A Wireless Sensor Network Platform for Structural Health - open-ZB
Contents
1. A Wireless Sensor Network Platform for Structural Health Monitoring enabling accurate and synchronized measurements through COTS custom based design R Severino R Gomes M Alves P Sousa E Tovar L F Ramos R Aguilar P B Lourenco CISTER Research Unit Polytechnic Institute of Porto ISEP IPP Portugal e mail rars rftg isep ipp pt ISISE Research Unit University of Minho Guimaraes Portugal e mail raguilar lramos civil uminho pt Abstract Structural health monitoring has long been identified as a prominent application of Wireless Sensor Networks WSNSs as traditional wired based solutions present some inherent limitations such as installation maintenance cost scalability and visual impact Nevertheless there is a lack of ready to use and off the shelf WSN technologies that are able to fulfill some most demanding requirements of these applications which can span from critical physical infrastructures e g bridges tunnels mines energy grid to historical buildings or even industrial machinery and vehicles Low power and low cost yet extremely sensitive and accurate accelerometer and signal acquisition hardware and stringent time synchronization of all sensors data are just examples of the requirements imposed by most of these applications This paper presents a prototype system for health monitoring of civil engineering structures that has been jointly conceived by a team of civil and electrical and computer engineer2. Configuration application C amp C App briefly described in Section 6 1 provides the system user with a human machine interface HMI to configure the system and also an application programming interface API to integrate the WSN system with the data processing analysis applications The latter enable to infer about the reaction of the monitored structure to natural vibration or impacts as outlined in Section 6 2 4 WSN ARCHITECTURE As previously stated the proposed SHM system aims at sampling several accelerometers placed at different locations in a structure in a synchronized fashion Sampled data is to be stored in each Sensing Node until it is retrieved by a central node for processing To enable the analysis of the results namely the modal shape analysis it is crucial to guarantee the temporal correctness of the system 4 1 Guaranteeing Synchronization According to Cinque et al 2006 the maximum drift between samples should be computed as presented in 1 C C s SZYyi 1 N j 1 where C s is the clock of the i th sensor N is the total number of sensors and f is the sampling frequency The existing timers in the TelosB platform depend on a 32 768 Hz Citizen CMR200T quartz crystal Citizen 2006 This crystal features a drift of 20 ppm in relation to its nominal frequency This means that in the worst case there is a drift of approximately 20 us at every second Assuming a sampling frequency of 100 Hz re
3. SenSys 04 Baltimore MD USA Whelan Matthew J Gangone Michael V Janoyan Kerop D Jha Ratneshwar 2009 Real time wireless vibration monitoring for operational modal analysis of an integral abutment highway bridge In Engineering Structures Elson J Girod L Estrin D 2002 Fine grained network time synchronization using reference broadcasts In Proceedings of 5th symposium on Operating systems design and implementation OSDI 2002 pages 147 163 Boston MA USA Ganeriwal S Kumar R Srivastava M B 2003 Timing syne protocol for sensor networks In Proceedings of Ist international conference on Embedded networked sensor systems SenSys 03 pages 138 149 Los Angeles California USA Maroti M Kusy B Simon G Ledeczi A 2004 The flooding time synchronization protocol In Proceedings of 2nd international conference on Embedded networked sensor systems SenSys 04 pages 39 499 Baltimore MD USA Werner Allen G Tewari G Patel A Welsh M Nagpal R 2005 Firefly inspired sensor network synchronicity with realistic radio effects In Proceedings of 3rd international conference on Embedded networked sensor systems SenSys 05 pages 142 153 San Diego California USA Rowe A Mangharam R Rajkumar R 2006 RT Link A time synchronized link protocol for energy constrained multi hop wireless networks In Proceedings of the 3rd Annual IEEE Communications Society o
