Hardware-In-The-Loop - Mechanical Engineering
Contents
1. 3 4 5 6 7 8 9 10 11 122. 52 49 47 aaa eal 0 LGRN lt m S e p o m m 1 Z Lenzi Fig 7 2070 Controller Output Configurator B 2070 Controller Output Configurator The 2070 controller output configuration application was developed to match the URMS configuration convention The diagram used in the configurator to show the output file current assignments see Figure 7 is the same as that used in 1 to describe the physical output number for each output file slot for a Model 334 cabinet An output configurator was implemented in order to map every phase indication used in the simulation to the corresponding phase in the 2070 controller as recommended in 5 In this configuration application it is possible to independently read the phase of each active output When this feature is used in conjunction with the URMS Output File Test Output Signal Test and or Lights Test utilities it is easy to check if the simulation and the URMS output signal assignments match and if the phase states are read properly by the simulation C Freeway Simulator The constant velocity microscopic mainline traffic model described earlier is implemented in the freeway simulator application There are three components associated with this part of the simulation 1 the freeway simulation user interface shown in Figure 8 2 a freeway layout menu see Figure 4 left and 3 a vehicle menu The simulation interface is used to observe the movement of
3. University of California Berkeley horowitz berkeley edu P Varaiya is a Professor at Department of Electrical Engineering Uni versity of California Berkeley varaiya eecs berkeley edu dual loop detectors it will be possible to estimate the queue length 2 The second method estimates the queue length using a vehicle re identification algorithm 3 4 This scheme is based on matching individual vehicle signatures obtained from Sensys wireless sensor arrays placed at the two ends of the on ramp Before deploying a 2070 controller with a modified URMS in the Hegenberger on ramp it must be thoroughly de bugged In addition the modified software must be tested and approved by D4 engineers before it can be used on the field The unmodified URMS software has already been debugged and tested by Caltrans engineers before its release for preliminary testing in the field using a traditional traffic controller suitcase tester device see Figure 1 However one of the main drawbacks of this tester is the need to manually operate mechanical switches to simulate loop detector sig nals This debugging and testing approach becomes difficult and sometimes inappropriate when coordination of signal actuation is required as will be the case for the field test To debug and test the modified URMS it will be necessary to recreate the dual detector signals used to measure vehicle speed upstream of the on ramp with good accuracy For this reason it wa
4. Hardware In The Loop On ramp Simulation Tool to Debug and Test the Universal Ramp Metering Software Rene O Sanchez Roberto Horowitz and Pravin Varaiya Abstract An on ramp simulation system that can be used to debug and test the Universal Ramp Metering Software URMS is presented The tool includes a simple car following microscopic traffic model for the on ramp and a Controller Interface Device CID which interfaces a standard personal computer with a 2070 traffic controller The CID consists of the low cost and commonly available National Instruments NI USB 6501 24 Channel Digital I O device and a basic circuit that interfaces the 5 Volt TTL logic from the Digital I O board to the 2070 controller The resulting hardware in the loop simulation tool systematically reads the phase states from the controller and changes detector states based on the cars trajectories as displayed on the on ramp simulator With this tool it is possible to check the performance of the 2070 controller running the URMS as if the traffic controller was operating on a standard on ramp managed by Caltrans Finally the real time nature of this tool is discussed based on a quantitative analysis of the simulator performance running on the Windows XP operating system I INTRODUCTION A hardware in the loop simulation HILS system was developed as a tool to assist in the completion of a ramp metering field test This field test has been proposed in order to imple
5. W standard functions that can easily be accessed from the simulation The custom made circuit was designed to interface the 5 Volt TTL logic from the digital I O board to the 2070 con troller This circuit was built using a modular IC breadboard socket SN706 TTL hex inverter buffers drivers with open collector high voltage outputs one 7805A voltage regulator and a 12 Volt power supply Two main goals of the CID design stage were portability and low cost The portability was ensured with the use of a small USB DIO board that can be used in most personal computers The low cost was achieved by using one of the cheapest data acquisition boards on the market The components to build the CID presented in this paper cost less than 200 U S dollars B Software Interface Module The software interface model provides the linkage between the CID and the traffic simulation program The NI USB 6501 board used to build the CID comes with drivers that can be used to develop customized applications using NI LabVIEW These drivers serve as the software interface module and do not require any modification when used in the HILS tool C Microscopic Simulation Engine The simulation engine was developed using the NI Lab VIEW development environment Before deciding to create a custom traffic microscopic simulator commercial simulator packages were considered However the time steps used in these simulators were not low enough for the resolution d
