment since they are prepared to interchange data with the eJournal Ue1s s quej 99441 BY Suisn UOISSAS IJOLWIY apedey S UOISIB AJS q 7 ANSIA DID opeunue SS a a 660119 fosa 1331 SPI payoeuucs S J9AJ9 QE YU BPOW AJOWOY QET SANS TN 18 61C luxu cH TOL 995 ouny Ta es sauvrpraca A W E Pozj tuuzas OP luulas una mei oci c et sauerprasa uruns og10 139H J ES 96 zdund 0 69 36 dung B 9 8Z ux zH 19887 wu 1H Sc sdy 16 zdung 9 1duing pass s3uerpras3 urumyu oHI109H Nanso MINAS 20 VINIL 1 Spucoss I rf 6d squetpnasy WITNIS OHI0029H J oz ost 001 os 0 Q KORY 0 eov 0 OA sausrprasa UFUNS SIOISH O C squetprasz UIUBJESIO 9H J 001 Zw 001 ZEW 001 LE LW sauerprasa wm nuod OSA 0d TomNoD KINOS UR TP sz wpupeu j aum O PIDIO OI Copia ujpray 09pia says i i j i owt ody Aq souls Joppo3 o noy s19p 0j user rumor aanv ry E naaa l qeumopa 3utodjes H ZH LH jueuiDe1j ejeq OUTUTETTETETTUTHETTMT 330 0507014 23955 ssaJboud ui sse SS9U9S IP Avg se JUS LI S 0808 p qet Aday sjeuuo 40d pue TALH wajsks shunjooq Jeunope sseds ajosuos uoqe ndivew Ieuinore ay 0 uonejuaun3o 3410 ssad9y aaneioqejjoa
161. image registroMotor m Vargas J Yesterday Text file GraphPos gif Vargas J Yesterday GIF image Figure 5 12 The eJournal module of the eMersion environment The eJournal GUI is composed of three sub spaces This division corresponds to the functionality provided by each sub space A simplified description of them is provided below eJournal Sub space It allows students to have access to the eJournal space from other groups as long as these groups have granted them access This 5 2 eMersion A novel approach from EPFL 129 feature fulfills the essential characteristic of web experimentation environ ments of Discretionary Collaboration seen in Section 5 2 1 Furthermore a trash icon serves to receive Fragments deleted by users The Fragments deleted can be reloaded or definitively deleted Finally the internationaliza tion using the I18N feature of Struts framework of the application is also included Languages currently available are English French and Spanish Folders Sub space The Folders sub space helps organize and deploy Frag ments inside the eJournal Users can create folders where they put and classify their Fragments as they do in the hard disk of a computer A folder with special features named Inbox is automatically created to re ceive data Fragments from other web components of the experimentation environment The Fragments deployed in the Fragments sub space can be filtered by type and creation date from t
162. in features of the environment are described in the next section The functional architecture of each module that constitutes the system is also pre sented C 1 3 Interface and eMersion functionalities Figure C 1 shows the remote experimentation environment facade The function ality of each work module that compose the system is described hereafter C 1 3 1 Navegation bar The navegation bar is located on the top of the eMersion Web application This window allows users to get access to the set of Web modules which compose the environment The Objective field contains a description about the fundamental objective pursued with the experimental activity On the other hand each icon located inside the Navegation area is used to access to the different Web modules of the system and at the same time to the documentation associated to each practice The icons are described below gt ls The client interface of a remote and virtual lab it also so called Tele manipulation console is obtained by pushing this icon This window allows students to carry out all the laboratory experiences such as they did it in the f2f laboratory 238 C eMersion Management The eJournal the collaborative and shared space for handling data fragments and user groups is obtained by clicking on this icon In this workspace students can save store and retrieve documents such as work reports images with the evolution of the experiments and or dat
163. ingle tank process Furthermore the same simulation running as a Java applet embedded in a HTML page can be seen in Figure 3 7 In this sense the advantages of using Ejs are obvious since the tool allows the on line execution of virtual labs through any Java enabled browser Home page for SimpleSingleTank Mozilla Firefox Archivo Editar yer Historial Marcadores Herramientas Ayuda O 2c day O fileIIC JEjs3 46 061119 Ejs Simulations SimpleSingleTank app SimpleSingleTank html Contents mank eyel Single tank process Simulation E Created with Easy A Java Simulations 100 150 200 250 300 D Manual Automatic sp 0 085 u 3 281 Ti 0 8 y 1 2 3 You can control it using JavaScript For example using buttons Play I Pause II Reset I Step Figure 3 7 Publishing the virtual lab of the single tank process as an applet 3 2 Adapting virtual labs for remote experimentation 39 3 2 Adapting virtual labs for remote experimentation Virtual laboratories alone do not cover all the benefits of hands on traditional laboratories where students can use real equipment Moreover it is also a fact that virtual labs using only simulated systems can not replace working with real processes in a conventional laboratory This means it is fundamental that virtual labs possess tele operation functions since the acquisition of experimental skills using real processes is of param
164. into eJournal Otherwise it is saved in the local hard disk Console _saveText registroMotor m generaMfile gt El DialogLogin iil E Y StringBuffer sb setLength 0 3D Drawables e Ok Cancel e NN WEE T uo BESA Figure 7 1 Ejs built in methods to link the applet to the eJournal workspace 152 7 Integration of Layers Figure 7 1 shows the use of these methods in the DC motor prototype They are invoked via the action property of buttons SaveGraph and SaveRegister When these methods are invoked they are internally launched in an indepen dent Java thread in order not to interrupt the normal execution of the main loop of the Ejs application 7 http Hab dia uned es 8080 eJournal Folder View Mozilla Firefox eJournal Default Folders Inbox x All S Since S ij bu L3 LJ L3 Active Folder Filter by Type Filter by Date Refresh New Rename Delete zip Fragments O Copy Move Delete Rename Import Export Share Send Assign Finalize Submit Mote Name Author Task State Annotation O SpeedControl gif Estudiante x E Today GIF image Creation O opentoop m Estudiante Today Text file gt GraphPos gif Estudiante x E Today GIF image Figure 7 2 Saving data from the Ejs console to the eJournal space Figure 7 2 shows where the buttons are located Here it is imp
165. ions since we are focusing on the technical aspects of the building of the web based laboratories using the approach mentioned in the previous chapter 4 2 Prototype I The Three tank system 81 4 2 2 The three tank system s virtual lab in Ejs The step by step creation of the virtual laboratory for the three tank system is described hereafter The starting point implies to define the process model used in terms of their variables which hold the different possible states of the process and relationships among these variables Defining the Model A detailed explanation of the main Ejs s windows whereby the model is defined is presented below This description focuses on the Ejs sections that will be modified later to add tele operation capabilities to the virtual lab Variables Eight pages define the model s variables Common Tanks Pipes Leaks Pumps Status PID Control setLanguage Each has a name that identifies it according to its purpose see Figure 4 3 Common Tanks Pipes Leaks Pumps States PID control setLanguage Name Value Type Dimension xmin D int xmax 120 int ymin 30 int ymax 100 int r lengthx xmax xmin int lengthy ymax ymin int t 0 double t_ejs D double dt 5 double g 981 double fps 25 int npoint 100 int skip D int colorWater java awt Color BLUE Object showLoginDialog false boolean Com
166. ious sections described the authentication process that occurs when a book ing is carried out locally by the administrator in the database Now the extension of this approach toward a flexible scheme is presented Based on Figure 6 5 the teacher administrator has been replaced with an automatic bookings modality Figure 6 9 depicts the new scenario Client applet 4 dao for bookings java Centralized Bookings DB LabVIEW i Identity checking module Physical resource Middleware interface Lab Server any location Figure 6 9 A flexible scheme for bookings and authentication process This scheme enables students to book themselves The basic idea consists of filling out the reservations database automatically from the client side by means of a web interface The full system is composed of three new applications which all provide this functionality In the case of the first application a Java applet was developed to perform new bookings on the client side see Client applet for booking in Figure 6 9 With regards to the second application a centralized server application also developed in Java manages reservations synchronism and communications between the client applet for bookings and the Lab Server see Bookings Main Server in Figure 6 9 Finally an additional Java module located in the Lab Server was developed see Java Interface in Figure 6 9 This module informs the Bookings Main
167. iro P Soares F Monteiro C Carvalho V amp Vasconcelos R 2009 A Remote System for Water Tank Level Monitoring and Control a Collaborative Case study in Proceedings of the 3rd IEEE International Conference on e Learning in Industrial Electronic ICELIE Porto Portugal Callaghan M J Harkin J McGinnity T M amp Maguire L P 2006 Client Server Architecture for Remote Experimentation for Embedded Systems In ternational Journal of Online Engineering 2 4 9 Bibliography 183 Casini M Prattichizzo D amp Vicino A 2004 The Automatic Control Tele lab A Web based technology for Distance Learning EEE Control Systems Magazine 24 3 36 44 Cavaness C 2005 Jakarta Struts Desarrollo de aplicaciones web con Servlets y JSP Anaya Multimedia S A Cefalo M Lanari L Oriolo G amp Venditelli M 2003 The REAL Lab Re mote Experiments for Active Learning in Proceedings XLI AICA Annual Congress Trento Italia Christian W amp Esquembre F 2007 Modeling Physics with Easy Java Simu lations The Physics Teacher 45 475 480 Costa Castell R Vall s M Jim nez L D az Guerra L Valera A amp Puerto R 2010 Integraci n de dispositivos f sicos en un laboratorio remoto de con trol mediante diferentes plataformas Labview Matlab y C C Revista Iberoamericana de Autom tica e Inform tica Industrial RIAI 7 1 23
168. is on waiting mode 3 3 4 API Java to control LabVIEW applications Applying Figure 3 23 to this scheme a JiL enabled Java client must implement two Sender and Receiver communication tasks to interact with the published VI The Sender task is to send control commands or state commands to set the values of VI controls whenever the user interacts with the client s user interface The Receiver task periodically reads the values of VI indicators sent from the Sender of the JiL server For example in a web enabled control application values sent from the Java client can be controller parameters and commands to start stop the feedback control loop and the data stream received back will con tain data on the state of the process the level of a tank the temperature in a heat exchanger the angular position of a motor etc These communication tasks can be easily implemented in a Java program by using the methods provided in a utility class called Jil class in the package called jil Tables 3 2 and 3 3 list the methods provided by Jil main class The first method to invoke on a Jil object is the connect method which es tablishes the connection with the server Then openVI O is used to open a remote VI and report the list of control and indicators of the remote VI to be exchanged This internal list acts as a buffer to improve the communication performance 3 3 A new approach to connect Java and Lab VIEW 67 This means that by reporting
169. ities in engineering education Specific conclusions The JiL Server methodology proposed in Chapter 3 to create web based labora tories experimentation layer and the developed software tools were successfully applied to develop the following web based remote and virtual laboratories for control education 176 9 Conclusions and Future Research Virtual and remote lab of a DC motor a temperature control system and a coupled threetank system in the Automatic Control Laboratory of the Department of Computer Science and Automatic Control of the UNED Virtual and remote lab of a one tank system located in the Department of Computer Science and Languages of the University of Almer a UAL Virtual and remote lab of a roto magnet system controlled by magnetic fields located in the Department of Systems Engineering and Automatic Control of the Technical University of Catalu a UPC Virtual and remote lab of a four variable system a multifunctional indus trial equipment designed and developed by the Automatic Control group of the University of Le n UNILEON eMersion as software tool for publishing remote and virtual labs covers all the design requirements of the e learning layer Specifically eMersion allows to maintain the pedagogical aspects of interaction and col laboration of traditional hands on laboratories when moving towards a flex ible learning model based on the Internet eMersion organ
170. izes and puts all the complementary learning resources to use virtual and remote labs at the disposal of students User s manual theory assignments task protocols and a complete instruction manual of the global system are available at the experimentation environment Furthermore the eJournal provides a shared space where students can save and organize the results of their practical activities and collaborate amongst them and with their teachers The automatic bookings system was successfully integrated into eMersion and used by students when they perform their practical activities The application design suggested helps to optimize the use of physical resources available by including 9 2 Future work 177 A simple mechanism to book resources to work on the real systems located in the laboratory A straightforward module of encriptation and user authentication A centralized backup for bookings E mail notifications of new reservations and or cancelations Finding out about students perception during the learning process is a very important matter to assess an experimentation environment as a valuable learn ing teaching tool For this reason during the two pilot experiences carried out with students UNED pilot experience and AutomatL bs project students were asked to fill out an evaluation questionnaire to determine the pros and cons of the experimentation environment Such results encouraged us to repeat the
171. js Model of the three tank virtual lab Custom methods of the main system variables in a text file during a practical session Fi nally the internationalization of the application is also completed through the use of the Java 118N feature Java 2009 Thus the fifth and last page describing the model contain methods used to switch between English and Spanish languages Defining the View The GUI of the three tank system virtual lab is created based on the set of graphical elements provided by Ejs see Figure 4 6 In order to keep a uniform structure in each lab prototype the interface is divided into four zones whereby each area contains a set of graphical elements designed for a specific function Tree of Elements y Elements for the view e Simulation View 2 amp MainFrame Containers Interface Menu gt Y MenuBar Ji o ES Console SplitPanel LJ de En O PanelPrincipal r Drawables Ez DrawingPanel Basic Bodies Fields y Tank1 eo T A tt EERE a TankSumi r Other S Setpoint1 Control 3D S Setpoint2 q Aa F3 e T setpoint1 T setpoint2 BE mmm Dinon b d Figure 4 6 Ejs View of the three tank system virtual lab 84 4 Prototypes Developed Figure 4 7 shows the applet s main window The Graphical view displays the plant s schematic representation in virtual mode
172. k of the student to his her own workspace in the eJournal snapshots from the evolution signals scopes GIF images and text files representing experimental data registers received from the experimentation console By this option is possible to classify the fragments by these three types by date This option allows to classify fragments depending on the creation date Note It is necessary to refresh the eJournal after each filtering Refresh Updates the main window that contains the data fragments of the eJournal Each time a new fragment is saved or any thing is modified in the eJournal pushing this button is mandatory 242 C eMersion Management New Creates a new folder Be carefull not to use special characters of Windows operating system Rename Modifies the name of a folder The Inbox folder can not be renamed Delete Eliminates a folder The Inbox folder can not be deleted Subspace Fragments The main functions in this subspace are listed below Copy This option allows to copy a set of fragments between eJournals or folders of the same eJournal Move lo move data fragments between eJournals or from one folder to another in the same eJournal Delete By using this option students can delete data fragments Once deleted they are deposited in a new space where is possible to restore or delete them definitively http lab dia uned es 8080 Fragment list Mozilla
173. kings The following booking has been saved Plant 6 Motor de corriente continua 2 UNED Date 28 10 2008 Saved Hours 14 00 15 00 You will receive an email with final confirmation in a Few minutes in this mall box E bvargast bec uned es Select the bookings that you want to delete and press Delete button Plant Date Stat 7 Motor de corriente continua 2 UNED 28 10 2008 zoo E Motor de corriente continua 2 UNED 28 10 2008 i400 You vil receive an email with the final confirmation in a few minutes in this mall box hvargas bec uned es e Stage 5 f Stage 6 Figure 6 12 Interfaces involved in the reservation process 146 6 Access Control to Experimentation Resources 6 3 2 The automatic bookings system server An administrator for the booking system is in charge of the server configura tion Figure 6 13 shows the graphical user interface of the Bookings Main Server whereby the administrator can configure the system s main parameters This server located at the Department of Computer Science and Automatic Control of the UNED is responsible for storing all the reservations made in the system This way this computer must be connected via the Internet to every Lab Server available So any Lab Server connected to the Internet could benefit from this bookings service Plants management o gt 4 Bla Bola y vga UPC
174. l and the view is assembled Once the description is ready then Ejs is enabled to generate the Java program a stan dalone application or an applet reproducing the phenomenon under study A general view of the simulation s algorithm working behind the applications created by Ejs can be seen in Figure 3 2 The code is divided into two stages of execution Initialization and Evolution The first one is devoted to the declara tion of variables and the initialization code needed to describe the initial state of the system The second one covers the algorithms that define the dynamic behaviour of the system during its evolution Inside the Evolution is where both the graphical state of the system and the interactive changes from the user are carried out Two blocks are intended for this task In the first one the view recovers the current values of the variables from the model and transforms them into a graphical representation model view communication The second block is always waiting for user s interactions view model communication The algo rithms of the Evolution are iteratively executed while simulation is playing to get the expected result 3 1 Development of virtual laboratories 30 Declaration of variables Variables panel Initialization of variables Value column in the Variables panel A Execution of the algorithms in the Initialization panel i Initialization Y Evolution i Execution of the algorithm