4. ation is of major importance for this kind of monitoring applications Xu et al 2004 Lynch et al 2006 Cinque et al 2006 Whelan et al 2009 This means that samples from all sensors must be acquired in a synchronized way in order for the data analysis algorithms to provide consistent results 3 2 Snapshot of the System Architecture The system architecture was designed in order to satisfy the identified application requirements and is illustrated in Fig 1 considering a prototype system composed by four Sensing Nodes Each Sensing Node is composed by a TelosB node Crossbow 2009 with a signal acquisition board SAB attached to a MEMS acceleration sensor see Section 5 Sensing Node 1 MEMS SAB TelosB Sensing Node 2 MEMS SAB TelosB Sensing Node 3 mems SAB TelosB leo Coordinator Data Node Analysis TelosB Apps Sensing Node 4 MEMS SAB TelosB Fig 1 Snapshot of the System Architecture All four Sensing Nodes communicate with a Coordinator Node also a TelosB node via a standard communication protocol IEEE 802 15 4 The Coordinator Node supervises the network and nodes activities e g node configuration start stop sampling and guarantees a tight synchronization between all nodes it also forwards the configuration parameters and dispatches the acquired data to the Command amp Configuration Application C amp C App The WSN architecture is described in Section 4 The Command and
5. ation scenario 6 1 Command and Configuration Application In order to provide the necessary HMI and API for the data analysis applications a Command and Configuration Application C amp C App was developed Fig 6 Coordinator ree Tee e Shae Se I Data Node aoe Analysis TelosB 8 Ss eS Apps Fig 6 Command amp Configuration Application The available controls of the C amp C App enable full control over the acquisition configuration parameters i e axis selection sampling rate sampling period sampling duty cycle etc and also provides a quick evaluation of the presence of the system nodes Several additional features are also built in to assist the user with relevant information on the network and acquisition parameters configuration One additional goal of the C amp C App was to provide a convenient interface between the WSN and the data processing analysis application The implemented mechanism allows a transparent interface with the system in a very similar with the previously used which are typically serial data interfaces To complete the data acquisition process a VI routine was developed in Labview Labview 2006 for the interpretation and conversion into standard units for receiving the messages from the serial port as well as their local storage in the central station 6 2 Experimental Modal Identification Tests A single degree of freedom structure represented by an inverted pendulu
6. cation C amp C App Implementation of the Sensing and Coordinator Nodes software was done in nesC Gay et al 2003 over the TinyOS operating system TinyOS 2010 The open ZB implementation of the IEEE 802 15 4 protocol has been used MASS 2007 Open ZB 2010 Fig 2 presents a message sequence chart of the application Coordinator Sensing Node 1 ensing Node EN Idle _ configuration Ready Paes Acquiring Dataremains in SAB Stop Transmittin Done Fig 2 Message sequence chart The WSN application commutes between 6 states as follows 1 Idle As soon as the nodes are powered they enter the Idle state At this stage the open ZB IEEE 802 15 4 stack is initiated and the nodes try to synchronize and associate with a PAN Coordinator The Channel Scan feature of the protocol stack is disabled since the network topology is fixed 2 Ready As soon as every node is synchronized the user signals the Coordinator to initiate the Ready state This is done by changing the information in the IEEE 802 15 4 beacon payload Each Sensing node receives the beacon parses the payload information and immediately checks the presence of a SAB The Coordinator is then signalled by each node concerning its readiness Upon the reception of this message the Coordinator informs the C amp C App about the state of each node 3 Acquiring When every node is configured the user can star
7. cular case the outputs of the Triaxial accelerometer are multiplexed by a 3 1 multiplexer The selected analogue signal then crosses the initial buffering and programmable gain stages Then an analogue 8 order Butterworth filter limits the signal s maximum frequency to 100 Hz to avoid undesired aliasing effects Then the filtered signal goes through a final conditioning stage and enters into a high resolution 24 bits ADC The digital circuitry connections arrows connected to the microcontroller MCU represent its relation towards the MCU internal architecture as briefly described next The MCU is responsible for controlling all the SAB hardware which includes the procedures for proper ADC behaviour handling the samples storage until WSN platform request and additional samples pre formatting Note that the voltage converter inverter that supplies the analogue circuitry is directly connected to the MCU enabling on off control The input multiplexer the programmable gain amplifier PGA and the high resolution ADC are connected to the MCU by several GPIO lines The data transmission from the MCU to the flash memory is achieved through the serial peripheral interface SPI bus The MCU connects with the WSN platform by its internal UART hardware and a couple of two GPIO lines 6 TEST AND VALIDATION This section describes how the proposed SHM system and the underlying architecture was tested and validated in a real applic