6. ainline loop detectors actuation B On ramp Model A simplified car following traffic model based on 11 was used to simulate vehicles on the on ramp This is a simple model specifically conceived for a homogeneous highway in which the nth vehicle follows the same trajectory as the n 1 th vehicle except for a translation in space and time It was necessary to incorporate the ramp metering traffic signal into the model which can be considered as an inhomogeneity by specifying rules of how vehicles react to the signal The rules that were specified are 1 a car in front of the traffic light must stop when the light is red and 2 only a predetermined number of cars can advance per green phase It was decided to use this model because it is simple but captures dynamics that are important for an accurate generation of detector signals There is a particular interest in testing algorithms that use vehicle speed close to the on ramp entrance to estimate queue length This model allows for changes in speed based on driver behavior parameters and the presence of vehicles ahead With this model it is also possible to introduce queue dynamics in the simulation a feature necessary for the accurate generation and timing of on ramp detector signals For the simulation it is necessary to know parameters related to the length of the on ramp the length of the loop detectors and the position of the loop detector with respect to the ramp entrance All
7. ccupancy and speed The HILS system presented in this paper has three basic components 1 a controller interface device CID 2 a software interface module and 3 a microscopic simulation engine A description of each component is presented below a Computer 2070 Controller Controller Interface Device CID Detector Detector States Phase States Phase States b Computer running Simulation Tool Fig 2 a On ramp simulation tool architecture b On ramp simulation tool setup A Computer Interface Device CID This device provides the interface from the 2070 traffic controller to the personal computer running the traffic simu lation The CID has two main elements 1 the NI USB 6501 device and 2 a custom electronic circuit The NI USB 6501 is a portable digital Input Output de vice which provides data acquisition and control capabilities With plug and play USB connectivity the NI USB 6501 is very versatile and can be used in most personal computers The NI USB 6501 has 24 single ended digital lines which comprise three DIO ports In this tool two ports are con figured to generate detector signals and one port is used to read the phase states output from the controller This device was chosen because of its low price portability and because when used with LabVIEW it provides a straightforward procedure to interface with the simulation engine Signals can be sent and received using LabVIE
8. e sired for this application e g CORSIM uses a second time step while VISSIM can not go lower than 100 milliseconds This limitation was one reason for developing a microscopic simulation specifically for an on ramp freeway system with a time step between and 10 milliseconds Another reason was to have the flexibility to customize the simulation engine to complement some features of the URMS e g configuration and testing menus Q O D d o LO On ramp Deman Traffic Light State JJ 10 A data available from D4 available from 2070 Controller o 9 A ae Legend T1 L1 L Leading Detector A 95 1 cal e Sasi cenit Boe et ig eee eee ees ee eee Bie Pessage Delecioi eree a a E L3 D Demand Detector T4 L4 LO On ramp Leading Detector TO On ramp Trailing Detector Q Queue Detector amama Detection Station data available from PeMS Fig 3 a Hegenberger Rd loop on ramp to 880 southbound b Hegenberger on ramp mainline layout used for the simulation tool HI THE MODEL In order to simulate the Hegenberger on ramp freeway system see Figure 3 a it was necessary to use a simplified layout that would capture the detector location and the ramp characteristics Figure 3 b shows a simplified configuration of the Hegenberger onramp freeway system that follows the NTCIP typical on ramp layout as close as possible 8 which is also the standard configuration used in t