175. l task 4 2 4 The virtual and remote version of the laboratory In order to let the three tank system s virtual lab to connect and manage the local control VI published by the JiL server certain programming modifications must be made in Ejs Much of this work has already been explained in the previous chapter Section 3 3 6 and therefore the description of this case study reinforces the correct application of this approach Steps taken to link the virtual Ejs application of the three tank system to the VI published by the JiL server are described hereafter Step 1 A new External Variables Page to link the Ejs and the LabVIEW variables is created Figure 4 12 shows how this was done for the three tank system On the upper part the URL 1abview 62 204 199 106 2055 gt threetank threetankReal vi points to the VI threetankReal vi pub lished by the JiL server If the URL is correct then Ejs receives the list 4 2 Prototype I The Three tank system Pumps States PID control Vars Lans setLanguage Video LabVIEW Common Tanks Pipes Leaks l ee totemai Fle 1atven62204 190 1062056 ireetariatvestanieati Name Value Type Dimension Connected to t_lab D double time ind gt hi lab D double hi ind h2_lab D double h2 ind h3_lab D double h3 ind ul lab D double usatl ind u2_lab D double usat2 ind reflreal lab 0 double refreall ind re
176. lement of the MVC paradigm and fuses one part in the view and the other one in the model Thus applications are created in two steps 1 the model to simulate by using the built in simulation mechanism of Ejs and 2 building the view showing the model state and answers to the interactive changes made by users From here on the term view will be used to refer to graphical user interfaces GUI 3 1 Development of virtual laboratories 33 mm oe e 5S we E T Bry bB er ERES Ea su 12m es Comment Verlat algorithm for the problem Figure 3 1 Simplified model view control paradigm in Ejs Ejs has been chosen in this dissertation to create the virtual labs and the client interfaces for remote experimentation for four main reasons 1 The tool offers a set of built in numerical methods to solve ordinary differ ential equations The chosen method is automatically called by the built in ODEs editor to solve the models 2 The MVC paradigm is assembled by Ejs This process is hidden to users which means they do not need advanced programming knowledge to get started with the tool 3 The composition of complex graphical user interfaces is carried out with a simple drag and drop actions using predefined graphical components Thus Ejs practically eliminates t
177. lished the remote VI is then started by calling the _external runVI method Figure 4 14 depicts the modifications to make in the Evolution pages of the virtual lab In this case an if else instruction has been added to differen tiate between simulation or remote mode When working in remote mode the _external step method is invoked to obtain the values of all indi cators from the JiL server Otherwise the state of the process is updated by solving the ODEs modelling the process Control Dynamics Frames persecond if it is connected REMOTE MODE MAX if _external isConnected external step jelse if it is not connected SIMULATION MODE ul controlSignal pidControl1 hl setpointi 01 20 uZ controlSignal pidControl2 h2 setpoint2 Q2 if manualControl Q1 ul 15 Q2 u2 manualol Q1 manualQ2 Q2 10 refreall setpointl refreal2 setpoint2 jelse Q1 manualol 2 Q2 manualQ2 setpointl hl MIN setpoint2 h2 FPS 25 Autoplay Comment Figure 4 14 The _external step method is called in every iteration of the Evolution during the connection As described the external synchronize method is used to report any user interaction with the GUI to the JiL server B Properties for Pump1 element of type Slider Variable Graphical Aspect Variable manualol Ticks le size
178. lization and control options of the interface are described below Menu options Through this menu upper part of the interface users can choose the lan guage of the interface Spanish or English the control mode manual or automatic and finally are able to save data registers obtained during a working session by using the eJournal options available Graphical representation of the DC motor The rotational movement of the steel disk is reproduced by a copy of this component a photo of the disk This image provides the same visualization effect when working in either simulation or remote mode Control panel of the laboratory A set of graphical components to control the dynamic behaviour of the mo tor is located under the process scheme Users can choose between position or speed control see Figures 4 31 a and 4 31 b They can also set both the setpoint value in closed loop control and the manual voltage applied to the motor when working in open loop Three numerical fields allow to change the parameters of the PID controller on the fly Four buttons allow the users to play pause or reset the simulation of the process when working in virtual mode The latter allows connection to the real plant to work in remote mode Some text fields are also added in the upper and lower part of the control panel to display connection state messages 4 4 Prototype III The DC Motor 109 Plotting panels On the right side two signal scopes sho
179. lowing characteristics 3 3 A new approach to connect Java and Lab VIEW 63 To minimize network traffic by sending data only when needed To set and get parameters with different data types from the client side through the use of variable names hiding the TCP implementation details To send metadata information to ease the data packaging and parsing To decrease the delays caused by the communication network and make better use the bandwidth the exchange data between client and server is done asyn chronously This key idea is based on the fact that users do not interact with the application continuously For this reason communication between client and server is unidirectional in every step of evolution the server sends the state of the process to the client being bidirectional only when users interact with the graphical interface the client sends input variables only when the user makes an interaction over the simulation Client Server Communication module 4 Sender Parser Thread 4 Thread TCPIIP link Asynchronous exchange Acquisition and Control loop Thread Receiver Thread Figure 3 23 Command based architecture As Figure 3 23 shows the server side performs three tasks the Command Parser the Sender and the Control Loop The Command Parser receives com mands from the client interprets them and executes actions requested When no request is received the
180. lowing students to gain a better insight of the process Users can interact directly with this view for instance by moving the apparatus around its centroide and obtaining different perspectives of the structure see Figure 4 24 b Heater and blower have also been represented at the end of the apparatus The color of the heater changes and the blower turns around its axis according to voltage applied E heatflow system e Journal Control Language 15 seconds U P I D Status video No video in simulation mode _ Augmented INTRO SENSOR PD J Setpoint S1 C 53 U volt 3 5 a E 10 15 20 2 30 seconds Disturb it 3 turbance volt 3 Setpoint S1 C 153 36 U Heater volt 3 46 Play E Pause JC Reset JC Connect Time sec 31 3 Status Lab Simulation mode BB heatflow system e Journal Control Language T1 T2 T3 Setpoint 10 seconds U P 44D Status video No video in simulation mode E Augmented contor SENSOR PID gt Setpoint S1 C 63 f U volt 4 1 a i 10 15 seconds Disturb It 3 Fps 25 e pa Temp S1 C Setpoint S1 C 63 35 LO 7 d Temp S2 9C 53 1 U Heater volt 4 05 Play Pause Reset Connect Temp S3 C 5178 Time sec 192 Status Lab Simulation mode ia no connected f4 b Another 3D perspective of the view Figure 4 24 The web based laboratory of the heatflow sy
181. mber 2008 Title of project AutomatLObs Red de investigaci n para la innovaci n docente en Au tom tica mediante laboratorios virtuales y remotos 1 4 Publications Awards and Projects 15 Honorable mention in the final competition Science in Action in the modality Science and Technology Title of the contribution AutomatL bs Una red de laboratorios virtuales y remotos para la ense anza en control au tom tico Granada Spain September 2009 Research projects The results obtained in the framework of this dissertation have been supported by different research projects Interactive tools to the modelled visualization simulation and control of hybrid systems 2004 2006 Spanish Ministry of Education and Science CICYT Ref DPI2004 01804 National University for Distance Education of Madrid Spain Directed by Prof Sebasti n Dormido Bencomo Control of complex systems for logistic and production of goods and ser vices COSICOLOGI CM 2005 2008 IV PRICIT Autonomous Region of Madrid Spain Ref 0505 DPI 0391 Directed by Prof Sebasti n Dormido Bencomo Event based modelling simulation and control 2007 2012 Spanish Min istry of Education and Science CICYT Ref DPI2007 61068 National University for Distance Education of Madrid Spain Directed by Prof Sebasti n Dormido Bencomo Environment Global Architecture Overview The first step before starting to develop a remo
182. ment Page comment Figure 4 3 Ejs Model of the three tank virtual lab Variables Evolution In this section two pages were created Dynamics and Control The first one shows the differential equations editor of Ejs with the defini tion of the mathematical relationships representing the three tank process see Figure 4 4 a The second page displays a small fragment in Java showing the logic that governs the system during its evolution see Fig 82 4 Prototypes Developed Frames Dynamics Control per second Indep Var t_ejs es increment ct e MEE State Rate dh ejs dhi hi h3 A13 Aperl 20 d t ejs d h2 d tejs dh2 h2 h3 A32 A20 Aper2 15 dhs d tejs dh3 h1 h2 h3 A13 A32 Aper3 10 d hsumi d t ejs dhsumi hi h2 h3 Aperi Aper2 Aper3 A20 5 MIN FPS 25 Solver Runge Kutta 4th order x Events 0 Autoplay Comment a ODEs Fanaa Dynamics Control per second if it is not connected SIMULATION MODE E MAX double ul controlSignal pidControll hl setpointl 01 double u2 controlSignal pidControl2 h2 setpoint2 02 if manualControl 20 Q1 ul m Q2 u2 manualol Q1 15 manualo2 Q2 refreall setpointl m 10 refreal2 setpoint2 jelse Q1 manualoQl 5 Q2 manualQ2 setpointl hl MIN setpoint2 h2 FPS 25
183. ment The receiver method recovers the incoming data sent from the LabVIEW server These methods are integrated into the definition of the Model and View of the virtual lab of the single tank process in Ejs Once the methods have been added to Ejs the programming logic that discriminates between working on simulation or remote mode is created The updating of the variables in simulation is carried out based on the evolution of the mathematical model or in the case of remote mode on real measurements obtained from the server 3 The Experimentation Layer Listing 3 3 Java code at the client side to receive data from the server side 1 public void receiver if connected try time in readFloat read time from server level in readFloat read level from server qautomatic in readFloat read input flow from server catch java io IOException e System out println Method receiver Error receiving data OMAN DA FW ND PRR eB oer 0 Listing 3 4 Java code at the client side to send data to the server side public void sender if connected try out writeBoolean m a write control mode out writeFloat qmanual write input flow in manual out writeFloat Kp write proportional gain out writeFloat Ti write integral gain out writeFloat Td write derivative gain OoOmA NOahWwWN EH m o out writeFloat ref write setpoint out flush flush data to client catch java io IO
184. mmed at the client side Figure 3 20 shows the camera SONY EVI D31 used to display video images in the virtual and remote control laboratory of the three tank system presented in the following sections of this chapter Although most existing IP cameras in the market are all in one solutions that is camera and video server in this case the hardware to capture video images the camera itself is connected to an external video server able to deploy the frames through its network interface to the users Video server and camera are connected using their serial ports Figure 3 20 Camera SONY EVI D31 with external video server AXIS 2400 The external video server AXIS 2400 allows to connect up to four cameras which can be used independently Furthermore the AXIS video server is prepared so as to receive incoming command from clients to make specific functions on the camera It allows for example to change the zoom view to move the camera to switch between cameras connected to the server etc All these functions can be performed from a remote graphical user interface by means of HTTP requests The second alternative to obtain visual feedback from the remote process is a low cost option The main drawback of this approach is that it requires a higher programming effort since it is necessary to create a software to acquire the images and publish them on the Internet However it allows for a better control of the video streaming For example in
185. mote visualization module Firewall E Remote audio module pel Internet p Audio streaming l Video streaming c client 2 3 Users data E Resources booking C Moddes gt Experiments E Bookings data base Instrumentation a Didactical set up ae ul Figure 2 3 Server side implementation requirements 22 2 Environment Global Architecture Although there are a high number of hardware software tools that can be used for the purpose of the present dissertation we used LabVIEW from National Instruments for the implementation phase see next chapter From a software design point of view the server is composed of a set of modules some of them optional described below A data exchange module This software remains in a listening state waiting for remote connections from users It receives commands and queries from clients and makes them effective over the physical system Responses are retrieved from the real plant through the instrumentation hardware and fi nally are sent back to the client The link between the server and clients is established via Internet and therefore the TCP IP protocol suite becomes the backbone for the information exchange This module which is labelled with in Figure 2 3 is of a paramount importance as it is in charge of implementing the listening server that is al ways waiting for remote connections from clients Furthermore the module
186. multiplexer 4 independent 12 bit D A convert ers 8 bit digital input port and 8 bit digital output port 4 quadrature encoder inputs with single ended or differential interface and 5 timers counters The card is designed for standard data acquisition and control applications and optimized for use with Real Time Toolbox for MATLAB However in order to standarize the use of LabVIEW as a software tool to interact with hardware a LabVIEW driver to control the card has been written see Figure B 3 Figure B 2 Humusoft data acquisition board MF614 and LabVIEW driver hudaglib Figure B 3 LabVIEW driver for Humusoft data acquisition board MF614 Because of its small size and low power consumption the MF614 can be used not only in desktop computers but also in portable computers Eight single ended 12 bit analog input channels Four 12 bit analog output channels 224 B Hardware Description Sampling rate up to 100 kHz 8 digital inputs 8 digital outputs Programmable A D ranges Four quadrature encoder inputs differential Five counters timers Low power consumption Driver for Real Time Toolbox for MATLAB included Driver for Real Time Windows Target Driver for xPC Target Driver for Windows Specifications Analog Inputs Channels 8 single ended a A D converter 12 bit 10 usec conversion time Input ranges 10 V 5 V 0 10 V 0 5 V a Trigger mod
187. n mechanism A communication scheme between two computers based on the Internet Suite Protocols is shown in Figure 3 10 which represents a typical scenario in remote experimentation over the Internet a client application receiving information from a server application s k blocks and sending data to it c k blocks through the Internet The dialogue between both hosts can be described as an abstraction in layers of the path followed by data from the source to the destination The data are first generated at the application level and then passed down to the underling layers Throughout this transit the data are wrapped with the functionality pro vided by each layer through a new set of protocols and definitions The number of layers depends on the network device and its functionality For instance in Figure 3 10 routers are represented just by two layers Link and Internet Effec tively routers only need the Link layer to connect physically to other equipments and the Internet layer to provide packet routing capabilities The IP suite is well known to most of the researcher community and therefore no additional information will be introduced here More information about IP Suite can be found in Dost lek amp Kabelov 2006 3 2 Adapting virtual labs for remote experimentation 43 3 2 1 2 TCP or UDP as data transportation mechanism User applications only see a transmission channel through the network Applica tion Programming Inter
188. n basado en web para estudios de ingenier a un caso pr ctico in Proceedings of the 1st Congreso Espanol de Inform tica CEDI EIWISA Granada Spain Guzm n J L Vargas H S nchez J Berenguel M Dormido S amp Ro dr guez F 2007a Distance Education Issues and Challenges NOVA Pub lishers pp 131 167 Hahn H H Spong M W 2000 Remote Laboratories for Control Educa 9 Bibliography 187 tion in Proceedings of the 39th TEEE Conference on Decision and Control Sydney Australia Heiming B amp Lunze J 1999 Definition of the Three Tank Benchmark Prob lem for Controller Reconfiguration in Proceedings of the 1th European Con trol Conference Karlsruhe Germany Hu W amp Bao A 2008 Application of Internet Based Collaboration Model in Engineering Education A Case Study in Proceedings of the 2008 Inter national Conference on Computer Science and Software Engineering IEEE Computer Society Washington DC USA pp 178 181 Jara C A Candelas F A amp Torres F 2008 An Advanced Interactive In terface for Robotics E Learning International Journal of Online Engineering 4 4 17 25 Jara C Candelas F Puente S Pomares J amp Torres F 2009b Practical experiences using RobUALab ejs a virtual and remote laboratory for robotics e learning in Proceedings of the 8th IFAC Symposium on Advances in Control Education ACE
189. n closed loop The Niederlinski theorem Niederlinski 1971 is used to check the stability of the system in closed loop The theorem postulates The final closed loop system obtained from pairing u1 y1 ua Ya and Un Yn for a n x n process is unstable if the cocient between the SSGM determinant and the product of the elements in the diagonal is negative that is _ SSGM Ge del NI lt 0 A 11 By using the previous formule for the threetank system it is obtained 2 4 x 0 95 0 925 x 1 167 0 5265 A 12 2 4 x 0 95 N Ithreetank As the NI obtained is positive we need additional information to guarantee the stability in closed loop This rule is simple and its use is convenient just steady state information is needed This information is easily obtained during the design process The interaction study of the variables suggests that a descentralized control strategy can be applied to control the system Decentralized control A controller objective in our system is to control the liquid level in both T1 and T5 From an inspection of the process equations it seems natural to use the pump flow rate Q1 t to control the level in the first tank h t and Qa t to control ha t In this context previous variables interaction analysis has demonstrated that such pairing of variables can be applied for controlling the system This strategy is often called decentralized control which is a simple