8. d Commands used to configure the SABs these commands are transmitted to the corresponding node and then directly forwarded to the SAB using regular IEEE 802 15 4 data frames 2 Network Commands used to manage the monitoring application There are two kinds of commands within the former category a Node Management commands b Application Management commands The Node Management commands are sent to the Sensing Nodes using regular IEEE 802 15 4 data frames during the application Ready state These include setting the behaviour of the node active passive remote reset channel selection and requesting onboard sensor reading temperature and humidity The Application Management commands are sent within the payload of the IEEE 802 15 4 beacon frames Fig 2 so that all nodes receive and process the command at the same time thus guaranteeing synchronization there is no contention in beacon transmission The commands are described as follows 1 IDLE This command indicates that the system is in dle state waiting for input from the User 2 READY It marks the beginning of the configuration phase for the nodes When receiving this command the Sensing Nodes wait for a configuration packet from the Coordinator including sampling rate period and time They also wait for a message to set their behaviour as active or passive 3 START This command triggers the beginning of the signal acquisition from the accelerometer
9. e Coordinator s beacon is set to IDLE Upon application input the Coordinator changes payload to READY signalling all boards When the Sensing Node is informed of the beginning of the Ready state it will immediately check for the presence of the SAB using its UART interface If the SAB responds the Sensing Node signals the Coordinator that everything is ready Otherwise it will signal the error using an Error Message with the respective error code Sensing Nodes are then activated and configured by the Coordinator Sensing Device Application Open ZB Stack TinyOS RADIO EIEL UART I O Connections for Sync Fig 4 Architecture of a Sensing Node Sampling is started by sending the START command in the beacon payload When the sampling time expires the Coordinator changes its beacon payload to send the STOP command Upon reception of the GET command the Sensing Nodes initiate the transmission of the sampled data stored at the SAB to the Coordinator Node Finally the Sensing Nodes signal the Coordinator that the data transmission is over 5 SIGNAL ACQUISITION SUB SYSTEM A custom designed signal acquisition board SAB had to be conceived for supporting a a high resolution 24 bit ADC b enough memory for storing data samples MEMS sensors are quite appealing for WSN applications due to their low energy consumption low voltage operation small size and low cost Although there are several MEMS sensors in the mark
10. equencies were detected The last stage of the experimental operational modal analysis process consists on the estimation of the dynamic properties of the structures by means of their natural frequencies damping coefficients and mode shapes For this purpose a more refined data processing method was used which consisted on the evaluation of the time series recordings with 3 conventional and new developed sensors 7 CONCLUSIONS This paper describes a wireless sensor network WSN system for monitoring physical infrastructures Building upon the cons of traditional wired based solutions several solutions based on WSNs have been proposed but there was a lack of ready to use and off the shelf WSN technologies able to fulfil some more demanding requirements of these applications e g monitoring bridges historical buildings or vehicles structures This paper describes a solution that is mostly based on standard and off the shelf technologies namely in what concerns hardware platforms operating system and communication protocol Only a minimum set of custom designed signal acquisition hardware was conceived in order to serve as an interface between the accelerometers and the sensing nodes Our solution is low power and low cost and guarantees accurate and time synchronized measurements Future work will focus on extending the WSN architecture proposed in this paper in order to support a higher number of nodes and a wider region und