9. e 2070 controller Using this interface it is also possible to set and modify simulation parameters In the on ramp layout menu the dimension of the on ramp segment the detectors location their length and their separation can be set Finally the vehicle menu is used to determine the properties of the three types of vehicles present in the on ramp simulator S Simulator v1 1 vi BEX Eile Edit View Project Operate Tools Window Help Fearn elu We On ramp Simulator Freeway Simulator 2070 Input Configurator 2070 Output Configurator Debugging Tools On ramp Simulator On ramp Layout VehicleType Driver Behavior ON RAMP SIMULATOR Traffic Light Information Traffic Light Ramp Parameters Time Interval J 30 Seconds Loop Detectors Traffic Light Demand veh TI Assignment j Pin Assignment J16 RM Rate Veh TI t 15 Dual Trailing Dual Leading Red Lower Green Trucks on OL1 f RED f LGRNI Sf f n Qi sh ae Yellow GRN1 ai vea wf M7 NATIONAL p INSTRUMENTS LabVIEW Student Edition Fig 9 On ramp simulator interface V BENCHMARKING This tool was designed to simulate traffic conditions on an on ramp freeway system and update vehicle positions and detector states in real time In the context of this project real time means that the HILS system should simulate the displacement of vehicles check if the vehicles are on a detection zone and update detector
10. ent environment This software is composed of four elements 1 the 2070 controller input configurator 2 the 2070 controller output configurator 3 a freeway simulator and 4 an on ramp simulator When the program is run the user can decide if any configurator will be used If the configuration process for the inputs or outputs is skipped the configuration stored in the computer will be used by the program After the configuration process is completed or bypassed the on ramp and freeway simulations start In the following each component of the software is described S simulator vitvi Eile Edit View Project Operate Tools Window Help zj gt Een ga On ramp Simulator Freeway Simulator 2070 Input Configurator 2070 Output Configurator Debugging Tools 2070 Controller Input Configurator Breadboard Cable Layout 2j est ape ets e o a az Ap C1 46 C1 50 C1 49 C155 C151 C1 57 C159 C1 61 C181 C1 79 C153 C1 41 C143 C S JK JE gt gt Diu l 11 10 16 12 18 20 22 46 44 14 2 6 S ot AY Nc SA nc By mca Sy m3 By ra NC OY Nc BY Nc Ry UNC Sy 2 By o2 Sy oF C1 47 C1 48 C1 56 C1 52 C1 58 C1 60 C1 62 C1 80 C1 82 C1 54 C1 40 lt a ap gt gt lt Cc gt lt cs a 9 T7 13 19 21 23 45 47 15 z Pi B e i BBY nc M Ma Mr By Nc RSI Nc A ie KA EA Fig 6 2070 Controller In
11. ffect the aggregate mainline speed since it is calculated as an average over a URMS calculation interval For any practical purposes the 1 5 mph uncertainty will not affect the velocity estimation on the on ramp At Ataverage oO Atmax Atactual At1 1 ms 2 00 0 506 17 97 63 2 ms 2 09 0 465 17 98 99 3 ms 3 26 0 553 17 99 20 4 ms 4 19 0 426 17 99 60 5 ms 5 20 0 492 18 99 69 10 ms 10 25 0 476 23 99 76 25 ms 25 34 0 478 30 99 96 50 ms 50 93 0 271 56 99 91 TABLE I BENCHMARKING RESULTS OF 60 000 SIMULATION STEP RUNS FOR DIFFERENT Af V max tactuation tcontroller V controller On ramp 40 mph 408 32 ms 408 32 16 ms 40 1 5 mph Freeway 80 mph 204 16 ms 204 16 16 ms 80 6 4 mph TABLE II SIMULATED LOOP DETECTOR SIGNAL UNCERTAINTY FOR At 2 ms Lietector 1 8 m and Legr 5 5 m VI CONCLUSION This paper presented a hardware in the loop on ramp eval uation system for the URMS which consists of a personal computer running an on ramp microscopic traffic model a CID and a 2070 controller running the URMS The system was developed to assist in the debugging and testing process involved with the release of a URMS version for deployment in the field Since this tool was specifically tailored for a 2070 controller running the URMS it allows for an easy configuration of the system and a user friendly interface that matches or complements some of the URMS debugging features The paper also presented an analysis of the real time nat
12. he URMS The ramp layout had to be slightly modified to incorporate dual detection for queue length estimation A traffic controller operating on an on ramp in California is usually programmed to collect data from the on ramp de tectors set the traffic light phase states and collect mainline detection stations data sometimes multiple mainline detec tion stations In order for the simulation tool to generate the signals that a traffic controller would encounter in the field it was decided to simulate traffic conditions on the Hegenberger on ramp mainline system with two completely different models 1 a constant velocity microscopic mainline traffic model and 2 a simplified car following on ramp traffic model The 2070 controller is able to read the detector states set by both models and can update the on ramp simulation metering rate phase states With this approach on ramp traffic conditions do not have any effect on the mainline freeway simulation However simulated mainline traffic conditions may have an effect on the on ramp simulation depending on the ramp metering algorithms implemented in the URMS The simulation tool was designed in this way in order to test traffic responsive ramp metering algorithms like ALINEA 9 where mainline traffic conditions read by the controller are used to set the metering rate at the on ramp It should be noted that this tool only allows testing the open loop behavior of ramp metering algorithms as