190. n through the Internet This set of virtual and remote control labs have been developed using the approach introduced in Section 3 3 6 The steps followed to build the prototypes can be summarized as follows a Designing and implementing a Java applet based on an mathematical model of the real process and an appropiated user interface using Ejs b Creating a LabVIEW VI to control the real process in local mode c Publishing the LabVIEW VI using the JiL server d Modifying the applet to connect with the local VI published by the JiL server The applet s user interface is also adapted so that it can visualize either the data generated by the computer model or the data received from the VI that controls the real equipment These web based applications have been developed to explain PID control concepts in an introductory course on automatic control Nevertheless these could be modified at a later stage in order to introduce advanced control strategies and adapt the labs to higher level courses The framework based on the JiL server and Ejs presents certain advantages when the communication layer is separated from the control loop on the server side First of all a complete overview of the implementation of the three tank system control laboratory is provided in order to illustrate the details of the development methodology This approach is then replicated to describe the other two case studies 4 2 Prototype I The Three tank system
191. nce on Engineering Education ICEE Coimbra Portugal Salzmann C amp Gillet D 2008 From on line experiments to smart devices International Journal of Online Engineering 4 50 54 Salzmann C Gillet D amp Huguenin P 2000 Introduction to Real time Con trol using LabVIEW with an Application to Distance Learning International Journal of Engineering Education 16 255 272 Salzmann C Gillet D amp Mullhaupt P 2005 Real Time Interaction over the Internet Model for QoS Adaptation in Proceedings of the 16th IFAC World Congress Prague Czech Republic Salzmann C Gillet D Scott P amp Quick K 2008 Remote lab online sup port and awareness analysis in Proceedings of the 17th IFAC World Congress Seoul Korea Shinskey F G 1996 Process Control Systems Application Design and Tun ning 4th edition McGraw Hill Shirer D L 2001 LabVIEW 6i adds internet features to data acquisition environment IEEE Computing in Science amp Engineering 3 4 8 11 S nchez J 2000 Un enfoque metodol gico para la ense anza a distancia de asignaturas experimentales Analisis dise o y desarrollo de un Laboratorio Virtual y Remoto para el estudio de la Autom tica a trav s de Internet PhD thesis Universidad Nacional de Educaci n a Distancia Madrid Espa a S nchez J Dormido S amp Esquembre F 2005 The Learning of Control Concepts
192. nd its mathematical model The objective is to control the liquid level inside the tank manipulating the pump that regulates the incoming liquid flow The level controller which is labelled as LC in Figure 3 3 computes its output depending on the change of the liquid level inside the tank or and the tracking reference The system tank pump can be seen as two sub systems connected in cascade so that a simple PI control could be applied to reach the control objective The closed loop control scheme is appreciated in Figure 3 4 Figure 3 4 Closed loop control in the single tank process The continuous PI controller applied to the system has the form u t P IC Kp e t E f ear 3 1 e t h t egt 3 2 For simulation with Ejs the discrete counterpart of the continuous PI con troller will be used Thus the PI controller obtained from the discretization of the continuous model is as follows P t Kp hres t h t 3 3 I t ICE At 2 yep t At A t At 3 4 u t P t I t 3 5 The discrete controller includes a new variable At that represents the simula tion step This simulation step is chosen according to the dynamics of the process to satisfy the Shannon sampling theorem Oppenheim et al 1996 3 1 Development of virtual laboratories 37 The building of the virtual lab in Ejs implies the definition of both the Model and the View Figure 3 5 provides some snapshots of the Ejs wi
193. ndicators and select the one desired Checking the block diagram check box in this dialog enables Ejs to display the VI wiring diagram for additional information Establishing this simple connection between model variables and VI controls and indicators instructs Ejs to include the necessary Java calls to the library jil Jil class integrated in Ejs in the generated simulation in order to pass the values of the variables back and forth as required by the program Connection to and disconnection from the VI is however left explicitly to be coded by the author so as to adapt to this program s logic In this context the methods from jil Jil class have been encapsulated in a new reduced set of built in methods of Ejs that the author can use to program the logical sequence of connection with the Jil published VI Table 3 4 describes these methods and their functionality Table 3 4 Methods in Ejs to link with a Jil published VI Ejs built in methods to control the connection with JiL server _external connect Initializes the TCP IP connection with the JiL server _external disconnect Closes the TCP link with the JiL server _external runVI Starts the VI external stopVI Stops the VI _external step Performs the data exchange between Ejs and JiL Server _external synchronize Updates controls when user interacts with the view _external isConnected Returns the current status of the TCP IP connection This reduc
194. ndows to describe the model of the single tank process virtual lab Introduction Model View O Introduction Model View Single tank process Variables Initialization Evolution Constraints Custom zi Tank Picontrot Single tank process T SS a t o double tMax 200 double dt 0 1 double g 9 8 double R 0 0295 double This simulation shows the control of r 0 0015 double the level of a single tank height lo 2 double The figure to the right shows the ever 0 11 double configuracion of the system o E oe a Introduction O Introduction Model View Variables Initialization Evolution Constraints Custom Tank Plcontrol Name Value Type Dimension automaticMode true boolean setPoint 0 046913580246913555 double setPointX 0 03923705722070845 double deltaY lo double K low 2 72 6 double u lo double Kp 20 double Ti 0 8 double integral o double o double c Variables of the Pl control Introduction Model View _O Variables Initialization 2 Evolution O Constraints Custom b Variables of the tank O Introduction Model View O Variables Initialization Evolution Constraints Custom Frames Dynamics Control per
195. nism to make the VI variables controls and indicators accessible for Java applets Making the VI and Java communicate across the Internet requires modifying the VI s wiring diagram to include additional LabVIEW TCP IP blocks This requirement poses an impor tant setback for developers who want to transform existing LabVIEW based local control systems into web based ones The following sections present an alternative approach and a very promising one for the quick and simple creation of web enabled control environments which use LabVIEW on the local side Java applets on the remote client and TCP IP as the communication channel between both elements The solution is based on A stand alone application called JiL Java internet Lab VIEW server which acts as middleware to publish an existing LabVIEW VI on the Internet and 3 The Experimentation Layer A Java library file which Java clients can use to control and access the variables of the JiL published VI The novelty of this approach is that the controls and indicators of the VI can be accessed from the Java program by the LabVIEW developer in a fully transparent way that is without introducing any modification to the original VI Outcomes of this research have been disseminated in Vargas et al 2009a 3 3 2 Technical issues of the JiL Server middleware layer The built in local communication facilities of LabVIEW allow to include commu nication facilities at design
196. niversity of Catalunia 168 8 7 Four variable system University of Le n 168 8 8 Robot arm University of Alicante o o o 169 8 9 Ball and beam system Polytechnic University of Valencia 169 8 10 Satisfaction enquiry about the practical activities 170 8 11 Regarding the overall system llle 171 8 12 Quality of the web based laboratories len 172 8 13 The most important learning resource 2 ls 172 A 1 Open loop experiences for identification 199 A 2 Temporal window to calculate of the model parameters 200 A 3 Three tank system representation by transfer functions 201 A 4 Descentralized control strategy G s is the three tank system to be controlled C s is the controller to be designed and is composed of two elements C1 and C2 Ci controls the level in tank i The variables hi setpoint ei ui and hi are the setpoint to be reached the error signal the control signal and the liquid level in tank i respectively s a adsa oo s IR a a OR eR ee d 204 A 5 Tracking references Three tank system 205 A 6 Disturbance rejection Three tank system 206 A 7 Model of the DC Motor o e 208 A 8 Typical speed control system o oll 209 A 9 Control strategy for the DC Motor o 209 A 10 PI control for speed control on the DC Motor 210
197. nrich the visualization of the system This polygon has the same shape as the red sector that shows the position on the steel disk If we look at this element we can see the lag between video images and the augmented view is due to the fact that incoming data are coming from two different sources the IP camera and the computer connected to the real plant see Figure 4 32 b GE D C Motor eJournal Control Language Position 129514 16 48 seconds U P I D T T Status video Reading video from camera O Augmented r 151 U 0 1 i e UC Oh 040 PAR tie pee RO PID control parameters Kp 0 08 Ti 0 6 Td 0 Posiionp 152 5 Speed sec 1 4 O z pm O Speed Control Position Control Reff seg 151 7 Ulvolts 0 1 Time sec 22 6 Information Remote mode with labserver El connected TIME LEFT sec BE D C Motor eJournal Control Language Position T i 215 seconds U P I D Status video Reading video from camera Augmented PID control parameters Kp 0 08 Ti 06 Td 0 ous O Speed Control f9 Position Control Refs 2465 b 5 Time sec 224 4 Information Remote mode with labserver B connected TIME LEFT sec b Remote mode using augmented reality Figure 4 32 The web based laboratory of the DC motor in remote mode 4 5 Conclusions 111 So far we have described three real examples for which we applied the su
198. ntroller The VI scheme of this central diagram used for this example is presented in Figure 3 29 The diagram consists of a synchronous while loop with four controls connected to four indicators This has the simple and rather useless effect of showing any modification of a control in the corresponding indicator Figure 3 29 Wiring diagram of the example Once the VI is published using the JiL Server as described before the Java code in Listing 3 5 communicates with this VI The first lines of the initO method of this applet instantiate a Jil object and try to connect it to the VI running at the example uned es host at the 8080 port Once the connection is established the VI example is opened by invoking the openVI method see line 11 It then starts the execution of the VI The following try catch block sets 70 3 The Experimentation Layer the values of the four controls and flushes the buffer The third block reads and prints the value of the VI indicators The VI must have placed in the indicators the values previously sent for the controls according to the VI wiring diagram The final try catch block stops the VI and closes the connection Listing 3 5 Java code to communicate with the VI example of Figure 3 29 public class SimpleJiLTest extends javax swing JApplet String initCmd intin con boolin con stringin con doublein con intout ind boolout ind stringout ind doubleout ind public vo
199. o al ternatives were tested The first one considers the use of an IP camera with a built in video server dedicated to capturing images and post publishing them as video streaming through its network interface see Figure 3 19 option 1 while the second alternative considers using a conventional webcam with serial inter face RS 232 or USB directly connected to the server computer where the server computer could be the same host that controls the real plant The publishing of the video stream is not direct in this case since the image data must be acquired via webcam through its serial interface first and then socket techniques should be used to send these data together to the plant state data across the network see Figure 3 19 option 2 Most of the remote laboratories created during the development of the the sis project work with the first alternative suggested It was chosen for its low CLIENT Remote viewer via internet browser Applet Java INTERNET network i interface Option 1 NO IP camera with built in ir video server T s SERVER Didactical setup in the lab LabVIEW application Option 2 Figure 3 19 Visual feedback of the remote plant through video images 56 3 The Experimentation Layer maintenance the fact that it is easy to use and mainly because it is possible to control the zoom and vision angle of the mechanism by means of HTTP requests progra
200. ocal control of the DC motor real system Front panel 105 Local control of the DC motor real system Block diagram 106 Operations of the JiL server when sending data to a client 107 The web based laboratory of the DC motor in virtual mode 109 The web based laboratory of the DC motor in remote mode 110 xii 5 1 5 2 5 3 5 4 5 5 5 6 5 1 5 8 5 9 5 10 5 12 6 1 6 2 6 3 6 4 6 5 6 6 6 7 6 8 6 9 6 10 6 11 6 12 6 13 7 1 7 2 7 3 7 4 List of Figures Interaction model illustrating the two main modes of on line learning 116 Struts framework Flow of information request answer 120 General abstractions representing the eMersion organization 121 First level of abstraction in eMersion 121 Second level of abstraction in eMersion 122 Third level of abstraction in eMersion 123 Functional structure of eMersion o e 124 The Navigation bar allows access to the web modules of eMersion 124 The experimentation console of the DC Motor experiment 125 HTML Documentation of the DC Motor experiment 126 Examples of external web applications integrated into eMersion 127 The eJournal module of the eMersion environment 128 Point to point authentication protocol between a user and a re mote server providing authentication services 133 Algorithm of the point to poin
201. odule that allows to communicate and decode the data re trieved from the remote site and then render it in the interface Although it is not possible to detail all of the design options available these cover a big part of the characteristics that an experimentation environment should fulfill to satisfy pedagogical and learning requirements in practical engineering ac tivities More details can be found in S nchez 2000 2 2 2 Layer 2 The e learning layer The previous section presented the main features to take into account when analysing designing and building of the experimentation layer Also a second key aspect that should be addressed is the development and or use of a web based learning management system LMS to support the student s learning process This platform should organize user access to the experimentation modules avail able and to allow for students teachers to interact and collaborate with each other The implementation phase would require the following Simplifying the organization of user groups Offering notification services by email instant messaging news etc Providing all the necessary theoretical documentation such as practical guides task protocol instruction s manuals and any other information needed to perform a remote experimentation session 1A learning management system LMS is a software application designed for the administra tion documentation tracking and reporting of t
202. of the web component navigation bar The window is divided into four main areas Objective Navigation Current Task and Aware ness The first contains a description of the learning objectives pursued by course modules The second one covers a set of icons to enable the user to access the different modules that compose the system The third contains an indicator show ing the current task carried out by the user and a link to the Access Protocol a document describing how to handle the environment The fourth area contains additional web components http lab dia uned es B080 Supervision Welcome Hector Vargas Mozilla Firefox UTE Control de velocidad y posicion sobre un motor de corriente continua Navigation Current Task Awareness E Tasks in progress wa al a E i Md Access Protocol el Exi HIE Password Figure 5 8 The Navigation bar allows access to the web modules of eMersion 5 2 eMersion A novel approach from EPFL 125 Experimentation console The so called experimentation console web component contains the applet Java generated by Ejs following the development scheme suggested in the Experimen tation Layer Chapters 3 and 4 Students can conduct laboratory assignments as well as in a traditional lab that is running control experiments using a simulation of the target process virtual lab and or directly accessing the real laboratory through a local network or across th
203. olute path 4 Figure 6 8 JiL server with authentication Setting up and status visualization gt Message gt Message gt Message gt Message gt Message gt Message gt Message gt Message gt Message TCP Write in TCPAS vi 2JIL vi TCP Write in TCPAS vi gt JIL vi TCP Write in TCPAS vi gt JIL vi TCP Write in TEPAS vi gt JIL vi TCP Write in TCPAS vi gt JIL vi ADO has been closed ADO has been closed TCP channel has been closed TCP Write in TCPAS vi 2JIL vi clear e Execution State VI Run top level C Documents and Settings hector Escritorio MOTOR44PP5 JILsmart apps motor2 motor2Real vi b Status tracking logging 140 6 Access Control to Experimentation Resources 6 2 4 User authentication issues in Ejs On the client side the introduction of the authentication module in the Easy Java Simulations interfaces meant following two steps 1 To define a custom method that uses any encriptation mechanism in order to provide a first security level when the password is sent across the Internet In this case we used the encriptation algorithm known as base64 to match the eMersion platform Listing 6 1 shows how this algorithm is coded The non encrypted password is given so as to get the encrypted version Listing 6 1 Java method to encrypt a string in Base64 public static String encBase64 String passNoEncrypted Base64Encoder encoder new
204. on task It implements the necessary means to com municate in order to send and receive data to from a client application It constitutes the s k blocks and sends them to the client At the same time it receives the c k blocks to apply them to the system The more interesting task in the server side corresponds to the Control task since the acquisition of signals reading from sensors and writing to actuators and the closed loop control are carried out Moreover knowledge of systems the ory control and real time must be put into practice here Figure 3 12 shows the standard structure of a feedback control loop Op erations performed in a feedback loop follow a standard sequence Operations 3 2 Adapting virtual labs for remote experimentation 45 labelled a and b in Figure 3 12 respectively represent the measurement of the analogue output signal from the process and its conversion into digital format Once the conversion has been completed b the value is compared to the setpoint value c to obtain the error signal With this information the control algorithm computes the control signal d that is then converted to its analogue represen tation e and applied to the physical process This sequence of operations is known as the control task that must be carried out at a fixed time interval called the sampling period determined in accordance with the dynamics of the physical process This fixed sampling interval or cycle time defin
205. on task through RT FIFO queues blocks The video task is a non time critical activity since the loss of some video frames is generally acceptable for the user Vargas et al 2008 For most appli cations sending five images per second is enough to obtain an adequate visual feedback of the remote system Salzmann et al 2000 Section 3 2 5 presents in detail how the video task has been implemented Finally the communication task concatenates the current measurements state vector and the video frame video vector in a new vector This resulting vector is sent to the client using a Comm WRITE block In parallel the control vector is received through the Comm READ block from the clients and passed to the control task by means of RT FIFO queues The TCP protocol is used in both im plementations because it guarantees packet delivery and bandwidth adaptation albeit at the cost of extra transmission delays Lim 2006 Salzmann et al 2005 As it was mentioned in Section 3 2 1 1 a possible alternative would be the UDP protocol which provides a better control of the transmission delay However UDP neither has a guaranteed packet delivery mechanism nor a bandwidth adaptation mechanism thus the designer is responsible to implement these features So far the generic structure of the server side application using LabVIEW has been presented The client must be able to receive the information sent by the server and report changes in the control parameters