11. er monitoring still guaranteeing a tight synchronization between all nodes REFERENCES Pines Darryll J Lovell Philip A 1997 Conceptual framework of a remote wireless health monitoring system for large civil structures 1998 Smart Mater Struct 7 627 Lynch Jerome P Low Kincho H Straser Erik G 2000 The Development of a Wireless Modular Health Monitoring System for Civil Structures MCEER Mitigation of Earthquake Disaster by Advanced Technologies MEDAT 2 Workshop Lynch Jerome P Loh Kenneth J 2006 A summary Review of Wireless Sensors and Sensor Networks for Structural Health Monitoring The Shock and Vibration Digest v 38 n 2 pp 91 128 Ceriotti M Mottola L Picco G P Murphy A Guna S Corra M Pozzi M Zonta D Zanon P 2009 Monitoring Heritage Buildings with Wireless Sensor Networks The Torre Aquila Deployment 8th ACM IEEE Int Conf on Information Processing in Sensor Networks IPSN SPOTS San Francisco CA USA Cingue Marcello Cotroneo Domenico Caro Giampaolo De Pelella Massimiliano 2006 Reliability Requirements of Wireless Sensor Networks for Dynamic Structural Monitoring In International Workshop on Applied Software Reliability WASR 2006 pages 8 13 Xu Ning Rangwala Sumit Chintalapudi Krishna Kant et al 2004 A Wireless Sensor Network For Structural Monitoring In Proceedings of 2nd international conference on Embedded networked sensor systems
12. et capable of satisfying the requirements outlined in Sub Section 3 1 complete ready to use COTS devices are still scarce Some of the most suitable devices for these applications are commercialize by Advanced Sensors Calibration ASC Germany Crossbow USA and Silicon Designs Inc USA Among the referred manufactures portfolios the triaxial accelerometer model ASC 5631 002 Advanced Sensors Calibration 2009 was identified as an interesting solution characteristics outlined in Table 1 Table 1 ASC 5631 002 characteristics Fig 5 depicts the overall architecture of the SAB A common energy source e g battery supplies the COTS WSN platform and the SAB hardware The system voltages are then derived from this energy source Note that both the WSN platform and the SAB s digital section voltage regulator are independent of the remaining system voltages This arrangement allowed switching on off all the onboard analogue circuitry which enables a substantial improvement in the overall energy consumption LDO voltage regulator 3 3 V Energy source Voltage Converter Inverter 7 V 7 V Low noise voltage regulators 5 V 2 5 V 2 5 V 5 V Triaxial accelerometer X Y Z Multiplexer 3 1 WSN Platform GPIO UART Buffering Anti aliasing filtering Conditioning Stage High resolution ADC NDI ORo eye Fig 5 Sensor Acquisition Board SAB architecture In this parti
13. m is one of the simplest examples used by the civil engineers to explain the fundamentals of the dynamics of structures In this work this structure was also used as a tool to evaluate and understand the behaviour of the COTS WSN and the developed prototype for operational modal analysis of civil engineering structures As it is shown in Fig 7 the studied specimen consists in an inverted wooden pendulum with 1 70 m height built specially for testing purposes in the civil engineering laboratory at the University of Minho The pendulum was designed in such a way that its dynamic properties replicates the properties of the Mogadouro s Clock Tower an old masonry tower in the northern part of Portugal which was previously studied and presented in Ramos 2007 For comparison purposes both WSN platforms were evaluated considering as references conventional wired based systems which consist in high sensitivity piezoelectric accelerometers model PCB 393B12 PCB 2009 as well as the NI USB9233 NI 2009 as data acquisition board Fig 7 Laboratory system idealization experimental setups The initial tests were meant to observe the performance of the COTS technology on WSN platforms for dynamic monitoring studies With this purpose the accuracy of the time series recordings of these platforms MICA2 solution MTS400 board was evaluated using only one of the conventional accelerometers and mote placed at the top of the Pendulum The re