13. ment queue control on the Hegenberger Rd loop on ramp to 880 southbound in the Caltrans Bay Area District D4 to study its effect in minimizing queue and mainline density oscillations and enhancing performance This will be accomplished by using a 2070 traffic controller running a modified Universal Ramp Metering Software URMS which is a recently developed program that allows the 2070 traffic controller to function as a ramp metering controller for use throughout California 1 To prevent on ramp queues from spilling over into surface streets and interfering with the street traffic the queue length must be regulated If the queue length could be measured an asymptotically stable PI regulator can be designed to stabilize the closed loop queue dynamics 2 However the PI regula tor needs the current queue length as its feedback signal which unfortunately is not available in the field For the field test two different queue length estimation methods will be evaluated The first method is a queue length estimator based on a simplified model for the driving behavior of a vehicle that is approaching the end of the queue the vehicle decelerates at a constant rate from its cruising speed until it stops By measuring speed upstream of the on ramp using R O Sanchez is with the Department of Mechanical Engineering University of California Berkeley r2sanche me berkeley edu R Horowitz is a Professor at Department of Mechanical Engineering
14. on time Atactua occur Offsets introduce by having Atactua At at the ith simulation step can be removed at the i 1 th simulation step In order to show that the HILS system is a reliable tool to debug and test the URMS it was important to quantify the uncertainty introduced by not developing this tool on a real time operating system environment As a result a benchmarking procedure was used to characterize the real time nature of the software 8 simulation runs of 60 000 simulation steps each were executed using different desired time steps At The actual time step Atgeryqi Was recorded for each simulation step and stored into a file These data were used to determine the time reliability of this tool as a function of At The results of the analysis are presented in Table I and show that Atactual for at least 97 5 of the steps is within one millisecond of At The worst performance is observed when At 1 ms As At increases the percentage of Atactual that are within one ms of At increases while the standard deviation decreases Based on Table I and Figure 10 choosing At 2 ms provides the time resolution needed for the HILS tool while introducing an acceptable error on the generation of the detector signals In order to quantify the effect of time uncertainty on the detector signals a worst case scenario analysis for At 2 ms was performed The shortest signals generated by the simulator which are also the most affected b
15. put Configurator A 2070 Controller Input Configurator The 2070 controller input configuration application was developed to match the URMS configuration convention The diagram used in the configurator to show the Input File current assignments see Figure 6 is the same as that used in 1 to describe the physical input number for each input file slot for a Model 334 cabinet An input configurator was included in this tool because it is necessary to map every detector used in the simulation to the corresponding detector in the 2070 controller as recommended in 5 In this configuration application it is possible to independently change the state of each active input When this feature is used in conjunction with the URMS Input File Test utility it is straightforward to check if the simulation and the URMS signal assignments match and if the detector states are read properly by the 2070 controller amp Simulator v1 1 vi BEX File Edit View Project Operate Tools Window Help gt een a On ramp Simulator Freeway Simulator 2070 Input Configurator 2070 Output Configurator Debugging Tools lt D I TEHES TEE al o m N wo D D 2 m m Z H wf vf 1 R AIN A l 9 10 11 12 13 14 C1 97 C1 94 C1 91 C1 40 C1 88 C1 85 C1 33 53 50 48 Fey C1 98 C1 95 C1 101 C1 89 C1 86 C1 100 ze C1 99 C1 96 C1 93 C1 90 C1 87 C1 84 55
16. s decided to design and build a hardware in the loop simulation system to replicate in real time the Hegenberger on ramp detector signals Fig 1 Traffic controller suitcase tester used to evaluate the URMS Il HARDWARE IN THE LOOP SIMULATION The hardware in the loop simulation HILS concept has been used to create a simulation tool to test the URMS running on a 2070 controller A particular feature of this type of architecture is that the traffic simulation model does not implement any control logic instead it controls traffic flow in the simulation based on the phase states produced by the traffic signal control equipment Simultaneously the traffic signal control equipment uses the detector signals generated by the simulation to update its control logic see Figure 2 a 5 HILS has been used in the past to interface with traffic signal control equipment for testing purposes however previous systems focused on testing intersection control software The simulation time step used in these systems is on the order of seconds and equipment is used primarily to simulate loop detector signals used by traffic controllers to determine car presence and a rough estimate of occupancy 6 7 The tool presented in this paper is primarily designed to generate through simulation traffic detector signals for an on ramp freeway system see Figure 3 b with sufficient resolution to allow the 2070 controller to accurately calculate volume o
