Development Of An Electric Vehicle Hardware-in-the-Loop .
Proceedings of the ASME 2013 International Design Engineering Technical Conferences andComputers and Information in Engineering ConferenceIDETC/CIE 2013August 4-7, 2013, Portland, Oregon, USADETC2013-12263DEVELOPMENT OF AN ELECTRIC VEHICLE HARDWARE-IN-THE-LOOPEMULATION PLATFORMSara MohonClemson UniversityGreenville, SC, USAArun VenkitakrishnanClemson UniversityGreenville, SC, USABeshah AyalewClemson UniversityGreenville, SC, USAPierluigi PisuClemson UniversityGreenville, SC, USAABSTRACTINTRODUCTIONHardware-in-the-loop emulation allows for rapidtesting of new control algorithms and hardwarecomponents without the need for a full vehicleprototype. This paper outlines an electric vehicleemulation platform consisting of hardware andsimulation components carefully integrated togetherin order to emulate the operation of the vehicle’spowertrain. The main hardware components of theplatform are two AC synchronous induction motorsand their inverters/controllers, a rotary torque sensor,a battery pack, a programmable DC power supply,and a programmable DC load. The simulationcomponents consist of a real-time data acquisitionsystem with a control design and modeling interface.The control structure consists of speed and torquecontrol mode selections for the two motors, vehicleand driver simulation models, as well as afeedforward component that improved the operationof the emulation platform during a typical EV mption, pollutant emissions, and the impact onclimate change continue to stimulate broad researchand development efforts into more sustainablevehicle propulsion systems. Electric propulsionsystems are one such major solution being pursuedby large and small vehicle manufacturers alike.While the electric vehicle (EV) itself is not a newconcept, the components (energy storage, drivemotors and associated power electronics) and theassociated systems integration issues continue toevolve rapidly. To help with the assessment ofdesign options, a variety of EV modeling andsimulation tools have been developed [1-2].However, the validity of pure simulation-based toolsdepends to a large extent on prior test data thatsupport component and system modelingassumptions. One approach that reduces the need forprior test data is using available hardware within thesimulation framework. This later approach has cometo be known as Hardware-in-the-Loop (HIL)simulation.1Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/28/2015 Terms of Use: http://asme.org/termsCopyright 2013 by ASME
An EV HIL platform generally consists of somehardware components from the EV class underconsideration, a real-time control and dataacquisition unit, and a computer with simulationsoftware that executes some system or subsystemmodels. The issue under investigation determineswhich hardware components will be included. Thesimulations will then retain the relevant model of therest of the system dynamics. The interface betweenthe simulation and the hardware components ishandled with suitable real-time control and dataacquisition units.In this paper, we focus on describing a HILplatform intended for emulating the drivetrain of aneighborhood electric vehicle (NEV), but the setupcan be scaled-up for larger EVs. The platform hasbeen assembled using mainly off-the-shelf EVdrivetrain components, namely, an AC machine andits inverter/controller, and the battery pack. Thepropulsion power is absorbed by another ACmachine with its own inverter/controller and isconnected to a programmable DC load and powersupply. This configuration is analogous to a typicaldynamometer set up. Using a carefully structuredcontrol strategy, commands are sent to the twoinverter/controllers via the real-time control unitwhile the vehicle’s forward dynamics aresimultaneously simulated.In the literature, one can find similar EVemulation platforms built and analyzed for specificpurposes. An EV HIL platform built at University ofBrest in France [3] used an AC motor as theemulated EV in torque control mode and a DC motoras a chassis dynamometer in PWM and hysteresiscontrol modes. The authors showed that the DCmotor in hysteresis control mode yielded bettercurrent tracking over PWM control mode because nospecific PID gains were necessary for the AC motorcontrol.Another EV HIL platform was built at Universityof Maribor in Slovenia [4] that used an AC motor asthe EV in torque control mode and a DC motor as achassis dynamometer in torque control mode. Theobjective