14. n Sensor and Ad Hoc Communications and Networks pages 402 411 Reston VA USA Paek Jeongyeup Chintalapudi Krishna Govindan Ramesh Caffrey John Masri Sami 2005 A Wireless Sensor Network For Structural Monitoring Performance and Experience In Proceedings of the Second IEEE Workshop on Embedded Networked Sensors EmNetS IT Sidney Australia Crossbow 2009 TelosB mote platform datasheet Online at http www xbow com Products Product_pdf_files Wireless_pdf Tel osB_ Datasheet Citizen 2006 Tuning Fork Crystal Units CMR200T CMR250T datasheet Online at http www citizencrystal com images pdf k cmr pdf TinyOS 2010 TinyOS website Online at http www tinyos net Open ZB 2010 Open ZB OpenSource Toolset for IEEE 802 15 4 and ZigBee Online at http www open zb net IEEE 802 15 WPAN Task Group 4 TG4 2010 Online at http grouper ieee org groups 802 15 pub TG4 html Gay D Levis P Von Behren R Welsh M Brewer E Culler D 2003 The nesC language A Holistic Approach to Networked Embedded Systems In Proceedings of the Programming Language Design and Implementation Advanced Sensors Calibration 2009 Capacitive Accelerometer ASC5631 Preliminary datasheet Germany Cunha A Koubaa A Severino R Alves M Open ZB an open source implementation of the IEEE 802 15 4 ZigBee protocol stack on TinyOS 4th IEEE International Conference on Mobile Ad hoc and Se
15. network performance The IEEE 802 15 4 protocol provides a_standard based solution for synchronization beacon enabled operation mode that fits the application requirements Section 3 1 Thus it has been selected for the WSN communication infrastructure A Coordinator node officially named PAN Personal Area Network Coordinator schedules channel access and data transmissions in a messaging structure the Superframe This node is also responsible for periodically transmitting a beacon frame announcing the start of the Superframe IEEE 802 15 TG4 2010 Upon beacon reception each Sensing Node triggers an external GPIO General Purpose Input Output pin on its Signal Acquisition Board SAB in order to synchronize it 4 2 Communication Architecture The prototype system consists of five TelosB Fig 1 nodes These hardware platforms feature a TI MSP430 16 bit microcontroller a CC2420 RF transceiver IEEE 802 15 4 compliant 48 kB of Program memory in system reprogrammable flash 10 kB of EEPROM two UART communication ports and I2C They also include in board light temperature and humidity sensors which might be useful for some SHM application scenarios Four nodes act as Sensing Nodes and control the corresponding SABs while one node acts as the Coordinator Node assuming network management including network configuration and synchronization data collection and interfacing with the Command and Configuration appli
16. nodes c not relying on standard communications protocols commonly they use IEEE 802 15 4 compliant devices that neither implement the IEEE 802 15 4 medium access control MAC nor ZigBee protocols d not building upon de facto operating systems OS for WSNs platforms e g TinyOS Contiki e not relying on COTS technologies more cost effective Examples of relevant work follow highlighting some of their limitations The system proposed by Xu et al 2004 which was re evaluated by Paek et al 2005 despite using a reasonable sampling resolution 16 bits Jacks an explicit synchronization mechanism between the sensing devices The implementation provides a posteriori time correlation of the samples which is not satisfactory for some operational modal analysis algorithms that require that samples from all sensors are acquired simultaneously Researchers at WSU SL e g G Hackmann et Al proposed a system based on iMote2 platforms which may present some system lifetime limitations due to their energy consumption Additionally no strict sensor data synchronization is supported forcing to correlate data a posteriori and validation was just based on external stimulus not addressing the natural vibration or on simulation Whelan et al 2009 described an innovative system composed of twenty sensing nodes deployment in a highway bridge Nevertheless the system uses a non standard communication stack and the WSN platform micro
17. nsor Systems MASS 07 Pisa Italy October 2007 pp 1 12 Aguilar R Ramos L Louren o P B Severino R Gomes R Gandra P Alves M and Tovar E Operational Modal Monitoring of Ancient Structures using Wireless Technology Proceedings of the XXVIII International Modal Analysis Conference IMAC 2010 Jacksonville Florida USA 2010 Labview LabView User Manual Release 8 0 National Instruments Corporation USA 2006 NI User Guide and Specifications www ni com Accessed December 2009 PCB Product Catalogue Accessed December 2009 Ramos L Damage Identification on Masonry Structures Based on Vibration Signatures PhD Thesis Universidade do Minho Guimaraes Portugal 2007 Van Overschee P and De Moor B Subspace Algorithms for the Stochastic Identification Problem Proceedings of the 30th Conference on Decision and Control Brighton England 1991 Welch P D The Use of the Fast Fourier Transform for the Estimation of Power Spectra A method Based on Time Averaging over Short Modified Peridograms IEE Transactions on Audio and Electro Acoustics 1967 G Hackmann F Sun N Castaneda C Lu and S Dyke A Holistic Approach to Decentralized Structural Damage Localization Using Wireless Sensor Networks IEEE Real Time Systems Symposium RTSS 08 December 2008