17. states with a time equal or less than the actual time it would take vehicles to travel the same displacement on a real on ramp freeway system Furthermore it is desired to achieve the smallest possible simulation time step At in order to increase the resolution of the detector signals sent to the 2070 controller The real time nature of the hardware in the loop simu lation HILS tool is limited by the performance of the Windows XP operating system which only permits a 1 ms time resolution Even though an actual HILS simulation step Atactual may take less than 1 ms this time is usually larger since Windows XP does not have sufficient real time capability to effectively implement such precise timing 5 To compensate for this limitation the simulation was designed so that the timing of the simulator would be based on three time stamps 1 a reference time stamp obtained at the beginning of the simulation run t 2 a time stamp recorded in the i 1 th simulation step t _ and 3 a time stamp obtained in the current ith simulation step t To update any quantity that needs the total simulation time the difference between t and t is used For quantities that need the time increment between the i 1 th and the ith simulation steps e g to calculate position increments the difference between t and t _ is used This configuration helps maintain accurate simulation timing even when variations in the actual simulation step executi
18. there is no interaction of the on ramp and the mainline model Fig 4 left Freeway layout right On ramp layout A Freeway Mainline Model A constant velocity microscopic mainline traffic model was used to model vehicle trajectories on the mainline In this model the cars of a given freeway lane travel at the same speed and their position is updated every simulation period The car trajectories start at the beginning of the freeway segment and end when the car reaches the end of the freeway segment given by the user specified freeway length see Figure 4 left The cars are generated based on the flow specified for every calculation interval The URMS software calculates aggregates of mainline data every 30 seconds For this reason parameters for a given lane can be updated every 30 second calculation interval in the simulation However the calculation interval should be set taking the data used to feed the simulator into account For example if PeMS 10 data are used the calculation interval should be set equal to the time granularity used in the data set The model for the freeway can be very simple because the main objective of this part of the tool is to generate loop detector signals that the controller can read to calculate aggregate values for each calculation interval and use these aggregate values as the input for traffic responsive ramp metering controllers Fig 5 M
19. these parameters can be specified using the simulator on ramp layout menu as shown in Figure 4 right For the Hegenberger on ramp these parameters were obtained from Goolge Earth and 12 C Vehicle Loop Interaction The loop detector signals generated by the simulation tool are actuated based on vehicle positions Both simulations have the location of the loop detectors with respect to the beginning of the freeway segment and the beginning of the on ramp respectively When any of the data points representing a vehicle is on the detection zone specified by the location of its leading and trailing edge the detector signal is triggered The interaction between loop detectors and cars occurs in real time Whenever the display in the simulator shows an active detector the detector signal read by the controller for that specific detector is active as well see Figure 5 This is a desired feature for a debugging tool since it helps to visually identify what is the state of each signal going into the controller In order to recreate the signals generated in a real on ramp freeway system more realistically three types of vehi cles can be generated in the simulator 1 cars 2 pickups and 3 trucks Each vehicle has an independent length shape and probability of occurrence The particular shape of each car can be observed in Figure 5 IV SIMULATION TOOL The simulation software was developed using the NI Lab VIEW developm