206. onsists of three plexiglass cylinders which are serially interconnected by stop valves The pumps 1 and 2 suck distilled water from the basin thus supplying the two cylinders on the left and right hand side of the set up The level of the liquid inside the three cylinders is supervised via piezo resistive pressure sensors with integrated test amplifiers The basin and the three cylinders present a unit manufactured out of plexiglass see Table B 2 Table B 1 System parameters Parameter simbol vame Gravity acceleration u 0010 om see B 1 The Three tank system datasheet 221 Dimension sizes and weight of the system Length 1270 mm Width 360 mm Height 880 mm Weight 40 kg Pumps Voltage 12 Volt a Current consumption 1 4 A Pump performance 7 l min Pressure 1 4 bar Sensors Supply voltage 8 V Range 0 70 mbar Output voltage 1 6 V Actuator with signal adaption unit The amplifier the mains supply unit and a signal adaption unit are contained inside of a 19 box The box offers a free slot for an electrical disturbance module Mains 230 V 50 60 Hz on request 110 V 50 60 Hz Amplifier inputs Pump control signals 0 10 V Amplifier outputs Voltages for the pumps 12 V PWM Inputs of the signal adaption unit 222 B Hardware Description 3 sensor signals range 1 6 V 2 control si
207. or ind booleancon boolean boolean control con booleanind boolean boolean indicator ind stringcon String string control con stringind String string indicator ind doublecon double double control con doubleind double double indicator ind boolean stop con Figure 3 32 Table of Ejs variables linked to VI controls and indicators The labview keyword identifies the file as a JiL enabled remote VI The IP_address and port values indicate the URL at which the JiL server is listening Finally the VI_relative_path value indicates the location of the VI in the JiL server Once provided the External File field Ejs requests the set of controls and indicators of target VI to JiL server In the case of each variable declared in the table by clicking on the Connected to cell the dialog window in Figure 3 33 pops up allowing the author to browse BB List of external variables C block diagram LabVIEW Indicators LabVIEW Controls int control string indicator stop Connecting the LabVIEW n indicator boolean control ouble indicator string control indicator int indicator to double control the intcon variable on Ejs onnected to model variable intcon Input Output int indicator ind Figure 3 33 List of controls and indicators of a remote VI as offered by Ejs 3 3 A new approach to connect Java and Lab VIEW 75 the list of VI controls and i
208. or can create Courses to which groups of teachers tutors and or students can be associated see Figure 5 5 It is important to notice that students could attend courses coming from different spaces p Space admin Global Administrator L Group of Professors Group of Tutors D D gt n Nor M 1 wa y yo i D 1 PED E y A V Group of Students Groups of Students Figure 5 5 Second level of abstraction in eMersion The third and last level of abstraction corresponds to the Modules created by a Course administrador or professor module can be seen as a laboratory used for a course Several modules or laboratories can be associated to a course in a similar way as occurs in any engineering course where one or more laboratory sessions are carried out to explain different concepts Figure 5 6 depicts the ideas above mentioned As we can see the professor administrator of a course can manage several modules Each module can have one or several groups of students aiming at completing their laboratory assignments using the graphical user interface of eMersion this being the last branch of the tree to configure 5 2 eMersion A novel approach from EPFL 123 N gt Professor admin I Space admin I 1 Global Administrator Selected group Selected group Selected group of students of students of students Figure 5 6 Third level of abstraction in eMersion The organizational structure of the web
209. ortant to notice that Ejs methods internally check whether the Java application is loaded as an applet If this latter condition is satisfied then data fragments will be saved in the eJournal database and will remain available for a later analysis Otherwise the fragment is saved in the local hard disk of the client computer 7 2 2 Defining types of data fragments The virtual and remote control applications developed in this dissertation can save data in two formats as mentioned in the previous section 1 Generic text data files txt with the values of some variables 2 Image files displaying the scope area of Ejs applications gif 7 3 Linking all Work modules in eMersion 153 The temporal piecewise of data to save is what exists in the internal data buffer of the application at the moment of invoking the corresponding method Circular buffers were used to keep the most recent data ready to be saved When data are saved in text format arrays from these circular buffers are turned into strings to use the _saveText method Once the data has been converted to string the Java API to manipulate strings is used in order to structure its presentation Figure 7 3 displays two examples of data fragments saved from an Ejs console Figure 7 3 a shows an example of a generic text file and Figure 7 3 b shows the image of a signal scope of an Ejs view saved as an image file Bl registroMotor m WordPad Position Archivo Edici n Ver In
210. ount importance for students of physical sciences and engineering Garcia et al 2008 At the same time the design and imple mentation of a remote laboratory are more challenging due to safety and fault tolerant aspects which must be taken into account in its development Casini et al 2004 The following sections show how the development of web enabled control labs for remote experimentation has been addressed A brief review of some concepts and technical considerations is offered before discussing the solutions adopted in this work 3 2 1 Background Carrying out the remote operation of any physical device is challenging if we take into account the amount of technical aspects that must be solved Technical issues such as performance interaction level visual feedback real time control user s perception safety and fault tolerance etc are in most cases research topics by themselves Salzmann amp Gillet 2002 In general remote experimentation over real equipment through the Internet implies being aware of the current state of a distant system to change the value of any input parameter of the remote system and to perceive the effect of this change with a minimum transmission delay A client server approach is commonly used to provide this functionality to virtual labs Callaghan et al 2006 Zutin et al 2008 S nchez 2000 Salzmann et al 2000 In particular when a student is conducting an experiment in a 40 3 The Expe
211. pache Foundation Website http commons apache org logging Apache Pool 2009 Apache Foundation Website http commons apache org pool Apache Tiles 2009 Apache Foundation Website http struts apache org 1 x struts tiles 182 9 Bibliography Apache Validator 2009 Apache Foundation Website http commons apache org validator Aranda J Dormido S Morilla F Ruip rez P amp S nchez J 1998 Virtual Control Laboratory for Distance Learning in Proceedings of the 2nd II ITAP Workshop on Distance Learning Conception and Exploitation of the Virtual Laboratory in the Framework of the Virtual Campus Academical and Indus trial Vision WESIC 98 World Congress Girona Spain Astrom K J 2006 Challenges in Control Education in Proceedings of the 7th IFAC Symposium on Advances in Control Education ACE Madrid Spain str m K J amp Hagglund T 2006 Advanced PID Control ISA AutomatL bs 2009 AutomatL bs Website http lab dia uned es automatlab Blume P A 2007 The LabVIEW Style Book Pearson Education Prentice Hall Bourne J Harris D amp Mayadas F 2005 On line Engineering Education Learning Anywhere Anytime International Journal of Engineering Education 94 1 131 146 Bristol E H 1966 On a new measure of interactions for multivariable process control IEEE Trans on Automatic Control 11 133 134 Brito N Ribe
212. pplication When using Ejs to create the remote interface for an existing virtual instru ment there is really no need to specify a model The behaviour of the process is determined by the real equipment Variables of the model are meant to help communicate with the VI controls and indicators The panel indicating variables in Ejs was edited to enable authors to enter the URL of a VI published by the JiL Server Ejs then automatically connects with the VI and retrieves the list of controls and indicators and then offers it to the author so that s he can choose which of them s he wants to read and write by linking them to the variables declared in the Ejs table of variables This process was automatized using an interesting feature of the command based architecture Metadata information is shared between client and server as a previous step to the data exchange By doing this client and server are informed of the exchange format beforehand i e which variables will be written 12 MetadataControls 9 0 Variable de m1 3 The Experimentation Layer Variable Variable Variable Variable nen J I m119 DE 19 cn 9o T com Figure 3 30 Metadata information in LabVIEW for VI controls in Figure 3 29 and read and know what their data types and order are This means the task of packaging and decoding one variable is simple and known by both ends server and client in the first step of the communication Figure 3 30 shows t
213. pulating pump 1 Q1 and pump 2 Q5 respectively The control algorithm in each loop is a standard PID controller see Figure 4 10 with set point weighting anti windup mechanism and filtering of the derivative action str m amp H gglund 2006 Proportional CG da e kp fi ui Ti s x KF usat Integral Saturation kp Td s ud Td N s 1 Antiwindup Derivative Figure 4 10 Continuous PID controller Simulink model Figure 4 11 shows the discrete version of the PID controller programmed in LabVIEW Figure 4 11 a depicts the subVI developed for the local control of the three tank system and Figure 4 11 b shows the internal implementation of the algorithm programmed using a Formula Node structure of LabVIEW A key issue to take into account when creating the local control application is how to stop the VI in a safe way This is a critical point since this expensive piece of equipment will be remotely manipulated by students and therefore safety actions should be designed to avoid damages in the hardware For this reason actions taken should fulfill certain requirements as mentioned in Salzmann amp Gillet 2008 Some of these requirements are The Formula Node is a text based node available in LabVIEW to perform mathematical operations in a block diagram with a C like language Formula Nodes prove useful to write complex equations and use existing text based code 4 2 Protot
214. r part of this window shows which user is connected Finally Figure 6 8 b shows a tracking report of con nections disconnections internal messages of the server and the absolute path of the VI target I JiL Server 1 0 File Options Help m Is there any user connected Start Authentication Time Slot Bookings id username email password starttime endtime valid 1 rvargas hyvargas bec uned es cmVuYTEzNzY 01 01 2009 10 00 00 16 07 2009 22 49 00 true 185 demo hyargas bec uned es ZGVtbw 01 01 2009 10 00 00 01 07 2009 10 00 00 true Tr Current client user demo iL Server 1 0 File Options Help m starttime 01 01 2009 10 00 00 endtime 01 07 2009 10 00 00 a Bookings visualization Is there any user connected Start Messages Area Connection report clear Internal Messages A 128 178 5 154 3222 disconnected at 01 04 2009 15 40 30 Server Stopped at 02 04 2009 16 08 27 Starting Server at 02 04 2009 16 08 28 Server Stopped at 02 04 2009 16 08 32 Starting Server at 02 04 2009 16 08 59 128 178 5 154 1539 connected at 02 04 2009 17 12 08 Openned reference of motor2Real vi at 02 04 2009 17 12 08 128 178 5 154 1539 disconnected at 02 04 2009 17 13 43 128 178 5 154 1550 connected at 03 04 2009 9 33 17 Openned reference of motor2Real vi at 03 04 2009 9 33 17 w Application abs
215. r s actions Servlet A AA Remote users Controller b Redirect 8 http answer sad DATABASE 8 views JSP mum um um um Views Figure 5 2 Struts framework Flow of information request answer for instance a query to a database the servlet controller links with the model of the application to get the result Then once the request has been executed a jsp view is generated based on the information previously obtained Finally the view is sent to the client who visualizes the answer thus closing the request response loop Struts integrates other developments making it one of the better known and used frameworks to build web applications Apache Validator 2009 Apache Logging 2009 Apache Pool 2009 Apache Tiles 2009 5 2 3 eMersion abstractions and concepts We have already mentioned that eMersion is a web application that focuses on the teaching and learning of students in subjects with a high tech content It specifically supports hands on activities carried out with students through the Internet in a flexible context In this sense the eMersion architecture involves some abstractions and concepts which can be found in a university from an orga nizational point of view The structure of the environment suggests a first global view of the system Figure 5 3 depicts this first level of abstraction as a starting point of the eMersion architecture Concepts presented can be summarized in the following three points
216. raditional classroom provides numerous benefits Sharing resources and knowledge is also another advantage in a society each time more globalized in need of more flexible mechanisms to interact and collaborate In this sense the evolution of Internet has changed the education landscape dras tically Riva 2001 Bourne et al 2005 Rosen 2007 Anderson 2007 Hu amp Bao 2008 What was once considered Distance Education is now called On line Education whereby the method of teaching and learning is based on the use of the Internet at any given time or place to complete educational activities A specific example of application of this new teaching model is the National University for Distance Education UNED Compared to other Spanish univer sities this institution has the largest number of students due mainly to the fact that distance education allows to enrol students who can not use the traditional educational system but wish to obtain a degree or improve their professional skills without having to change their lifestyle It is currently possible to find a wide set of universities with on line presence around the world OUA 2009 London External 2009 UDIMA 2009 OU 2009 UNED 2009 Open UA 2009 UNAD 2009 UOC 2009 This confirms the viability and the importance of computer assisted teaching and learning through Internet The implementation of the distance learning model is not an easy task in the case of engineering and physi
217. raining programs classroom and online events e learning programs and training content 2 Environment Global Architecture Guaranteeing a sequence of activities that students must carry out during an experimental session Tasks can be of two types 1 tasks in simulation mode and 2 tasks in remote mode The former are the tasks that students must carry out prior to performing the experiments in the real plant This should be done with a graphical user interface allowing students to work in a simulation mode environment the objective being to get an adequate previous insight about the process involved Students would then reduce time spent on activities if using the real plant Remote access should not be allowed if the student has not satisfactorily completed previous tasks set in simulation mode Should the student s work be evaluated positively by the teaching staff access in remote mode would then be granted Managing students and the assessment of their work as well as the uploading of reports and their tracking Including an automatic booking system to schedule access to the labora tory s physical resources Figure 2 4 shows the structure of an e learning platform e Learning layer for publishing remote and virtual control laboratories Experimentation layer on the Internet e Learning Platform On line documentation practical guides tasks protocol etc f Time booking system to use the physical resourc
218. raphical interface In simulation mode the state of the system and its associated variables are updated based on the evolution of a mathematical model of the process Otherwise in remote mode these variables are updated according to the real plant changes in the remote lo cation Video feedback should be included in order to provide distant users with a certain sense of presence in the laboratory 2 2 A systematic two layer approach 21 Events sheduling to program faults in the system should be scheduled The systems could be enabled to analyze systems in presence of noise or distur bances measurements The robustness of the system could then be evalu ated under these anomalous operating situations Finally it is highly recommended that users define experiments in an easy way For example a programmed change of the setpoint could be required to observe the process response in different operating points 2 2 1 2 The server Requeriments and specifications The server side applications are running on a computer located close to the phys ical system equipped with the needed hardware software interfaces to access sen sors and actuators from the real plant Additionaly the server should implement the interfaces protocols and programming modules needed to connect with the outside world i e with the clients see Figure 2 3 Lab Server z Exchange data module client 1 C wee 7 Access management module Re
219. re A 13 At 60 sec the position reference is changed to 200 to observe the dynamic behaviour of the system The step response presents an under damping characteristic with an overshoot of around 21 The steady state is reached in almost 8 sec introducing a second step change at 70 sec Similarly the steady state in this last test is reached around 8 sec It is also possible to appreciate that in the steady state the motor oscillates slightly around the setpoint Motors usually present a dead zone due A Modelling and Control angle sim ref pos angle real position degrees volts 140 62 64 66 68 70 72 74 76 78 time seconds control signal sim control signal real 60 Figure A 13 62 64 66 68 time seconds Tracking reference in position control DC motor to the friction of the mechanism The dead zone is the input signal range to the motor where no action occurs in the output The dead zone for our didactical setup under study is about 0 3 volts 160 T T T T vel sim g 40r o ref vel 5 120 E vel real 2 gt 100 2 3 80 o o 9 60 40 j 1 L 1 1 1 L 1 L 58 60 62 64 66 68 70 72 74 76 78 time seconds 10 T T T T T 8 control signal sim control signal real eL 4 2 o Enn uu gt ab 4 2 J 0 1 1 i 1 1 1 1 f 1 58 60 62 64 66 68 70 72 74 76 78 time seconds F
220. reafter 7 4 Aremote laboratory integrated into eMersion Figure 7 5 shows a complete view of the experimentation environment eMer sion during a practical session with the three tank system in remote mode As mentioned in Section 5 2 4 the environment is composed of five independent web applications navigation bar eJournal experimentation console on line informa tion and external applications The navigation bar gives access to the other environment s web resources From the link Access Protocol users can obtain a complete user s guide of the environment The eJournal resource provides a shared workspace for users to communicate and collaborate during the learning process The eJournal allows students to save retrieve and interchange their experimental results and documents Fur thermore the presentation of results and discussing with teaching staff can be done using the options provided for these purposes They can also organize the information collected during experimental sessions as well as an on line reposi tory works Work tracking and awareness could be implemented based on this information A collection of HTML pages accessible from the same environment permits students to visualize all the documentation about the practical exercise More over each web page includes a PDF document which contains the same informa tion than the web page This option can be used by students to download the experiments as electronic f