was to implement a control strategy thatused no speed sensors because they were consideredimpractical sensors for a vehicle environment.Instead, a speed observer was designed and slidingmode control was used to effectively transition intothe field-weakening region of the AC motorcontroller.A more complex hybrid EV HIL platform wasbuilt at Korea University of Technology andEducation in South Korea [5]. The platform usedone DC axial flux motor (AFM) with variable air gapin torque control mode as the vehicle. Another DCAFM was used in speed control mode as adynamometer to load the vehicle. This platform wasused with two different simulation programs toemulate a series hybrid EV with four AFMs in thevehicle wheels and a parallel hybrid EV with batteryand fuel cell energy sources with two AFMs in thetwo front wheels.A mechanical load emulation platform similar tothe one we describe in this paper was built atUniversity of Nottingham in England [6] in 1999.The platform consisted of two very small (0.55kW)AC induction motors coupled by a torque sensor.One motor acted as a dynamometer and the other asthe electric drive but both were controlled withvector control. The authors presented a feedforwardspeed-tracking control strategy for the dynamometerthat can be used for a wide range of nonlinearmechanical loads. This work was improved upon in[7] by using nonlinear control methods.In this paper, we describe a HIL EV emulationplatform design, which has similarities in intent andpurpose to the above reviewed works. The focuswill be less on the low-level control of the individualmachines and more on the system integration ofreadily available components that can be used foreducational and research purposes. In building theplatform, we used off-the-shelf AC machines andtheir inverters/controllers, with limited knowledge ofthe low-level control techniques other thanController Area Network (CAN) and otherinterfacing definitions. Typical uses of the platforminclude: 1) efficiency mapping of motors,controllers, and battery packs, 2) analyzing howvehicle design parameters (including mass, dragcoefficients, etc.) affect performance such asregenerative braking capabilities, and 3) evaluatingvarious driver models for analyzing the effects of2Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/28/2015 Terms of Use: http://asme.org/termsCopyright 2013 by ASME
driver behavior on the overall energy efficiency ofthe EV drive.The remainder of the paper is organized asfollows. First the hardware layout is described.Then the HIL platform control structure and thesimulation models are discussed. This is followedby some experimental results and discussions on theplatform control strategies considering a typicalurban driving cycle. Then an experimental analysisthat demonstrates one typical application of the HILplatform is presented. Finally, some conclusions areprovided.PLATFORM HARDWARE/SOFTWAREThe EV emulation platform presented in thispaper consists of readily available hardware andsimulation tools. Figure 1 shows a photo of theplatform. It consist of two 13kW 3-phase ACsynchronous induction machines from HPEVS [8],two 1236 AC motor controllers/inverters from CurtisInstruments [9], a 100Nm torque sensor fromInterface [10], a 48 volt battery pack from Rayovac,a 10kW programmable DC power supply fromAmrel [11], and a 6 kW programmable DC loadfrom NHR [12] (not shown in Figure 1). The realtime control and data acquisition system is dSPACEMicroAutobox (dS 1401) [13]. The simulationmodel is built in Matlab/Simulink and interfaced todSPACE ControlDesk software for visualization ofexperiments.of the drive motors is absorbed by the AC generator(similar machine operating in generator mode) andconnected to the combination of a DC load inconstant current (CC) mode and DC power supply inconstant voltage (CV) mode connected in parallel.This configuration was chosen instead of another(absorber) battery pack to avoid dealing withovercharging issues and to add some testingflexibility to the setup. Both electric machines arecapable of 4-quadrant operation. A rotary torquesensor is placed between the two machines andmeasures the shaft torque, which forms one input tothe vehicle simulation model to be discussed below.The 1236 AC motor controller came with aproprietary programming software made available bythe supplier, Curtis Instruments. The softwareallows for selecting torque or speed control modesand settings, motor encoder settings, motortemperature settings, and