18. processor does not run a known OS Additionally they provide no detail on the synchronization mechanism Ceriotti et al 2009 presented a very complete implementation of a SHM application that allows monitoring several phenomenon of interest when monitoring heritage buildings accelerations deformation and environmental parameters However the particularities of the system and its inherent customization level limit its application to a narrow type of structures Moreover the synchronization mechanism is based on a custom middleware and takes few advantages of the native functionalities of the communication protocol requiring a constant refreshment and storage of temporal information in order to maintain time consistency 3 SYSTEM OVERVIEW 3 1 System Requirements The aim of the system is to sample in a synchronized fashion multiple accelerometers placed at different locations in a structure and forward the data to a central station for later processing The most relevant application requirements were identified as follows XYZ accelerometer triaxial Max measurement range 1 g Minimum sensitivity 1 V g Typical resolution 1 mg Max resolution 50 ug Frequency response 3 dB 0 100 Hz Max sampling rate 100 Hz Max sampling drift between sensors 10 ms ADC resolution 24 bits 0 sample lost during sampling process Ensuring the correct synchronization of the sensing oper
19. rio under consideration operational modal analysis of Civil Engineering structures but also to other types of applications where mechanical constructions e g industrial machinery vehicles under stress natural or induced require structural integrity monitoring and or analysis The remainder of this paper is structured as follows Section 2 presents some related work in this area Section 3 provides a system overview emphasising the underlying application requirements Section 4 details the WSN architecture and related implementation aspects The hardware platform with particular emphasis on the signal acquisition board is described in Section 5 In Section 6 a comprehensive explanation of the application interface with the WSN and the application scenario is presented together with a discussion on the results of the tests carried out to validate the prototype platform Finally Section 7 draws some conclusions and outlines future work 2 STATE OF THE ART SHM has been a very active research area among both academics and industrialists especially in what concerns recent developments in WSN and Micro Electromechanical Systems MEMS Lynch et al 2006 Nevertheless existing solutions for SHM using WSNs present one or more of the following limitations a low sampling resolution typically 8 12 bits systems which invalidates SHM based on operational modal analysis b no explicit synchronization mechanisms between sensing
20. s The SABs are synchronized at each beacon and save the samples in its internal memory 4 STOP Upon reception of this command Sensing Nodes stop the data acquisition procedure command sent to the SABs and wait for further instructions 5 GET lt address gt The Coordinator polls each Sensing node with the GET command to trigger the transmission of the sample data stored at the Sensing Nodes SABs memory Each Sensing Node checks the address embedded in the beacon payload 6 RESET This command signals the end of an acquisition cycle After receiving this command a Sensing Node switches to the Ready state All commands are acknowledged by the Coordinator upon reception at the UART sent by the C amp C App 4 4 Sensing Nodes The Sensing Nodes Fig 3 control and synchronize the acquisition of the SABs and carry out the acquisition of the embedded sensors measurements temperature humidity voltage luminosity Pee ak EAN ov i Seat Ly bP M Fig 3 Sensing Node SAB and accelerometer The architecture of a Sensing Node is illustrated in Fig 4 All the application as well as the open ZB stack was developed in nesC over TinyOS Communications with the SAB are handled using the UART serial interface of the TelosB Two additional general purpose input output GPIO pins of the TelosB are used to enable the synchronization of the SAB and to control the communication flow At the beginning of the application th