221. receives indicators from JiL Ejs requests to stop the VI target to JiL JiL stops the VI target and requests to stop the VI gt JiL stops the VI targ Ejs closes the connection Ejs requests to close the VI target and the connection to JiL JiL closes the VI target in a safety way and closes the connection E Edition of simulation and execution ay a Continuous exchange of data e S Stop target VI and close connection Figure 3 31 Messages exchange between Ejs and JiL server All this flow of information between client and server is hidden to the developer and s he is only limited a few and simple steps Figures 3 32 and 3 33 illustrate how it applies to Ejs for sample VI in Figure 3 29 JDOM is an open source Java based document object model for XML that was designed specifically for the Java platform so that it can take advantage of its language features 74 3 The Experimentation Layer Note the External File field in Figure 3 32 with the URL of the sample VI file The syntax is lt labview IP_address port gt VI_relative_path ED Easy Java Simulations JILTest xml DER O Introduction Model View Variables Initialization Evolution Constraints Custom 4 E B LabVIEW Commons External File labview localhost 2055 jiltest JILTest vi Name Value Dimension Connected to intcon int int control con intind j int indicat
222. resented as an appropriate solution for the on line teaching of en gineering students T his paradigm can be applied taking into account the three following perspectives Pedagogical From this perspective flexible education means that students are given extended access to learning resources increased freedom to orga nize their learning activities and enhanced participation autonomy and collaboration 5 1 On line learning Pedagogical aspects 115 Technological From this perspective flexible education enables adequate exploitation of information and communication devices as well as infraestruc tures especially the Internet Organizational From this point of view flexible education relies on renewed study programs regulations as well as partnerships and collaborations with other institutions These three points of view on the flexible education paradigm have already been addressed throughout the analysis design and development of the present thesis In fact the UNED being a distance learning university the application of such teaching alternative modalities is the main goal for this academic institution and therefore big efforts are always made to pursue such objective Pedagogical aspects should be addressed first since the quality of teaching must be maintained when providing remote experimentation services in an open academic environment As there are different learning models available Maharg 2004
223. ril Text file lt Terminado Figure C 7 Fields description of the fragments table On the other hand by clicking on a fragment users can select open or save the fragment this action depends on the browser and type of fragment Some basic fragment types are described below a External Fragment Fragments imported to the eJournal from the user local hard disk for example reports in format doc or pdf a Image Fragments of image type obtained from experimentation con sole The format by default is gif Result Data registers obtained from a simulated or real experimental session An example would be the registers m containing measure ments The fields of the fragment container table provide information about each one of them a Name Name of the fragment Author Owner of the fragment a Task Shows whether fragment has been assigned to a task The action to assign a fragment to a task is done by clicking on this simbol Status Helps to visualize the state of a fragment a Creation Date and time of creating a fragment C 1 How to work with eMersion 245 Annotation Allows users to associate notes to fragments Each an notation is added as a attribute of a fragment By this mechanism students can generate sets of annotations associated to particular frag ments For example the creator of the fragment can link an annota tion to the fragment and make a question about the obt
224. rimentation Layer virtual manner s he works with an mathematical model of a process while in a remote working session relies on access to a real resource through the Internet by means of a server computer acting like a gateway Looking at Figure 3 8 we can appreciate what happens when a client appli cation keeps a connection with a remote server to control a target device The server side sends in a continuous flow the information blocks s k representating the current state of the plant and it receives information blocks c k containing changes carried out by a remote user in some of the system input parameters On the other hand the client side receives the state of the system sent by the server contained in s k blocks and at the same time keeps waiting for some user s interaction in order to report this change to the server side by means of new information blocks c k KO user s interaction actions user s perception states sk DU 0 s3 s2 si Figure 3 8 Stream of information between client and server Different kinds of data streams can be exchanged between client and server see Figure 3 9 Usually the block s k is composed of measurements that repre sent the current state of the system a few bytes and video images kilobytes to provide visual feedback to the remote user Here it is important underline that inclusion of visual feedback is fundamental in the remote handling of a system
225. rol strategies have been developed Manual controller This operating mode allows studying the system re sponse in open loop without control actions Simply the changes carried out by users will be directly applied on the system 210 A Modelling and Control PI controller A PI controller is enough to control the engine speed Notice that although the designed controller is a PI the full PID control algorithm has been programmed to test the effect of a derivative action as well The PI controller is as follows 1 C K t A 21 9 Kp 1 A 21 where K is the proportional gain of the controller T is the integral time of the controller Finally the closed loop control structure can be seen in Figure A 10 DC Motor reference e U A W gt Kpl BE gt Tis T8 1 Figure A 10 PI control for speed control on the DC Motor As known if controller parameters are well tuned the system is stable and non steady state error occurs In particular the roots of the system in closed loop characteristic polynomial see the equation A 22 1 Ak AK gp P T TL 0 A 22 are responsible of the oscilatory behaviour of the tachymeter output under step changes in the input voltage Thus a simple pole placement controller design for first order systems can be applied With Kp 0 08 T 0 6 for 7 0 253 sec and A 27 sec vol
226. ropean Control Conference ECC 2009 Budapest Hungr a Agosto 2009 6 H Vargas J S nchez y S Dormido 2008 Proyecto AutomatLabs Red Interuniversitaria de Laboratorios de Control Autom tico a trav s de In ternet XIII Latin American Congress on Automatic Control CLCA M rida Venezuela November 2008 7 H Vargas J S nchez S Dormido Acceso Extendido a Recursos de Ex perimentaci n a trav s de Internet La Experiencia AutomatLQbs XXIX Jornadas de Autom tica Tarragona Espa a Septiembre 2008 8 G Farias S Dormido F Esquembre H Vargas S Dormido Canto 2008 Laboratorio Virtual para la Ense anza de T cnicas de Reconocimiento de 1 4 Publications Awards and Projects 13 10 11 12 13 14 Patrones XIII Latin American Congress on Automatic Control CLCA M rida Venezuela November 2008 S Dormido J S nchez H Vargas S Dormido Canto R Dormido N Duro G Farias M A Canto y F Esquembre 2007 An lisis desa rrollo y publicaci n de laboratorios virtuales y remotos para la ense anza de la autom tica II Congreso Espanol de Inform tica CEDI Zaragoza Spain J L Guzm n M Berenguel F Rodr guez H Vargas J S nchez S Dormido 2007 Desarrollo de un entorno de experimentaci n basado en web para estudios de ingenier a un caso pr ctico II Congreso Espa ol de Inform tica CEDI Zaragoza Spain J S
227. rs General Parameters TCP Port 2055 2 Open Front Panel Smtp server smtp uned es Package Size 1 C Start JIL at launch Sender s email adminlab dia uned es Sending Rate ms 100 Recipients l Ce Authentication Parameters Y Use AutomatL bs Authentication Data Source Name motor Updating Delay sec 20 Check DB Connection to DB P b Edit General Options disabled when started Figure 6 7 Setting up options in JiL server The Edit General Options window was modified to configure authentication parameters in the JiL server Figure 6 7 a shows the new settings added to this configuration zone highligthed with a red elipse Once the database manage ment system MySQL has been installed and the tables created a new data source should be created through an ADO connection With this setting the JiL server is able to connect to the database A user only needs to provide the DSN 6 2 A simple point to point authentication protocol 139 reference Data Source Name and check the connection to the database pushing the Check DB button If the connection is successful an indicator led labelled Connection to DB is turned on When the server is started all the configuration options are disabled see Figure 6 7 b Other two visualization windows Authentication and Output Messages ar eas can pop up from the Options menu The table Time Slot Bookings in Figure 6 8 a shows the new incoming bookings the lowe
228. rument o e 65 Part of the JiL Server wiring diagram that obtains the list of a VI controls and indicators using two Invoke Node blocks 65 Modular view of the Jib Server approach 68 Scheme of a generic VI for feedback control purposes 69 Wiring diagram of the example o 69 Metadata information in LabVIEW for VI controls in Figure 3 29 72 Messages exchange between Ejs and JiL server 73 Table of Ejs variables linked to VI controls and indicators 74 List of controls and indicators of a remote VI as offered by Ejs 74 The D TS200 three tank system by Amira 79 The three tank system From left to right tanks 71 75 and T3 are serially connected to each other by pipes 80 Ejs Model of the three tank virtual lab Variables 81 Ejs Model of the three tank virtual lab Evolution 82 Ejs Model of the three tank virtual lab Custom methods 83 Ejs View of the three tank system virtual lab 83 GUI of the three tank system virtual lab 84 Local control of the three tank real system Block diagram 86 List of Figures 4 9 4 10 4 11 4 12 4 13 4 14 4 15 4 16 4 17 4 18 4 19 4 20 4 21 4 22 4 23 4 24 4 25 4 26 4 27 4 28 4 29 4 30 4 31 4 32 xi Local control of the three tank real system Front panel 87 Continuous
229. ry the influence of the electromechanical pieces to measure and acting on the motor are taken into account see Figure A 7 U s A Wes Uw s l Dep D A Amplifier Motor Tachymeter A D1 Th s Ute conver Integrator Position sensor gain A D2 ola Figure A 7 Model of the DC Motor A 2 The DC Motor Modelling and control 209 where both the gain A and the time constant 7 of the engine can be calculated through the open loop step response of the system in speed K and Kp are the tachymeter and the position sensor constants respectively A 2 2 Design of the PID controllers Speed control The DC motor is commonly used in processes where an angular speed is required for example in lamination processes Figure A 8 shows the schematic view of a typical speed control system The power amplifier receives a direct voltage U that excites an induced magnetic field The output U s from the tachymeter is a voltage proportional to the angular speed reference C e Controller Motor Power Amplifier U s Tachymeter Figure A 8 Typical speed control system Figure A 9 illustrates the closed loop control sheme for this process where C represents the controller used to close the loop DC Motor reference f e gt C Uy G s W gt Figure A 9 Control strategy for the DC Motor Two cont
230. s 2009a Ejs wiki Website http www um es fem EjsWiki Ejs 2009b Free online course about Ejs Website http www euclides dia uned es simulab pfp index_en htm eMersion 2009 EPFL Website http emersion epfl ch Esquembre F 2005 Creaci n de Simulaciones Interactivas en Java Pearson Prentice Hall Esquembre F Gillet D Salzmann C Rekik Y amp Dormido S 2007 Comprehensive Collaborative Web based Experimentation Integration of Dis tributed Teleoperation and Simulation Services for Supporting Active Learning in Higher Education http e spacio uned es fez view php pid bibliuned 715 Fan R Cheded L amp Toker O 2005 Internet based SCADA a new approach using Java and XML Computing amp Control Engineering Journal 16 5 22 26 Farias G Keyser R D Dormido S amp Esquembre F 2009 Building Remote Labs Using Easy Java Simulations and Matlab in Proceedings of the 10th European Control Conference Budapest Hungary Ferreira J amp Cardoso A 2005 A Moodle extension to book online labs International Journal of Online Engineering 1 2 1 4 Fleming W H 1989 Future directions in control theory A mathematical perspective Society for Industrial and Applied Mathematics Garcia J Lopez D amp Ordu a P 2008 Mobile Devices and Remote Labs in Engineering Education Eighth IEEE International Conference on Advanced
231. s in the Constrains panel Graphical representations of the results model view communication Interaction of the user view model communication Execution of the actions in the Evolution panel Execution of the user s action Figure 3 2 Simulation algorithm of any Ejs application To illustrate the ideas aforementioned a simple virtual control lab of the single tank process created by Ejs is introduced hereafter The structure followed in the creation of this simple but very descriptive example will be extrapolated to the development of every study case presented in Chapter 4 A simple virtual control lab in Ejs The single tank process As mentioned in Section 3 1 1 detailed knowledge of the underlying analytical model of the process is a mandatory step before starting with Ejs The creation of the virtual lab will therefore be addressed after the model description Model of the Process dh a K SUL ca Ul at d Aaya Variables and Parameters h Liquid level inside the tank m a Cross section of the outlet hole m A Cross section of the tank m g Gravitational acceleration 9 8 m s u Voltage applied to the pump volt K Pump parameter m volt Ref Level Controller Container Figure 3 3 Control of a first order system The single tank process 36 3 The Experimentation Layer Figure 3 3 shows the scheme of the single tank process a
232. s is to generate a VI to control the real system locally Figures 4 8 and 4 9 show the block diagram and the front panel of the LabVIEW application created according to the structure described in Section 3 3 In particular the block diagram in Figure 4 8 shows the flat sequence structure of LabVIEW that divides the execution flow of the application into three stages 1 Initialization of the hardware 2 Acquisition and closed loop control HIL Hardware in the Loop 3 Reset of the hardware Figure 4 8 Local control of the three tank real system Block diagram THREE TANK SYSTEM REAL 8 8 8 h2 h3 29 9023 19 9414 25 1172 hi Figure 4 9 Local control of the three tank real system Front panel 88 4 Prototypes Developed The first and third stages are self explanatory However the intermediate stage contains some technical issues described below The LabVIEW application implements a descentralized multivariable control strategy see Section A 1 2 in Appendix A This kind of control strategy can be applied to systems for which the interaction between variables is minimal and therefore every measured vari able can be controlled by means of one manipulated variable in independent control loops In this case the liquid level in tank 1 Hi and tank 2 H3 is regulated by mani
233. s the on line learning of students in control engineering education will be described 114 5 The e Learning Layer 5 1 On line learning Pedagogical aspects The potential of web based experimental applications used as pedagogical sup port tools in control engineering education has already been described in previous chapters By using these tools engineering students can be required to complete different tasks and or activities during experimentation sessions However these activities must be carefully designed so that they avoid inherent drawbacks that exist when carrying out distant experiments In this particular case students must conduct their practical assignments without the help of an on site in structor Therefore additional tools are required for virtual and remote control laboratories so as to Provide all the neccesary documentation to carry out a complete remote practical session Provide a common workspace where instructors and students or students only can interact synchronously or asynchronously in a similar way than in traditional labs To give the possibility to create workgroups where users can collaborate and interact among them such as they would do it in a face to face class To address this problem complementary web based resources should be ex plored for a convenient use of web based labs As mentioned in the introduction of this dissertation the flexible learning paradigm is p
234. s two examples of this feature Figure 5 11 a shows the external toolkit so called SysQuake Remote which was integrated into eMersion at EPFL to sustain interaction during the learning http 62 204 199 141 HTML Test Page Mozilla Firefox ius Make a booking alala Frequency Domain Identification Em Frequency Domain Identification Select date and plant ido h e e Pl rar ana T ieu 03 Sun Mon Tue Wed u ri B tae gain i DE 0 2 4 5 NB ve s mt A n Time est y 08 16 18 Y CIA Time et pg E 20 Que y check Time est 7 0 F Select time Ae EA EE EEEE OT t T wta Sample 57 e 18 00 19 00 arg G time Model first order 00 ax 0 e T em 19 00 20 00 20 Excitation V Oclosed 00 63 00 00 08 00 9 14 00 15 00 e 20 00 21 00 Data P Um a o w B 00 100 e 15 00 16 00 lt gt 21 00 22 00 9 cc Graphics small C large coms ma Cat o pzz Revert G u LE Gem uem e e e 17 00 16 00 e 23 00 00 00 PF cay L en Available Selected 7 he parkis med J ee YP 0309 Sys Wak Ste v em Occupied Not available a SysQuake Remote from EPFL b Bookings System from UNED Figure 5 11 Examples of external web applications integrated into eMersion 128 5 The e Learning Layer process Gillet et al 2005 Figure 5 11 b depicts the automatic bookings system developed in this dissertation to manage students acc
235. sary to include a module that allows access to authorized users only These systems can be of two kinds either traditional or on demand S nchez 2000 The first class is of a similar nature to the one that can be found in con ventional laboratory in a presential context Timetables are available and each user must access the physical resource needed in the timeslot that the system administrator has assigned her him In the second class when a user wishes to access a piece of equipment in the laboratory s he must book a free timeslot using web applications provided by this service Since certain users need more time than others to complete their tasks the system should be flexible enough However a minimum degree of inteligence is required for the system to avoid abuses from users who will book more timeslots than needed and guarantee a minimum number of hours for everyone This piece of work will describe a first access protocol to the laboratory s phys ical resources For this purpose a client application is authenticated against the remote server computer directly connected to the plant and which also contains an authorized users list kept in a database Such a list is filled out manually by an administrator usually the teacher or a tutor with these privileges enabling access to students who fullfil certain requirements Finally once the access protocol has been defined the inclusion of an au tomatic mechanism to populate the datab