20. ure of the simulator that shows that for all practical purposes Windows XP limited real time capabilities do not affect the performance of the HILS tool Future research involves adding more versatility to the tool and enabling communication with the controller in order to increase syn chronization Symbol Name Unit At desired simulation time step ms Ataverage average simulation execution time ms A eee actual simulation execution time ms Ate maximum recorded time step ms oO standard deviation ms V mnax maximum velocity used in simulation mph Ldetector loop detector length m Lear average regular car length m tocrianon theoretical detector actuation time based on V max ms tcontroller detector actuation time recorded by 2070 controller ms Veontroller Velocity calculated by 2070 controller mph TABLE III LIST OF SYMBOLS REFERENCES D Wells and E Torizno URMS Universal Ramp Metering Software User Manual tech rep Traffic Operations Caltrans Version 1 07 2008 X Sun Modeling Estimation and Control of Freeway Traffic PhD thesis University of California Berkeley 2005 K Kwong R Kavaler R Rajagopal and P Varaiya A practi cal scheme for arterial travel time estimation based on vehicle re identification using wireless sensors in Transportation Research Board 89th Annual Meeting 2009 K Kwong R Kavaler R Rajagopal and P Varaiya Arterial travel time estimation based on vehicle re identification
21. using wireless sen sors Submited for publication to TRC D Bullock B Johnson R B Wells M Kyte and Z Li Hardware in the loop simulation Transportation Research Part C Emerging Technologies vol 12 no 1 pp 73 89 2004 Z Li B Johnson A Abdel Rahim and M Kyte Hardware and software design of an automated testing tool for traffic controllers in Intelligent Transportation Systems Conference 2006 ITSC 06 IEEE pp 1525 1530 Sept 2006 E Kwon S Kim and T M Kwon Pseudo real time evaluation of adaptive traffic control strategies using hardware in loop simulation in Industrial Electronics Society 2001 IECON 01 The 27th Annual Conference of the IEEE vol 3 pp 1910 1914 vol 3 2001 NTCIP 1207 2001 v01 17 National Transportation Communications for ITS Protocol Object Definition for Ramp Meter Control RMC units tech rep ASSHTO ITE NEMA 2001 H Hadj Salem J M Blosseville and M Papageorgiou Alinea a local feedback control law for on ramp metering a real life study pp 194 198 May 1990 PeMS PeMS website url http pems eecs berkeley edu accessed 2 3 2009 2009 G F Newell A simplified car following theory a lower order model Transportation Research Part B Methodological vol 36 no 3 pp 195 205 2002 T O Program Ramp Meter Design Manual tech rep California Department of Transportation 2000 1 2
22. vehicles through the defined freeway segment With this interface information related to the simulation can be accessed and it is possible to set and modify freeway lane parameters independently In the freeway layout menu the dimension of the mainline segment the detector location and the detector separation can be set Finally the vehicle menu is used to determine the properties of the three types of vehicles present in the simulation The vehicle parameters for the mainline simulation can be different from those used on the on ramp simulator Freeway Simulator Freeway Layout Vehicle Type MULTIPLE LANE FREEWAY SIMULATOR Simulation Info Simulation Info Lane 2 Parameters Lane 3 Parameters Simulation Info Simulation Info Lane 1 Parameters 144271 Parameters Parameters 2 low Veh TI Speed mph Wy NATIONAL p INSTRUMENTS LabVIEW Student Edition Fig 8 Freeway simulator interface D On ramp Simulator The simplified car following traffic model described earlier is implemented in the on ramp simulator There are three components associated with this application 1 the on ramp simulation user interface shown in Figure 9 2 a freeway layout menu see Figure 4 right and 3 a vehicle menu The simulation interface is used to observe the movement of vehicles through the on ramp This component also displays information related to the simulation including the phase state output by th
23. y the time uncertainty are those of the smaller cars traveling over a loop detector at the speed limit There is a speed limit specified for the freeway and one for the on ramp Assuming that D D o nD pS D oO 5 i E a a o N b D o 5 D R a a o N U ha T T T T x Number of Iterations o N A T T x f Io N i nN A f x Number of Iterations 3 4 5 6 7 8 9 10 11 12 138 14 15 16 17 4 5 6 7 8 9 10 11 12 13 14 15 16 17 x10 5 ms x10 10ms r r r D D S b N N Number of Iterations o 5 6 7 8 9 10 11 12 13 14 15 16 17 18 10 11 12 13 14 15 16 17 18 19 20 21 22 23 10 50 ms Ab 4 rae 4 0 5 so 55 56 Number of Iterations Oo N gt D 562 53 54 Execution Time ms Execution Time ms Fig 10 Histograms of Atactua for multiple simulation runs 60 000 iterations each using different At Atactual Atmax the propagation of the maximum error of 16 ms is shown in Table I Based on the data collected from the benchmarking procedure the uncertainty propagated to detector signals generated by the simulators is within 16 ms which when used by the 2070 controller to calculate velocity would yield a 6 4 mph uncertainty for the freeway simulator and a 1 5 mph uncertainty for the on ramp simulator The 6 4 mph uncertainty in the mainline speed computations may seem significant However it will not considerably a
Download Pdf Manuals
Related Search