21. s It merges the benefits of standard and off the shelf COTS hardware and communication technologies with a minimum set of custom designed signal acquisition hardware that is mandatory to fulfill all application requirements 1 INTRODUCTION Structural Health Monitoring SHM and damage identification at the earliest possible stage have been receiving increasing attention from the scientific community and public authorities Damage identification is relevant to all engineering fields as service loads and accidental actions may cause damage to the structural systems Pines et al 1997 Conventional monitoring systems used for these applications in civil engineering studies involve large number of wires copper or fibber optic cables and centralized data acquisition systems with remote connections As damage is a local phenomenon and in order to achieve high accuracy it is important to monitor the structural behaviour at fine grained level Thus a sufficiently large number of measuring points is necessary The fact that the conventional sensor platforms use wires increases the cost of the monitoring systems and creates difficulties in their maintenance and deployment Adding to the fact that the cost of traditional wire based monitoring systems is driven by the number of sensors the installation time and installation costs limit the scale of deployment of such systems Lynch et al 2006 From experience the installation time of a struc
22. s With this purpose the effect of an impulse force was registered using one conventional accelerometer and one new sensing node both located at the top of the pendulum The tests were carried out considering a sampling rate of 100 Hz and sampling time of 10 s The results are shown in Fig 9 Acceleration mg Acceleration mg 150 0 50 0 25 0 00 Yt 0 25 75 4 New Prototype of WSN New Prototype of WSN Conventional Wired Based Systems 0 50 Conventional Wired Based Systems 150 ae 0 1 2 3 5 0 i 7 8 Time s Time s a b Fig 9 Time domain series recorded using the developed prototype of WSN platform a High amplitude excitation recordings and b lower amplitude excitation recordings As it was shown even for signals with amplitudes below than 0 25 mg the records from the new developed WSN platform and the conventional wired based accelerometers presented a remarkable degree of similarity The subsequently stage consisted on the verification of the accuracy of the frequency content of the acquired signals with the developed WSN platforms Considering the same pair of sensors located at the top of the pendulum and 30 s of sampling time experiments in two excitation scenarios were carried out random impacts tests vibrations with amplitudes below 5 mg and ambient noise tests vibrations with amplitudes below 1 5 mg The Welch Spectrums Welch 1967 of the time series record
23. s were calculated and are presented in Fig 10 Conventional Systems New Prototype of WSN Excited Tests Ambient Tests located at the top of the pendulum using parametric time domain techniques such as the Stochastic Subspace Identification SSI method Van Overschee and De Moor 1991 Fig 11 shows the results of this analysis only for the case of random excited system Conventional Systems New Prototype of WSN 1 Mode Shape i eS 2 Mode Shape 3 Mode Shape Fig 11 Experimental modal analysis results under excited environment Tests new WSN Platform The first two mode shapes of the structure were identified with no uncertainties However there was registered a light difference in the 3 mode shape which will be further investigated in future stages of the present research project Table 2 summarizes the results of the experimental modal identification studies performed in the pendulum using the conventional wired based systems and new WSN platforms Table 2 Modal Identification Results Conv Systems New Prototype Error of WSN Fig 10 Frequency domain results Tests new WSN Platform The results evidenced the high accuracy of the resultant frequency domain spectrums calculated from the records of the new developed system With this respect even in the case of ambient noise tests outstanding similarities in the content of fr
24. sults in a sampling period of 10 ms For keeping the drift bellow 10 ms according to the application requirements it will be necessary to synchronize every 500s at most This result imposes the existence of a synchronization mechanism in the WSN so that all nodes have the same time reference There already exist some mechanisms to achieve synchronization in wireless networks The simplest approach is to use the Global Positioning System GPS as the source for a universal clock GPS can provide extremely accurate timing but requires special typically power hungry receivers and a clear sky view Many of the proposed protocols solve the synchronization problem by transmitting in band synchronization information Typically these involve creating some form of hierarchical organization and use it to distribute timing information There are several in band time synchronization schemes in the literature where some providing good accuracy are RBS Elson et al 2002 TPSN Ganeriwal et al 2003 or FTSP Maroti et al 2004 Notably the work from Werner Allen et al 2005 is the only practical synchronization strategy that does not require nodes to construct a hierarchical organization but it can take an unbounded number of broadcasts to achieve synchronization Another approach to this problem is RT Link Rowe et al 2006 a TDMA like protocol that can use an out of band synchronization mechanism avoiding in band solutions that reduce