236. second Indep Var t Increment lat e Cus State Rate 20 d level pa gr ode level 10 5 MIN FPS 25 sotver Euler y Tolerance Events 0 d Dynamics of the process ODE O Introduction Model View O Variables Initialization Evolution O Constraints Custom Equations pee Dynamics Control persecond i automaticMode P I action LAM u Kp setPoint level integral 20 if u 0 u 0 15 Update integral action integral integral Kp dt Ti setPoint level 10 n 5 MIN FPS 25 public double ode double level if level gt 0 return a A Math sqrt 2 g level Kflow A u else return Kflow A u Comment Mass balance and Bernouilli s law for the single tank Additional Libraries e Controller f The ODE as a custom method Figure 3 5 Defining the MODEL of the single tank process In Figure 3 5 a a brief description of the simulation and complementary in formation about the system can be included js takes this information and uses it to create HTML pages with the description on the Java applet generated Figures 3 5 b and 3 5 c show respectively the declaration of the physical vari ables associated to the tank and the variables related to the PI controller The evolution of the system is defined in Figures 3 5 d and 3 5 e in two pages The first one is used to specify the mathematical model Figure 3 3 and
237. sertar Formato Ayuda Del amp l 4 e gt Data from console in remote mode Date Wed Apr 22 16 26 15 CEST 2009 UNED Remote Lab The DC motor 2 data format t th w u ref dat 50 90 44 D 52 54 56 58 60 62 64 66 68 70 51 00 43 D seconds 51 10 s 44 D 51 20 i 43 D U P I D 51 30 A 46 D 10 51 40 A 44 51 50 E 44 D A 5 51 60 A 48 51 70 44 2 51 80 E 44 D 3 0 51 90 A 45 52 00 42 0 5 52 10 t 45 52 20 44 D 10 NR 1 36 dk 52 54 56 58 62 64 66 70 Para obtener Ayuda presione F1 Seconds a Generic text data file m b Image data file gif Figure 7 3 Types of data fragments in the eJournal space Special care was given to ensure the text data file has a Matlab compatible structure This means students can load and analyze the data in the Matlab workspace without extra manipulation Furthermore a header as comment has been added at the top of the file This header provides information about whether the data was saved in simulation or remote mode the date when this was done the remote equipment used and what the name and order of the variables are 3 Linking all Work modules in eMersion As mentioned in Section 7 1 the management of work modules in eMersion is closely related to the linkage of all web components of the environment mod 154 7 Integration of Layers ule is part of a course in eMersion and a course can contain several
238. simulation of any physical system developers must model for instance through a set of differencial equations the system and then solve the equations by numerical integration methods Software tools such as Matlab Simulink Matlab 2009 SysQuake SysQuake 2009 Mod elica Modelica 2009 and others help build virtual laboratories as they provide predefined built in numerical methods which can be used to solve the models By means of them experts in different subjects can generate simulations with a certain degree of complexity to be used by teachers in classrooms with the goal of making the class more attractive and useful for their students Kumpaty amp Haeg 2007 Guzm n et al 2008 Urqu a et al 2005 Mart n 2007 However most applications conceived with these software tools can only be used in local mode i e the core of the simulation resides in the same machine where it is executed and therefore these tools are not prepared to be distributed on line through a communication network such as the Internet The natural and intuitive way to face this problem is to look for a technology that covers all aspects related to the use of simulations in local mode that could be distributed on line and take advantage of the web based simulation technologies available In relation to this point Java technology a programming language created by Sun Microsystems has gained enormous popularity since its launch Some of the advantages of using Ja
239. ssemble control parameters input variables of the system Send the control parameters to the server side Task 3 Render Display the information renderized and detect the user interaction Figure 3 16 Concurrent tasks in the client side The virtual lab of the single tank process is used as an example again to explain the aforementioned ideas Measurements by themselves can faithfully represent the dynamic behaviour of the remote equipment and therefore the content of this stream will be fundamental to display results on the client side Thus in order to simplify the explanation provided visual feedback is not taken into account here Section 3 2 5 is completely devoted to explaining how this part of the development has been addressed The first step is to define the format of the exchanged variables in accordance with the server Figure 3 17 shows one possible format for the data t h qa a system state m a am Kp Ti Td ref b control parameters Figure 3 17 Structure of exchanged data packets between client and server 92 3 The Experimentation Layer The current time t liquid level inside the tank h and the input flow to the tank in automatic mode qa constitute the states of the single tank system measurements The control vector is divided into the control mode m a the input flow to the tank in manual mode qm the PID parameters Kp T and T4 and finally the setpoint val
240. ssing button Top Tank1 BB three tanks system DER eJournal Control Language H1 H2 H3 Setpoint L 200 300 seconds Pump1 Pump2 Status video No video in simulation mode _ Augmented PID VIDEO rs A13 50 A32 50 A2 50 Aleak1 0 Aleak2 0 Aleak3 0 e Qa Sus Pump1 19 1 Pump 45 8 Fps 20 Q1 cm3 sec 19 08 Q2 cm3 sec 458 100 200 300 400 seconds H1 mm H2 mm 199 67 A mm 5 SP1 mm 300 SP2 mm 200 Play Pause IC Reset JC Connect Time sec 450 H3 mm 24 Status Lab Simulation mode a no connected 4 a Final GUI in simulation mode three tanks system eJournal Control Language a H1 H2 H3 Setpoint Y L 300 seconds Pump1 Pump2 T T Status video Reading video from camera O Augmented contor PD VIDEO A13 50 A32 50 A2 50 L 100 200 300 seconds H1 mm 298 69 H2 mm Aleak1 0 Aleak2 0 Aleak3 0 Pump1 18 31 Pump 44 56 Fps 25 Q1 cm3 sec 18 31 Q2 cm3 sec 44 56 SP1 mm 300 SP2 mm 200 Play Pause lese Time sec 450 3 H3 mm 248 41 TIME REMAINING min 19 Status Lab Remote mode with labserver E connected lt i b Remote mode with video feedback Figure 4 17 The three tank system virtual and remote control laboratory 4 2 Prototype I The Three tank system 95 Finally Figure
241. stance Education Research Trends Nova Science Publisher pp 131 167 7 S Dormido H Vargas J S nchez R Dormido N Duro S Dormido Canto F Morilla M A Canto G Farias 2009 Compartiendo recur sos de experimentaci n a trav s de Internet la experiencia Automatl bs Book Chapter La UNED ante el EEES Redes de investigaci n en inno vaci n docente 2006 2007 Editorial UNED 1 Introduction Objectives and Structure Conference papers 1 H Vargas Ch Salzmann D Gillet S Dormido Remote Experimenta tion Mashup 8th IFAC Symposium on Advances in Control Education Kumamoto Jap n Octubre 2009 2 M Guinaldo H Vargas J S nchez E Sanz S Dormido Web based Control Laboratory The Ball and Beam System 8th IFAC Symposium on Advances in Control Education Kumamoto Jap n Octubre 2009 3 L Diaz G Ramos H Vargas R Costa A Virtual Remote Laboratory to illustrate the Internal Model Principle for periodical signals 8th IFAC Symposium on Advances in Control Education Kumamoto Jap n Oc tubre 2009 4 H Vargas J S nchez y S Dormido Experimentaci n Remota en Au tom tica Nuevos Entornos Basados en la Web 2 0 XXX Jornadas de Autom tica Valladolid Espa a Septiembre 2009 5 H Vargas J S nchez y S Dormido The Spanish university network of web based laboratories for control engineering education The AutomatL bs project Eu
242. stem in virtual mode 102 4 Prototypes Developed Three small vertical rectangles placed along the 3D representation simulate the sensors and the color inside the duct 3D cube covering the sensors represents the air temperature A WebCamImage element displaying the output of a web cam see Figure 4 25 a was also added to the view Since the air heating process is not visible to the human eye the video image is enhanced with a coloured 3D image representing the air temperature BB heatflow system e Journal Control Language T1 T2 T3 Setpoint T T s i L 1 20 25 30 35 40 45 seconds U P 1 D Status video Reading video from camera O Augmented CONTROL SENSOR PID Parameters 20 25 30 35 40 45 sa 1 seconds Kp C volt 0 1 J 90 5322 oC Ti sec 2 Temp S1 C Setpoint S1 C Td sec 0 Temp S2 C 46 64 U Heater volt 3 63 Temp S3 9C 45 49 Time sec 45 1 Status Lab Remote mode with labserver a connected TIME REMAINING min 10876 a Remote mode with video feedback BB heatflow system e Journal Control Language HEN T1 T2 T3 Setpoint i i 30 35 seconds U PD Status video Reading video from camera e Augmented CONTROL SENSOR PD Choice a sensor to close loop 30 35 40 45 seconds Close loop around sensor 1 901215373 i eC O Close loop around sensor 2 emp 319 Sp RA O Close loop around sensor 3 Temp 52
243. sters can be obtained from the Ejs applications 7 2 1 Ejs built in methods to save data in eJournal Client side Java applications can easily be integrated into eMersion with the help of predefined built in methods of Ejs Esquembre et al 2007 These meth ods were used in this dissertation to save data fragments in the eJournal space Methods used and their functionalities are described below public boolean _saveImage String fileName String compName Saves the image of an element compName of the view to the file fileName public boolean _saveText String fileName String textData Saves the text tertData to the file fileName Code for SaveGraph Action Saving data as an image file If the application is loaded as applet from eMersion then image file is saved into eJournal Otherwise it is Eje Ejs motor4 xml E 7 7 7 saved in the local hard disk O Description velocityControl _savelmage GraphSpeed VEL jelse if positionControl _savelmage GraphPos P0 Tree of Elements E Simulation View MainFrame y Y MenuBar f Ok Cancel Ex eJournal n SaveGraph m m Code for SaveRegister Action SaveRegister sbuilderSave sbuilder e Be Control i Y Saving data as a text file 4 If the application is loaded as applet from eMersion gt Ele Language then data is saved
244. studio XXVII Jornadas de Autom tica Almer a Spain J S nchez H Vargas S Dormido 2006 Recursos de Aprendizaje basa dos en el Web para la formaci n ocupacional en sistemas de Regulaci n y Control XXVII Jornadas de Autom tica Almer a Spain H Vargas G Far as S Dormido and J S nchez 2006 Web based learning resources for automation technicians vocational training illustred with a Heatflow and liquid level laboratory 7th IFAC Symposium on Advances in Control Education ACE Madrid Spain N Duro R Dormido H Vargas S Dormido J S nchez and R Pastor 2005 The Three Tank System A Remote and Virtual Control Labora tory using Easy Java Simulations 44th IEEE Conference on Decision and Control and European Control Conference CDC ECC Sevilla Spain N Duro R Dormido H Vargas S Dormido J S nchez 2005 El sis tema de tres tanques Un laboratorio virtual y remoto usando Easy Java Simulations I Spanish Congreso on Computer Sciences CEDI EIWISA Granada Spain Awards PRODEL prize to the best paper about Control Education presented in the XXIX Jornadas de Autom tica Tarragona Spain September 2008 Title of paper Acceso Extendido a Recursos de Experimentaci n a Trav s de Internet La Experiencia AutomatL bs Prize from Consejo Social y Fundaci n UNED to the best project on teaching innovation 2008 Madrid Spain Dece
245. sue on Control Engineering Education 21 1122 1133 Dormido S amp Esquembre F 2003 The quadruple tank process an interac tive tool for control education n Proceedings of the 7th European Control Conference Cambridge UK Dormido S Esquembre F Farias G amp S nchez J 2005b Adding interactiv ity to existing Simulink models using Easy Java Simulations in Proceedings of the 44th IEEE Conference on Decision and Control and the European Control Conference Sevilla Spain Dormido S Vargas H S nchez J Dormido R Duro N Dormido Canto S amp Morilla F 2008 Developing and Implementing Virtual and Remote Labs for Control Education The UNED pilot experience in Proceedings of the 17th World Congress The International Federation of Automatic Control Seoul Korea Dost lek L amp Kabelov A 2006 Understanding TCP IP A clear and com prehensive guide to TCP IP protocols Packt Publishing Ltda Duro N Dormido R Vargas H Dormido Canto S S nchez J Farias G amp Dormido S 2008 An Integrated Virtual and Remote Control Lab The Three Tank System as a Case Study Computing in Science and Engineering 10 50 59 9 Bibliography 185 Eikaas T L Foss B A Solbjorg O K amp Bjolseth T 2006 Game based dy namic simulations supporting technical education and training International Journal of Online Engineering 2 2 Ej
246. t authentication protocol 134 Database to host bookings llle 135 Bookings registers in the accesslist table of the database 135 Global vision of the first stage of the booking process 136 LabVIEW Identity cheking module Extract of the block diagram 137 Setting up options in JiL server o 138 JiL server with authentication Setting up and status visualization 139 A flexible scheme for bookings and authentication process 141 State diagram of the booking process o o a 142 State diagram of the authentication process 143 Interfaces involved in the reservation process 145 Server site of the automatic bookings system 146 Ejs built in methods to link the applet to the eJournal workspace 151 Saving data from the Ejs console to the eJournal space 152 Types of data fragments in the eJournal space 153 Creating a new module included into a course in eMersion 154 List of Figures xiii 7 5 eMersion s facade Remote session using the three tank system 157 8 1 UNED pilot experience home page o e 162 8 2 AutomatL bs project network 164 8 3 AutomatL bs home page o e e 166 8 4 DC motor Miguel Hern ndez University 167 8 5 One tank system University of Almer a o 167 8 6 The Rotoiman system Polytechnic U
247. t be composed of three parts 1 The model which describes the process under study in terms of Variables which hold the different possible states of the process Relationships among these variables laws that govern the process expressed by computer algorithms 2 The control which defines certain actions that a user can perform on the simulation 3 The view which shows a graphical representation either realistic or schematic of the process and its states From a practical point of view Ejs helps developers follow the MVC paradigm through three main sections Introduction Model and View see Figure 3 1 The Introduction is where the developer can write any didactical or pedagogical information related to the application for example instructions about how to use the simulation exercises to complete questionnaires etc Ejs takes all this information and transforms it into HTML pages The Model covers all aspects related to the definition of the model and the control of the MVC paradigm de scribed above And finally the section View is where the graphical representation of the virtual laboratory is built using the palette of predefined ready to use com ponents of Ejs and where the level of interactivity is specified by modifying the internal properties of these elements In this sense Ejs goes one step further in the process of simplifying the build ing of a simulation since it eliminates the control e
248. t could always be used The JPEG format recovers single images from a remote camera through http requests whereas the MPEG format gets the images compressed as a raw flow of bytes 60 3 The Experimentation Layer 3 2 5 3 The Augmented Reality concept The augmented reality concept in remote experimentation tries to give an added value to the remote visualization module WebCamImage by overlaying the simulation view created using the set of graphical elements of Ejs over the video images Tree of Elements N e Simulation View other graphical elements onto the video images o fe Frame 2d DrawingPanel MATS t Camimage First element in the hierarchy in order to draw Eok 0 Square Circle M gk Tree of elements in Ejs After compilation Figure 3 22 Simple illustration of the augmented reality concept Figure 3 22 depicts the idea previously described From a practical point of view the development of this point has already been explained in Section 3 2 5 2 since its conception is strongly inspired by the WebCamImage element This component was built in such a way that it can be added onto Ejs drawing containers with coordinate axes such as DrawingPanels or PlottingPanels Consequently this effect is achieved when adding a WebCamImage element as a first object of a drawing container of Ejs since the captured images will always appear in the background behind the other graphical el
249. t experiments research on system identification techniques and control of systems with time delay could be addressed 214 A Modelling and Control A 3 1 Modelling A simple first order model To obtain a simple first order model we can apply step inputs to the heater and observe the temperature change at each sensor Figure A 15 shows one of the plots that are obtained In this case we see the temperature in the three sensors blue 5S1 green S5 red S3 with a fixed flow rate using V 3 0 Volts and three different voltages applied to the heater Step response in open loop 70 T T T T L L L L L 200 250 300 350 400 450 500 550 seconds 45 T T T T 4 lx g h S 3 55 v 3 2 5 I L 1 L L L 200 250 300 350 400 450 500 550 seconds Figure A 15 Step responses at the three sensors when V goes from 3 to 4 volts and a fan fixed flow rate voltage of 3 volts To design a feedback controller to control the temperature at a desired sensor location we need to have a model of the system Examining Figure A 15 we note that the step response can be approximated by a simple first order system of the form Ta G c m A 24 Vi Tns 1 Now suppose we want to control the temperature at 4 with a fixed flow rate using V 3 Examining Figure A 15 we note the following A 3 The Heatflow system Modelling and control 215 Applying V 4 volts to the heater res
250. taphor has been chosen at the EPFL to design the graph ical user interface GUI of the environment dedicated to web based experimentation 118 5 The e Learning Layer From a design point of view eMersion meets the majority of requirements for web based experimentation environments A complete review of the main characteristics of collaborative hands on activities in such learning environments is presented in Nguyen 2007 A summary of these essential features is provided below Hands On activities Support Virtual and remote labs should support hands on activities Such resources are paramount in engineering curricula since students need to have contact with mechanisms taught and materials used Component Integration Additional web based resources should be integrated into the same experimentation environment so as to support hands on ac tivities Experimentation by itself interaction and collaboration support as well as complementary information for experimentation process are some examples of these additional web based resources Multi session Experiment The system should allow to experiment in a flex ible way i e students could complete tasks set at anytime and from any location In addition the system should also allow several trial and error experiments to reinforce students understanding Types of Collaboration The system should place a greater learning respon sability onto the students Different coll