even battery specificationsettings. For data acquisition and recording of motorspeed (RPM), current, voltage signals duringexperiments, the controller can be configured tooutput CAN messages for monitoring desiredvariables. The CAN language utilized was CANopenand Curtis Instruments provided the CANopenidentifiers for all the variables inside the controller(such as motor RPM, bus voltage, current, etc.).Special CANopen code was written in Simulink sothat the dSPACE data acquisition system could readthe RPM, current and other data during experiments.CONTROL AND SIMULATION STRUCTUREFigure 1: Hardware for the HIL platformThe EV is emulated by the AC motor on the left(Figure 1) powered by the battery pack. The outputFigure 2 shows the schematic of the control andsimulation structure for the whole HIL platform.The solid red lines denote power lines. The dashedred line denotes the interface between the hardwareand simulation components and functions. In orderto emulate the operation of the EV, the first machine(drive motor) was set to operate in torque controlmode. This mode was selected from the availablecontrol mode options of the 1236 motorinverter/controller. The mode selection is justifiedby the fact that a driver would normally command atorque output from the EV’s motor via theaccelerator and brake pedals. The driver is modeledseparately as a PID-type controller that requeststorque commands to follow a desired drive cycle3Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/28/2015 Terms of Use: http://asme.org/termsCopyright 2013 by ASME
over time. The driver torque command is sent to thefirst motor controller, which accelerates or brakesaccordingly. The motor controller uses DC powerinput from the battery pack but produces controlledAC output to the motor and therefore also includesan inverter.The second AC machine (generator) wasoperated in speed control mode and acts as adynamometer (absorber) for the power output of thefirst motor. The speed control mode is selected hereas the only remaining option for stable operation ofthe whole setup, as the torque control mode hasalready been selected for the first machine on thesame shaft of negligible inertia. It remains todetermine what the command speed for the generatorshould be and this is assigned to the vehicledynamics simulation part of the HIL platform. Notethat the AC generator controller also has an inverteras it connected to a DC bus.The simulation part is largely comprised of adriver model, vehicle dynamics model, andthrottle/brake control.Driver ModelThe drive cycle data provides the referencevelocity that the driver tries to follow. The driver ismodeled as a PID controller that compares thisreference drive cycle velocity with the simulatedvelocity of the hardware (generator). The PIDcontroller output is the driver torque command. Theparameters of the PID controller can be changed torepresent different kinds of drivers. The PIDparameters used for this paper are given in theAppendix.Figure 2: HIL platform with hardware and simulation componentsVehicle Dynamics SimulationIn this paper, the vehicle dynamics simulationessentially integrates the following forwarddynamics equation:dv1m Ftraction fr mg cosθ ρ ACd v 2 mgsin θdt2(1)The torque sensor emulates the traction torque atthe wheel of the EV when ignoring the small mass ofthe torque sensor and couplings. Given the real-timemeasurement, and known vehicle parameters such astire radius (R), operating gear ratio (i), and otherroad load parameters, the forward dynamics of thevehicle can be simulated. The road loads to be4Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/28/2015 Terms of Use: http://asme.org/termsCopyright 2013 by ASME
subtracted from the traction torque include therolling, aerodynamic and grade resistances.Given this simulated forward speed of thevehicle, the commanded rotational speed can becomputed and sent to the generator’s controller,which has been set in speed control mode.Throttle and Brake ControlThe motor and generator controllers receiveanalog wired throttle and brake inputs as commandsignals. The motor can receive throttle or brakecommands in torque control mode. The controllerconfiguration of the generator is such that it canthrottle but not brake in speed control mode (i.e., forthe generator, the speed commands are received asthrottle only). The common shaft can slow down bythe generator throttling less than the motor, themotor throttling less than the generator, or by themotor