25. sults of these tests are presented in Fig 8 Acceleration mg Acceleration mg 7 400 200 Commercial WSN Platform Conventional System 30 COTS WSN Platform Conventional Wired Based System 400 Time s Time s a b Fig 8 Time domain series recorded using COTS WSN platforms a low amplitude excitation recordings and b higher amplitude excitation recordings The results of the first test indicated the good performance of the commercial WSN platforms for measuring high amplitude vibrations As it was expected for signals with amplitudes below 20 mg the WSN platforms recorded only noise it is even feasible to observe the digitalizing lines due to the low resolution of the micro accelerometers and the ADC embedded However it is important to state that in SHM studies of civil engineering structures vibrations with amplitudes below 2mg are commonly found Moderate differences less than 5 were found in the frequencies detected with both systems wired and COTS WSN as well as meaningless results for the mode shape detection task due to the lack of the implementation of synchronization algorithms in the commercial WSN platforms Using the developed prototype of WSN platform a second round of tests were carried out considering the same inverted pendulum as case study The first test was aimed to observe the quality of the time series recordings of the developed platform
26. t the signal acquisition process by sending a command to the Coordinator that will signal the Sensing Nodes for start sampling through a beacon frame All Sensing Nodes trigger the SABs and re synchronize them at every beacon 4 Stopped The user sends a command to the Coordinator to stop the data acquisition process Again the Coordinator signals the network using its beacon at the beginning of the next Superframe All the nodes stop the data acquisition process when the beacon embedding this command is received The sampled data is stored in the SABs memory until the respective node is polled by the Coordinator 5 Transmitting After signalling the Stop state for the network the Coordinator initiates the Transmitting state by pooling a Sensing Node at a time for data Every message payload embeds 8 samples which are relayed to the C amp C App upon reception by the Coordinator 6 Done All Sensing Nodes signal the Coordinator upon completion of the Transmit state When the last Sensing Node informs the Coordinator that there is no more data to send the Coordinator enters the Done state 4 3 Coordinator node The Coordinator node is responsible for synchronizing the network and managing the application It also serves as a sink to the sampled data sent by the Sensing Nodes Such data is immediately forwarded to the C amp C App without any processing for later analysis The Coordinator supports two types of commands 1 Boar
27. tural monitoring system for bridges and buildings can consume over 75 of the total testing time and the installation labour costs can approach well over 25 of the total system cost Lynch et al 2000 These installation time and device costs can be greatly reduced via Miuicro Electro Mechanical Systems MEMS based sensors integrated in Wireless Sensors Networks WSN In this line the recent years have witnessed an increasing interest in a new technology based on WSN platforms as a low cost alternative for being applied in civil engineering structures Lynch et al 2006 Previous work from the same team collaboration between the CISTER and the ISISE research units focused on a SHM system strictly based on commercial off the shelf COTS technologies This enabled a preliminarily demonstration of the applicability of MEMS WSN based systems for operational modal analysis of structures Aguilar et al 2010 Such work allowed identifying three major limitations 1 the lack of enough sensitivity of the acceleration sensors 2 low resolution of the Analogue to Digital Converter ADC embedded in the WSN platform and 3 the lack of synchronization algorithms The SHM system illustrated in this paper solves the limitations from our previous work and blends both the advantages of using COTS and customized hardware and software technologies Importantly the proposed system architecture aims not only to respond to the application scena
Download Pdf Manuals
Related Search