251. tation 32 1 Baekground ts x km s el ee RUM 3 2 1 1 Communication protocols 3 2 1 2 TCP or UDP as data transportation mechanism 3 2 2 Building the server side o o 3 2 3 Using LabVIEW for remote experimentation 3 2 4 Linking a virtual lab to the server side 3 2 5 Visual feedback from the remote plant 3 2 5 1 Development of the Java library webcam jar 3 2 5 2 The WebCamImage view element in Ejs 3 2 5 3 The Augmented Reality concept 3 3 A new approach to connect Java and LabVIEW 3 3 1 Motivations co xke mne a eel bee he 3 3 2 Technical issues of the JiL Server middleware layer 3 3 3 Details of the JiL Server implementation 3 3 4 API Java to control LabVIEW applications 3 3 5 A very simple example of use o 3 3 6 Integrating the API Java in Ejs 34 Conclusions dial eso AA ie dp elu us Prototypes Developed 4 1 Case studies A short introduction 4 2 Prototype I The Three tank system o o o 4 2 1 System overview o e e 4 2 2 The three tank system s virtual lab in Ejs 4 2 3 Local control of the three tank system in LabVIEW 4 2 4 The virtual and remote version of the laboratory 4 3 Prototype II The Heatflow system lll 4 3 1 System overview 222A 55 57 58 60 61
252. te experimentation system is to define its global architecture In this sense from a software engineering point of view there are three mandatory stages to follow in every software develop ment process Requirements Analysis Specifications and Design and Architec ture This chapter describes these stages and tries to give a systematic approach to use for the appropriate implementation of a web based experimentation envi ronment with pedagogical perspectives Once the global structure of the system is specified requirements analysis and specifications the design and architecture process is then addressed Here the global architecture of the system is divided into several sub problems or compo nents that form the entire system In this given context this chapter introduces the systematic two layer approach to follow in the development process of the system the experimentation and the e learning layers 2 Environment Global Architecture Planning the structure of the system Figure 2 1 illustrates the typical application scenario of a remote experimentation system designed for pedagogical purposes The university provides the overall infrastructure required for remote experimentation services offered to students that is a set of didactical setups specially designed for hands on laboratories a set of server computers used to interface these processes with the outside world and a main server computer providing the complementary web based
253. ted in many works Hahn amp Spong 2000 Cefalo et al 2003 Casini et al 2004 Valera et al 2005 Eikaas et al 2006 G mez amp Garc a 2007 Costas et al 2009 Brito et al 2009 In fact in the last decade several academic institutions have explored the WWW to transfer their courses and experimental activities towards distributed contexts However most of these developments focus on technical issues related to the design and building of web enabled applications that allow to perform prac tical activities through the Internet virtual and remote laboratories In general these do not take into account the social context of interaction and collaboration generated in traditional hands on laboratories Nguyen 2007 Indeed direct contact with teachers and interaction with classmates are valu able resources that may lessen or even disappear when conducting hands on ex perimental sessions via web based laboratories This suggests the necessity to include this requirement when designing developing and deploying a web based experimental environment as that described in this dissertation 1 Introduction Objectives and Structure Remote experimentation and flexible learning The flexible learning paradigm is presented as an appropriate on line learning solution for engineering students Kazmer amp Haythornthwaite 2005 In Gillet et al 2005 such paradigm is analyzed by the authors from three different per spectives pedagogical te
254. th Ejs and published by a web server This applet controls the threetankReal vi through the JiL server 3 Web server to publish the Java applet created by Ejs A web server should be installed in order to host the web site that provides the Java applet embedded in an HTML page 4 Web browser to download the Java applet from server A web browser is used to enable clients to have access to the web based laboratory 4 3 Prototype II The Heatflow system 97 In the example of the three tank system the web site can be accessed from http threetank dia uned es whereby threetank dia uned esis the domain name assigned by the DNS server that manages the root domain So far we have seen in detail how to put the methodology introduced in the previous chapter to create a web based control lab of the three tank system into practice Since these application are very specific their implementation will differ in each case The four points mentioned in Section 4 1 should provide a set of guidelines for potential developers to follow Outcomes of this work was published in Duro et al 2008 Dormido Canto et al 2008 In the following section we will see how the same mechanism is used to create the other two lab prototypes However to avoid repeating the steps previously mentioned in this case study the explanation will exclusively describe the local control VI and the final GUI of each laboratory 4 3 Prototype Il The Heatflow system
255. the tanks cross sections Q4 and Q denote the flow rates of pumps 1 and 2 Qij denotes the flow rates between tank T and T j 0 represents the system s output and Que i 1 2 or 3 represents the flows that leave the tank when its drain valve is open These three balance equations mean that the volume variation in each tank is equal to the sum of the flow rates that enter and leave the tank Yet flows Q13 Qa and Q20 are still unknown in Equations 1 2 and 3 To obtain them we use Torricelli s law Qij az Sn sgn hi hi y 2g hi hs A 4 where az is the outflow coefficient sgn x is the sign of the argument x and y is the acceleration due to gravity A 1 The Three tank system Modelling and control 199 The resulting equations to calculate the partial flows are thus Q13 az Sn sgn hi h3 y2g h1 ha A 5 Q32 az3 Sn sgn ha ha 2g ha ho A 6 Q20 Qzo Sn V 2gho A 7 The valves of the pipe connections the pumps and the leaks have been mod eled with the relationships which can be found in Shinskey 1996 The value of the physical parameters of the process are provided in the Appendix B Identification process To get a description of the process by transfer functions it is neccesary to make the identification from experimental data To get the experimental data different open loop experiments with the actual three tank system were done The identifi cation of the process is don
256. the window This general view shows the eJournal with all its functionalities divided into three different sub spaces The extended eJournal version is also shown in Figure C 3 Subspace eJournal Access to other eJournals Allows to get access to the eJournal of other groups this option will be not used in this practices C 1 How to work with eMersion 241 Trash The trash allows to see the data fragments deleted from the current folder These fragments can be reloaded in the eJournal workspace or definitively deleted Language The language of the interface can be changed by pushing on the flags located on the right top part of the interface Subspace Folders A folder represents a directory in the workspace The Inbox folder is a directory with special features It is created automatically with the workspace and is used to receive data from the different components of eMersion as for example data fragments from the experimentation console This folder can not be renamed or deleted It is possible to create and to rename new folders that allow students to organize the work carried out Next the basic functionalities of the eJournal in this sub space are described hereafter Filter Allows to choose the type of fragment to display in the main window There are two ways of filtering a by type There are different type of fragments They can be fragments files sent from the local hard dis
257. there is more than one possible approach to e learning and e teaching 4 Student centred learning Blended learning Z Social learning The student centred learning methodology aims at designing learning tasks from the standpoint of students with a clearly defined sequence of actions instead of a learning process that places full responsibility on the students themselves Kurhila et al 2004 This learning style relies on the effective arrangements of three different elements show tell do The results obtained of an X task set to students SHOW could be used as an example of this learning method The experiment conducted will then be explained using some keywords TELL And finally a research task for students to complete by themselves DO will be set This is a common learning rule for face to face laboratory sessions However if this learning method is applied in the case of virtual and remote labs students can benefit from flexible time and space and could even get instant feedback 116 5 The e Learning Layer The blended learning methodology is understood as a combination of self led learning and other learning resources live seminars or workshops with the pos sibility of forming workgroups Live seminars allow students to get a first insight into the contents of a lesson This will help them confirm that they understand the subject prior to discussing it in group This will therefore encourage them to come properly prep
258. tion The link to the theory and any other document should be indicated here The field Title provides a representative name of the subject of the documents The content of the document is indicated by means of its URL eMersion uses all this metadata information to create a module A module is composed of a set of JSP HTML pages based on the metadata information au tomatically created for the eMersion s course management system When a final user has access to the experimentation environment to carry out an experimental session s he is really accessing the web application that constitutes a working module Hence it is possible to say that a module represents the final facade of the web based experimentation environment An administrator course can create as many working modules as virtual and remote control laboratories s he has to develop For the purpose of this disserta tion three modules were created which correspond to the three remote laboratory prototypes presented in Chapter 4 For each of them all the neccesary documen tation was developed following recommendations described in Chapter 5 156 7 Integration of Layers The automatic bookings system client interface is easily integrated into a working module just by providing its URL The following section will show the module of the three tank prototype fully integrated into eMersion The interoperability of all web components described along this dissertation is shown he
259. to perform the angular position or velocity control of a load using the direct current supplied to the motor see Figure 4 27 The load is composed of an inertia steel disk fixed to the motor s axis and a viscous friction provided by a magnetic brake The engine is fed through a unity gain amplifier capable of delivering the power required to the motor 104 4 Prototypes Developed voltage to motor tachometer potentiometer speed position reference point Figure 4 27 The DC motor scheme A tachometer provides voltage uy in proportion with the angular velocity of the load A potentiometer with a reducer device provides a voltage ug propor tional to its angular position A similar potentiometer is also available to change the reference y directly in the equipment The modelling and control of this process is addressed in Appendix A 2 and no additional information is given here As with previous prototypes this chap ter pretends to further tackle the creation of web based control laboratories using the software tools and approaches provided in this dissertation 4 4 2 Local control of the DC Motor The intrinsic time constant of this mechatronic device defines a challenging prob lem because of the fact that is almost in the same order of magnitude as the Internet transmission time Again the JiL server approach provides a lot of help in the case of these applications by separating the server program in a communication
260. to provide a graphical user interface GUI for Java programs Swing was developed to provide a more sophisticated set of GUI components than the earlier Abstract Window Toolkit 3 2 Adapting virtual labs for remote experimentation 59 cation Here we can see that in addition to the properties inhereted from the Interactivelmage element three new parameters must be configured These pa rameters are related to the IP camera used since a Video object from the web cam jar library is created In the example illustrated in Figure 3 21 the value of these parameters are URL http IP_ADDRESS axis cgi mjpg video cgi MJPEG true delay 20 where URL is the http location provided by the video server AXIS 2400 to access the stream MPEG MJPEG is the reading format TRUE in this case and the delay between the reading of two images was set to 20 ms E Easy Java Simulations webCamImage xml O Introduction Model View Tree of Elements 1 Elements for the view e Simulation View E Frame T DrawingPanel W Camimage Result after compilation WebCamimage on a DrawingPanel container Setting the properties of Camlmage element You will receive output messages here A Trying to run simulation webCamImage I Figure 3 21 WebCamlmage view element on Ejs Remarks The MPEG format is not available in all IP cameras Should this be the case the JPEG forma
261. tor fed by a direct current source see Figure 8 4 with regards to speed and position The DC Motor is a common actuator in control systems and therefore although it is a simple and straightforward example it requires special attention The didactical setup is an electromechanical process intended to perform the angular position or speed control of a load using the direct current supplied to the motor as a manipulated variable Costa Castell et al 2010 8 2 The AutomatLQbs project 167 e Journal Control Lai Posici n angular 6 7 8 9 10 time Acci n de control Position Speed 3 Posicion 90 Speed rpm 0 8 Position Control Speed Control Referencia 90 VIVolts 0 Parametros de control Tiempo seg 10 Freno No L Saturaci n de la acci n de control Real 5 Ejecuci n en tiempo real Control PID en posici n a Hardware b Ejs view in remote mode Figure 8 4 DC motor Miguel Hern ndez University One tank system University of Almer a The four tank system represents one of the most complete processes to teach automatic control concepts Johansson 2000 Dormido amp Esquembre 2003 It can be configurated to enable the study of monovariable and multivariable systems In this case the plant is configured as a two tank system see Figure 8 5 a which objective is to control the liquid level in the lower tank using the upper tank to
262. ts a stable behaviour of the system can be obtained Position control In the same way the motor DC is used also when a fix angular position of the system is requiered For example to fix the position a robot arm requieres to A 2 The DC Motor Modelling and control 211 carry out an accurate position control of the engines in the joints Figure A 11 illustrates a typical scheme of position control with internal feedback of speed in which the output of the position sensor Ug s is a voltage proportional to the angular position of the motor axis Motor reference e troller U Power P corone gt Amplifier wW integrator Uo s position 0 sensor Figure A 11 Typical position control system The PI parameters of the system in position control mode can be easily ob tained from the scheme in Figure A 12 With Kp 0 08 T 0 6 a stable be haviour of the system in position control can be obtained DC Motor f e A Ww 1 reference gt K i 3 U WwW i 0 a Tis TS 1 S Figure A 12 PI controller for position control on the DC Motor Reference tracking Some experiments for testing the performance of the obtained PI controller pa rameters in both speed and position control have been carried out The first experiment starts with the system stabilized around the operating point 0 150 in position control mode see Figu
263. ts not available at their universities and b To increase the quality and robustness of the network of virtual and remote laboratories for a higher number of students and teachers with different teaching concerns 166 8 System Assessment AutomatLGbs Spanish thematic network on learning teaching in control engineering Espa a Mozilla Firefox archivo Y Ver Historial Marcadores Herramientas Ayuda e Q X qu 99 hetpittab ia uned es automatiabfindex_en html X MIG soos P AutomatL bs eit Home Project Information Connection Resources FAQ Acknowledgement 009 Hour 11 37 hrs Network of Automatic Laboratories Spain Optimized to 1024x7t 68 pxls or higher Get access to the experimentation environment AutomatL bs Search criteria Search C Google AutomatLabs le I About AutomatLabs AutomatL bs Project r Department of Computer Sciences and The AutomatL bs project is a network of virtual and remote laboratories For teaching and iha Automatic Control UNED A ut a t learning of control engineering It is based on the integration of the hardware software Main services of AutomatL bs at UNED om human resources provided for the university groups taking part into this project Shortly L AutomatL bs provides to students two elements web based tools for booking time slots in CEA Control Education Group order to carry on the practical experiences and a working environment purposefully d
264. tudy to a national scope academic year 2007 2008 to carry out a more exhaustive test of the ex perimentation system see Figure 8 2 The workgroups of the control engineering research area taking part in the AutomatL bs initiative were the National University for Distance Education UNED the University of Almer a UAL the University of Alicante UA the Polytechnic University of Valencia UPV the Polytechnic University of Catalo nia UPC Miguel Hern ndez University UMH and the University of Le n UNILEON The main challenge of this experience was the integration of the hardware software and human resources of the project participants into the web based ex perimentation environment hosted by the Department of Computer Science and Automatic Control of the UNED AutomatL bs 2009 Vargas et al 2009b 2010 The various groups committed to the following 1 To provide a physical system hosted by the participants to the AutomatL bs network This piece of equipment should be for the exclusive use of the net work during the months when the laboratory is open to students 164 8 System Assessment 27d Experience Integrating labs from other Spanish universities A ES Universita d Alacant Miguel Hern ndez Jniversidad de Alicantet UMH Server UNILEON Server UA Server UAL Server UNIVERSITAS Wait Wait Connection necti Connection Connection Office
265. ue ref If float data type is used to represent every variable except the control mode boolean the state and control vectors will respectively be 12 and 21 bytes In a second stage Java methods should be programmed to control the con nection with LabVIEW server Such as Figure 3 18 shows this implementation requieres the application of the following methods connect disconnect sender and receiver SERVER CLIENT connect sender receiver Ean Signal Conditioning Container Figure 3 18 Communication methods for remote experimentation Listings 3 1 and 3 2 show excerpts of the Java code for establishing and re leasing a connection with a server computer TCP sockets are used to access the network layer In Java the sockets programming is done creating an object and calling to its methods In the Listing 3 1 code the establishing of the connection is carried out in line 5 To create an instance of a Socket object the domain name or IP address and the service port in the remote server are required by the constructor Then in lines 6 and 7 the input and output stream buffers are created These buffers act as FIFO storing queues whose filling and empty ing depend on possible delays in the network communication The disconnection from server side is made by invoking the close method in the socket and the input output stream buffers lines 6 7 8 in Listing 3 2