braking.Figure 3 shows the scheme for the throttle andbrake logic as implemented on the simulation side ofthe platform. If the reference velocity minussimulated velocity is greater than zero, then themotor accelerates at the torque value given by thedriver PID block. Otherwise, the motor deceleratesfor two seconds. If the error is still positive after twoseconds, then the motor brake is applied. The motorbrake is applied gradually using an increasing ramp.When the error is negative then the motor begins tothrottle at the new desired torque. These adjustmentshelped ensure smooth operation of the platform.Figure 3: General throttle and brake logicRESULTS AND DISCUSSIONSVelocity TrackingThe hardware components used in the HILplatform are sized for NEVs, golf carts, etc. In orderto test the emulation platform, a drive cycle that ourspecific platform could achieve was needed. Ascaled-down version of the Japanese 10-15 modedrive cycle was selected to represent a typical stopand-go cycle with significant acceleration andbraking events. The rest of the vehicle and driverparameters used in the HIL EV emulations areprovided in the Appendix.Figure 4 shows the tracking performance for theemulated EV with a nominal mass of 150 kg. It canbe seen that the generator speed (actual RPMconverted to kph) tracks the simulated speed ratherwell, apart from measurement noise. The drivermodel also tracks the reference profile well with atracking performance that improves with timeaccording to the integral action of the PID model forthe driver. The corresponding motor and generatorcurrent profiles are shown in Figure 5. When the5Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/28/2015 Terms of Use: http://asme.org/termsCopyright 2013 by ASME
motor current becomes negative, current is being putback into the motor’s battery pack. This regenerativebraking current occurs when the vehicle deceleratesduring the drive cycle.Figure 4: HIL velocity profilesFigure 6: Zoomed-in view of first peaks with nofeedforward torqueThe feedforward is achieved by using the knownreference velocity and the known parameters of thevehicle. We simply use the discretized form given inEq (2), where V(t) stands for the reference velocityat the current time t and Δt is a time step selected toapproximate the derivative. The feedforward schemeis added to the simulation side of the HIL platformas shown in Figure 7.T (t) R m (V (t Δt) V (t)) 1 ρ ACdV 2 (t) fr mg i 2Δt (2)Figure 5: HIL current profilesCloser inspection of the drive cycle reveals thatthe simulated velocity lags the actual and referencevelocities as shown in Figure 6. The dynamics ofvehicle simulation block causes this lag. Thegenerator control loop also may contribute to thislag, but we assume it to be insignificant. Tocompensate for this lag, a feedforward torque can becalculated based an inversion of the forward vehicledynamics.Figure 7: Construction of feedforward torqueBest results were achieved with feedforward forΔt 1 second. Figure 8 shows improved ted. The lag of the simulated (actual)velocity from the reference has been significantlyreduced. The steady state errors could not beimproved much further without destabilizing theplatform with the current controller selections.6Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/28/2015 Terms of Use: http://asme.org/termsCopyright 2013 by ASME
under the feedforward.These results are asexpected. The motor draws more battery current toaccelerate a heavier vehicle resulting in increasedconsumed energy. The motor must also brake moreoften to achieve desired deceleration of the heaviervehicle resulting in increased regen energy.Figure 8: Zoomed-in view of first peaks withfeedforward torqueVelocity tracking performance of the nofeedforward case and the feedforward case werecompared using mean error and peak error betweenreference velocity and simulated velocity for thedrive cycle.The feedforward case showedimprovement in tracking performance. The resultsare given in Table 1.Mean Error Peak Error(kph)(kph)No Feedforward1.175.41With Feedforward1.124.55Table 1: Velocity tracking performanceAPPLICATION EXAMPLEEffect of Vehicle Mass on Energy ConsiderationsAs already stated, one application of the EVemulation platform is to study the sensitivity ofperformance metrics to vehicle design choices. Asan example, the HIL platform was used toinvestigate energy output and energy recovery abilityof the EV for different vehicle mass values. TheEV’s motor ability was evaluated using