266. ults in a temperature change of approximately 21 5 degrees 65 5 44 The time constant is the elapsed time to reach a 63 2 of the total change In this case that would be approximately 44 0 632 21 5 57 588 degrees approximately Examining the figure we see that 57 588 degrees is reached in about 17 9 seconds from the time when the step input is applied t 212 sec So a simple first order approximation for the open loop model is TA 21 5 A 25 Vi 17 95 1 The same procedure should be followed to get a simple first order model for sensors 2 and 3 A 3 2 Controller design Now the design of a feedback controller to regulate T with zero steady state error will be addressed Notice that the procedure for designing controllers for sensors 2 and 3 should be the same The controller is a simple PI of the form Ki V Kp Tiref T ref Ti A 26 Substituting into the first order model we obtain the closed loop transfer function T G K s Kj LA see A 27 Tiref 8 7 KpG 1 s GKk The denominator can be rewritten as follows K G 1 K ge Gots G 8 wos we A 28 T T In this way we can now obtain the gains K and K for a desired performance If we want to specify a peak time and damping ratio the following equation can be applied 216 A Modelling and Control T A 29 wo 1 E2 By replacing equation A 29 in equation A 28 the gains for the PI controller obt
267. ument and to extract the information Metadata information is not only useful to edit and connect variables in Ejs but also because it enables to JiL server to send just the bytes of the chosen variables by the developer during the definition of the model In this way there is no overhead in the data sent from server to client since the JiL server already knows what has to be sent to Ejs and how Otherwise getValues methods would need to send the name of the variable required in every iteration that increases data traffic during communication Figure 3 31 shows the exchange of messages between Ejs and the JiL server when setting up a connection with a VI target Ejs Client JiL Server Developer opens an External Page Ejs reports the path of the target VI to JiL JiL extracts a list of the controls in Ejs and provides the URL and P and indicators of target VI and the relative path of target VI sends the list to Ejs into an XML file Developer connects VI variables with Ejs variables and executes the application JiL recovers the connected variables runs VI target sends indicators and waits for any change in the controls JiL sends indicators to Ejs JiL sends indicators to Ejs Ejs receives indicators from JiL Ejs receives indicators from JiL Ejs receives indicators from JiL JiL sets controls and and sends new controls to JiL sends indicators to Ejs JiL sends indicators to Ejs Ejs receives indicators from JiL Ejs
268. urces somehow To overcome this problem an automatic booking and authentication system was developed A simple point to point authentication protocol to use laboratory resources is first introduced This current project was thought in such a way that its use can be extended to a network of remote and virtual laboratories Although this scheme can be applied to remote labs created using any technology in this dissertation we have integrated the point to point authentication protocol into the JiL server approach explained in Chapter 3 Finally a description of the transition towards a booking and authentification flexible scheme inspired in this simple authentication protocol will be provided The following chapter will describe how the final product the automatic bookings system has been completely integrated into the eMersion environment 132 6 1 6 Access Control to Experimentation Resources scheduling access to physical resources 6 1 1 Background In the case of on line labs that enable web access to real devices remote work benches access time should be booked in advance Ferreira amp Cardoso 2005 The idea is to grant access only to users who meet certain requirements Users for example must be enrolled on the corresponding course and to have been positively evaluated by the teaching staff who is in charge of the course in the simulation stage a prerequisite to access to physical resources For this reason it is neces
269. urrent tasks in the server side The control task is a time critical activity running at a sampling period of 20 ms with a priority greater than the other two threads The AI One PT block reads the analogue input signal from the sensor its output is compared to the setpoint input of the PID block and the result is fed into the 40 One PT block that sends the resulting value to the actuator thus completing the control task The data structure composed of the setpoint value the PID control parameters the command to the actuator and other variables is known as the control vector and is sent from the communication task to the control task through RT FIFO queues blocks RT FIFO queues act as a fixed size queue in other words the writing of data to an RT FIFO does not overwrite previous elements These 3Real Time FIFO queues are often used to share data among threads without breaking the determinism of their execution Communication task RT FIFO to share control vector D T LabVIEW 2 Style variable to share video stream Figure 3 15 Three loops running concurrently in the LabVIEW server 3 2 Adapting virtual labs for remote experimentation 49 variables are produced when users interact with the client interface The data array formed by the values sent to the actuator the measurement from the sen sor the current time and other variables are known as the state vector and is transferred from the control task to the communicati
270. utomatic bookings system server 64 Conclusions 122 4 so Gad he ee Re o eo riae gd tN x CDA ark ill 99 99 103 103 104 108 111 113 114 117 117 119 120 123 130 131 Contents 7 Integration of Layers 149 Tel Linking layers AE A AAA RE RR 150 7 2 Integration of Ejs applications in eMersion 151 7 2 1 Ejs built in methods to save data in eJournal 151 1 2 2 Defining types of data fragments 152 7 3 Linking all Work modules in eMersion 153 7 4 A remote laboratory integrated into eMersion 156 1 53 CONCIUSIONS lt a a ae ee a 158 III ASSESSMENT 159 8 System Assessment 161 8 1 A first experience from the UNED 162 8 2 The AutomatL bs project oo o o 163 8 2 1 Access to AutomatL bs llle 165 8 2 2 Remote systems available len 166 8 2 3 Analysis of results 22e 170 8 3 Conchisions s ease hike Bere vue eee ds 173 9 Conclusions and Future Research 175 9 1 Conclusions ica al ea A ee ea A 175 9 2 Buture work s sn taa a UE S m 9 177 Bibliography 181 APPENDICES 195 A Modelling and Control 197 A 1 The Three tank system Modelling and control 198 ALT Modelling it E a a ess 198 A 1 2 Multivariable control 0 ooo o e 201 A 2 The DC Motor Modelling and control 207 A 2 1 Modeling lees 207 Contents A
271. uts provide even more high speed performance on the Q8 Additional features include 2 on board PWM outputs two 32 bit general pur pose counter timers watchdog timers and the ability to reconfigure the 8 encoder counters as 24 bit counter timers When combined with WinCon 4 0 and a PC the Q8 yields a world of real time control possibilities Control virtually any hardware system Rapid Control Prototyping Hardware in the loop control Monitor and control production equipment Development of OEM real time applications Key Features 8 x 14 bit programmable analog inputs 8x 12 bit D A voltage outputs 8 quadrature encoder inputs 32 programmable digital I O channels Simultaneous sampling of both analog and encoder sections 2x 32 bit dedicated counter timers 8x 24 bit reconfigurable encoder counter timers 230 B Hardware Description 2x on board PWM outputs 32 bit 33 MHz PCI bus interface Supports Quanser real time control software WinCon 2000 XP Totem Pole digital I O for high speed Easy synchronization of multiple Q8 boards B 3 The DC Motor datasheet 231 B 3 The DC Motor datasheet B 3 1 Hardware Figure B 7 DC Motor hardware in the Automatic Control Laboratory UNED Figure B 7 shows the hardware components of the DC motor didactical setup The charge is composed of a inertia steel disk tight to the motor s axis and a viscous friction due to
272. va are Java is simple Programming languages are not simple but Java is con sidered a much simpler object oriented programming language when com pared to the programming language C Partially modeled after C Java has replaced the multiple inheritance in C with a simple structure called interface The use of pointers were also eliminated Java is distributed Java is designed to simplify the distributed com puting thanks to its integrated networking capabilities Writing network programs in Java is like sending and receiving data to from a file 3 1 Development of virtual laboratories 31 Java is portable platform independence One of the most com pelling reasons to use Java is its platform independence Java runs on most major hardware and software platforms including Windows Macintosh UNIX and Linux operating systems Java applets are supported by Java compatible browsers An applet is a program written in Java that can be embedded in an HTML page When a Java technology enabled browser is used to view a page that contains an applet the applet s code is transferred to the system and executed by the browser s Java Virtual Machine JVM Java is multimedia images sounds and animation The sizzle of Java is multimedia Sounds Images Graphics and Video In this growing age of multimedia new computers are known as multimedia ready with CD ROM drives sound cards 3D accelerator cards and other new special
273. ve Journal Trash peana E E Folders HectorS3rem m 3 inbox z fan Since 2 T iul hd bd bj HectorS3sim m 3 NS Active Folder Filter by Type Filter by Date Refresh Hew Rename Delete Zip a HectorHoySIM2 m 8 w Fragments r Copy Move Delete Rename Import Export Share Send Assign Finalize Submit Hote pesiertiognem m 3 a Name Author Task State Creation Annotation HectorHoySim m 3 Y madrid m Estudiante x 26 September Text file HectorHoy m 3 amp F portugal m Estudiante 26 September Text file HectorS3rem m Estudiante x 23 September Text file registroThreeTank m Y HectorS3sim m Estudiante x 23 September Text file PAGE Xj HectorHoySIM2 m Estudiante Li T 23 September Text file HectorHoyRem m Estudiante xa 1 23 September Text file Full Journal i HectorHoySim m Estudiante x 23 September Text file HectorHoy m Estudiante xj 23 September Text file registroThreeTankm Dormido z T 23 September Text file T Colos gif Dormido x 23 September GIF image gt Terminado Terminado Figure C 3 Reduced left and extended right views of eJournal In eMersion the initial view of the eJournal is a simplified version where students can see the last 10 fragments existing in the current folder as well as their state and authors Figure C 3 illustrates an image of this version Then the student can access to the extended version by clicking on the button located in the lower part of
274. w the temporal evolution of the main signals speed position reference and manipulated variables Indicators panel of instantaneous values At the bottom right corner of the interface a set of numerical fields display the current value of the system variables BE D C Motor eJournal Control Language angle 12 14 16 seconds U P 1 D Video status No video in simulation mode _ Augmented PID control parameters Kp 0 08 Ti 0 6 Td 0 Pos tion 149 7 Speed sec 1 7 My 38 O Speed Control Position Control Refl iseg 150 Ulvotts 0 3 gt Eli E r3 Time sec Status Lab Simulation mode no connected BI D C Motor eJournal Control Language 18 20 22 A seconds U P 1 D Video status No video in simulation mode PID control parameters Kp 0 08 Ti 06 Td 0 Position 164 9 Speed sec 30 Speed Control O Position Control Rates Uvots 1 1 C Jt 6 a Time sec b Status Lab Simulation mode no connected b Simulation mode in speed control Figure 4 31 The web based laboratory of the DC motor in virtual mode 110 4 Prototypes Developed Figure 4 32 a depicts the DC Motor interface working in remote mode with visual feedback The steel disk is graduated in degrees to visually check the correct position of the load A polygon element of the Ejs graphical library is used to e
275. will be analyzed that sustain and support the learning process of students in a remote experimentation context The software tool used to publish web based labs through Internet eMersion is also presented and described in detail More information on the use of eMersion can be found in Appendix C Chapter 6 The analysis design and implementation of an Automatic Booking System to organize and optimize the use of the physical resources available in laboratories is also presented The integration of this application in the remote experimentation environ ment eMersion is addressed in Chapter 7 Chapter 7 The integration of the experimentation layer and learning layer is presented All the web components described in previous chapters are cou pled in order to obtain the final remote experimentation environment for engineering educational programs Chapter 8 An assessment system is presented in this chapter The UNED pilot experience and the AutomatL bs project are two experiments carried out with real students aimed at evaluating how useful the system is as a teaching tool Results are analyzed and discussed at the end of this chapter Chapter 9 Conclusions and future research are presented 1 4 Publications Awards and Projects Journal papers 1 H Vargas J S nchez S Dormido C A Jara F Candelas F Torres Do cencia en Automatica Aplicaci n de las TIC a la realizaci n de actividades 1 4 Publications Aw
276. wo combo box elements located in this sub space A set of icons allows users to create folders rename and or delete them and finally compress the folder selected in the active folder selector into a ZIP file Fragments Sub space The third and last sub space of the eJournal contains a set of useful functions to handle data Fragments Students use the functions Copy Move Delete Rename Import Export Share Send Assign Finalize Submit and Note to manage fragments and at the same time interact and collaborate with other participants of the community of practices profes sors tutors other students A list of Fragments is deployed as a table in the lower part of this sub space Files are deployed in accordance with the filter and the active folder used at every moment The columns in the table provide meta information or attributes associated to each Fragment as for example Name Author Task Status Creation and Annotation The eJournal web component fulfills a key role in the introduction of the social learning model used in this dissertation Part of the ideas described here can be mapped into actions and situations that occur in a presential laboratory 130 5 The e Learning Layer 5 3 Conclusions This chapter has provided essential information about how to appropriately face the education of students in a distributed learning context through the Internet For this purpose the following topics have been discussed
277. work modules which somehow represent the different types of practices or laboratories avail able in the system When managing a course the administrator can create as many modules as s he needs To create a new module the teacher administrator of the course must fill out a web form see Figure 7 4 Course Administration System Welcome Profesor You are managing the course Practicas Y Home Modules Protocols Groups m Modules management Protocol name Choose module s logo Three tank system remote laboratory Examinar http m Protocol objective Module language Experimentation window size in pixels Level control over a coupled three tank svstem english x Height 550 Width 760 Protocol description file Experimentation URL C guideemersion html Examiner http lab dia uned es threetank threetankclient html Toolkit Documentation Title Booking System http bookings dia uned es clienthtn m Title Task Protocol http lab dia uned es threetank protocol html He Examinar p Examinar Title Interface MAnual http lab dia uned es threetank manuallhtml MEI Fie Examinar ZP Examinar Create module Reset Figure 7 4 Creating a new module included into a course in eMersion As seen in Figure 7 4 most of the setting parameters needed to create a new module are related to the URL of each web component that integrate the module A brief des
278. xperimen tation console Ejs applications 2 The integration of Ejs applications into eMersion to save data fragments to the eJournal space On the other hand the web experimentation environment requires to integrate some supplementary components to give access to other pieces of information requested during the learning process eMersion automatically makes the linkage between the web components above mentioned This task is carried out during the creation of a working module in eMersion A module is in a few words the final web interface eMersion s facade or cockpit where students work and therefore from where they should have access to any further web component From an eMersion management point of view the creation of a module is just a part of the configuration process however it is precisely there that the integration of every web component is carried out A complete summary of both the eMersion management and the cockpit usage can be found in Esquembre et al 2007 and Appendix C respectively 7 2 Integration of Ejs applications in eMersion 151 7 2 Integration of Ejs applications in eMersion Reports about the students work made during a remote experimentation session are required by the teaching staff These results are based on the data registers obtained when performing the practical experiments with the help of the exper imentation console This section is devoted to explaining how and what type of data regi
279. ype I The Three tank system 89 1 The physical equipment should be identifiable so as to define what kind of equipment is connected 2 The equipment should be fully controllable and diagnosticable by the local computer 3 For safety reasons the full controllability of the physical equipment may not be exposed to the outside world Controllability also implies that it is always possible to place the equipment in a known state 4 Other requirements such as reliability and maintainability should also be considered dt sec output limit PID parameters process variable setpoint H Initialization manual umanual a VI for PID control algorithm usat Discrete PID compute float32 bi ad bd ar up u ipre calculated coefficients bi kp tsjti ad tdj Ed N ts bd kp N ad ar ts tt j reset anti windup icalculate control signal up kp beta ysp y ud ad ud bd y yold if ti 04 fonly P D actions u up ud felse P I D actions u up ui ud saturation of control signal if u cumin usat umin yelse if u gt umax usat umax else usat u iupdate states iF ti 04 if manual 0 ui umanual usat umanual tracking For PI felse uizui bi ysp y ar usat u J with antiwindup yold y b LabVIEW Formula node Figure 4 11 Discrete PID controller with antiwindup protection pro grammed in LabVIEW In this particular context all the prototypes descr
280. ystem Modelling and control 201 Figure A 3 Three tank system representation by transfer functions Table A 1 contains the first order model parameters for each transfer function Table A 1 Parameters obtained from the identification process ELO A xm D 755 0 995 1 167 la s 0 3645 0 Finally the description by means of transfer functions is shown in the equa tions A 8 and A 9 46 5s H5 MS Umen te 4 8 1 167e749 5s o neto 048 A 9 A 1 2 Multivariable control From a practical point of view we can run controllers in two different modes manual or automatic In manual mode the operator directly manipulates the controller output typically by moving sliders that increase or decrease it In au tomatic mode a computer or some other automatism generates the controller out put signals Manual mode is most useful during the plant s startup and shutdown or in an emergency situation Conventional controllers have a manual automatic switch to transfer the controller from automatic mode to manual 202 A Modelling and Control Interaction study of the variables Before addressing the design of controllers to be used in automatic mode a vari ables interaction study has been done A multivariable system has a set of input and output variables which can be coupled with each other We talk about cou pling among variables when the change in one input variable affects to more than one output variabl

Download Pdf Manuals

Related Search

Related Contents

chapter one getting to know the basics  ミニ通 信 新計量法（SI単位）について（その 1）  【User Manual】 DP-102E RS-232 to EDX Interface - Lite  Istruzioni per l`uso di base  ソニーSxS™メモリーカード USB リーダー/ライター Sony SxS™ Memory  User Manual  

Copyright © All rights reserved. Failed to retrieve file