thefollowing metrics for the whole drive cycle: totalregen energy, total energy consumed from batterypack, and the regen ratio for vehicle masses rangingbetween 150 kg to 400 kg. The regen ratio will bedefined as the ratio of total regen energy over totalenergy consumed from battery pack. Table 2 showsthe performance results for feedforward case.The total regen energy, total consumed energy,and regen ratio increases as vehicle mass increasesTotalRegenEnergy (kJ)TotalConsumedEnergy (kJ)150 kg250 kg325 kg400 .836.628.41RegenRatio (%)Table 2: Select EV performance metrics withchanging vehicle mass for feedforward caseCONCLUSIONThe HIL EV emulation platform presented hereis built from off-the-shelf components foreducational and research purposes. This paperprovided the details on the system level integrationand the control structure for the platform. Some usesof the platform have been outlined and at least oneapplication study has been demonstrated.Some aspects of the platform can yet be refinedfurther. For example, more advanced control andcompensation methods could be sought to improvethe overall speed tracking performance, especially atsteady state. The vehicle simulation models and thedriver models could also be refined further per therequirements of the specific evaluation to beconducted on the elective vehicle.NOMENCLATURERimgaACdρtire radius (meters)gear ratiomassgravitational acceleration constantaccelerationfrontal areacoefficient of dragdensity of air7Downloaded From: http://proceedings.asmedigitalcollection.asme.org/ on 05/28/2015 Terms of Use: http://asme.org/termsCopyright 2013 by ASME
θfrFnetFtractionFRoad LoadFRRFaeroFgraderoad gradefriction coefficientnet force
Interface [10], a 48 volt battery pack from Rayovac, a 10kW programmable DC power supply from Amrel [11], and a 6 kW programmable DC load from NHR [12] (not shown in Figure 1). The real-time control and data acquisition system is dSPACE MicroAutobox (dS 1401) [13]. The simul
changing electric vehicle technology and its uncertain uptake. Better electric vehicle projections, factoring in minimum compliance with regulatory frameworks, would provide much greater certainty to plan for expanding the electric vehicle fast-charging networks. This requires analysis of future year-by-year electric vehicle penetration,
The Urban Development and Housing department of Madhya Pradesh has prepared “Madhya Pradesh Electric Vehicle (EV) Policy 2019” to accomplish the objectives of ‘Elec-tric Vehicle Initiative (EVI). Electric Vehicle Initiative (EVI), under Madhya Pradesh Electric Vehicle (EV) Policy 2019, is dedicated to accelerating the deployment of EV’s .
ASSESSMENT OF NEXT-GENERATION ELECTRIC VEHICLE TECHNOLOGIES I. INTRODUCTION The global electric vehicle market has been steadily expanding since 2010. Figure 1 shows annual electric vehicle sales globally from 2010 through 2015, with a breakdown by the 15 major automobile manufacturing groups with the greatest cumulative electric
2-wheeler, 3-wheeler and 4-wheeler Hybrid Electric Vehicles (HEV), Plug in Electric Vehicles (PHEV), Battery Electric Vehicles (BEV), and Fuel Cell Electric Vehicle (FCEV). 2.3.2. Electric Vehicle Battery refers to all energy storage systems used in the defined EVs above. This includes Lithium ion batteries, nickel metal
Research on Electric Vehicle Development Status, Policies, Trends and Issues in China . Zeyuan Lyu . Qingdao No.58 High School, China . Keywords: electric vehicle; electric engine; new energy. Abstract: Electric vehicles are a well-known technology in the 21st century. It is the most popular transportation tool for the government and citizens .
duppattas - six electric bell - two electric fans - three electric hand dryer-two electric iron - one electric lamps / wall-six electric organ-one electric shavers-two electric table lamp-two electric toaster-one empty receptacle
However, electric furnaces can accommodate central cooling easier than zonal electric heating, because the air conditioner can share the furnace’s ducts. Electric resistance heat can be provided by electric baseboard heaters, electric wall heaters, electric radiant heat, electric space heaters, electric f
Electric Vehicle Charging Micro-Climate Plan for the Greater Houston Area vii Acronyms BEV Battery Electric Vehicle—Vehicle powered 100% by the battery energy storage system available on-board the vehicle. CCID Charge Current Interrupting Device—a device within EVSE to shut off the electricity supply if
