PMSM Electric Drive System simulation for Parallel ... - IRNet Explore

PMSM Electric Drive System simulation for ParallelHybrid Electric VehicleM. Lakshmi Swarupa (1) , G. Tulasi RamDas (2) & P.V. Raj Gopal (3)(1) Assoc.Prof, Malla Reddy Engg college , (2) Vice-chancellor for JNTUK & (3) AGM, BHEL R&DE-mail : swarupamalladi@gmail.com , das_tulasiram@yahoo.co.in & pv_rajgopal@bhelrnd.co.inAbstract - As permanent magnet synchronous motor (PMSM) has the characteristics of high power density, high efficiency, and hightorque/ current units, so it is extensively used in electric vehicles/Hybrid electric vehicles driving. Energy management unit(EMU) inhybrid electric vehicle (HEV) demands for good motor driving and energy regeneration capability. Based on the mathematical modelof PMSM, this paper presents the PMSM vector control algorithm, studies HEV energy regeneration requirements for motors andanalyzes power generation work mechanism of PMSM. At last, combined with HEV control strategy, simulation of PMSM systemenergy regeneration research is done and the simulation results show that the system has good performance.Keywords- Permanent magnet synchronous machine, field-oriented control, reference model, PI, NN controllers.I. INTRODUCTIONRecently, as the world resources and environmentalissues are becoming increasingly prominent, manycountries have stepped up the development plans ofelectric vehicle (EV) [1]. Hybrid electric vehicle (HEV)and the fuel cell vehicle (FCEV) are the new focus.Motor is the core component in HEV driving. In pureelectric vehicles and fuel cell electric vehicles, it is theonly driving component. In HEV, it is the keycomponent to realize different control strategies. Motorsfor EV require high reliability, wide torque and speedscope and large power to volume ratio.The development of permanent magnetic materials,especially the rare earth material, makes the motorpower density and torque increase and the volumedecrease.PMSM is widely used in high efficiency, highpower density, high torque/current units and wide-speedoperation EV driving system. Energy management unitin HEV controls the energy flow according to the workstates of the vehicle, this requires the motor has a gooddriving system and energy regeneration capability.Researching on motor energy regeneration is of greatsignificance for improving energy efficiency andincreasing system availability. Based on the PMSMmathematical model, this paper presents the PMSMvector control algorithm, studies HEV energyregeneration requirements for motors, and analyzespower generation work mechanism of PMSM. Then thecontrol strategy of PMSM energy regeneration isproposed and simulation of the motor control system ismade .that the PMSM had its permanent magnetsmounted on the surface of the rotor. This type of PMSMhas therefore a uniform air gap and no saliency, henceLd = Lq. In the actual demo, it is assumed that thePMSM has an interior permanent magnets rotor.The impact of the buried-magnet configuration isrotor saliency that makes Lq>Ld and introduces areluctance torque term into the PMSM torque equation.To take advantage of the reluctance torque, the Idcurrent component is no longer set to zero has it is forthe PMSM with surface mounted permanent magnets.The PM Synchronous Motor Drive is composedof four main parts: The electrical motor, the Three-phaseInverter, the VECT controller and the Speed Controller.• The electrical motor is a 288 Vdc, 100 kW PMSM.This motor has 8 pole and the magnets are buried(salient rotor's type).• The Three-phase Inverter is a voltage sourceinverter, controlled by PWM. This block is builtusing the Universal Bridge Block.• The VECT controller block computes the threereference motor line currents corresponding to theflux and torque references and then generates acorresponding PWM using a three-phase currentregulator. When the nominal flux is required, anoptimal control is used in order to minimize the lineInternational Conference on Software Technology and Computer Engineering (STACE-2012), ISBN : 978-93-81693-68-1, 22 nd July 2012,Vijayawada42

PMSM Electric Drive System simulation for Parallel Hybrid Electric Vehiclecurrent amplitude for the required torque. When aflux weakening is needed, the amplitude and thephase of the current are changed to extend thetorque-speed operating range.• The Speed Controller is used in torque regulationmode. The normalized flux value is computed withthe speed of the machine in order to perform a fluxweakening control.The Torque limitation block is used to prevent thelimitation due to the torque-speed characteristic of thismotor for a 288 Vdc source. When the internalmachine's voltage reaches the inverter voltage (becausethe desired torque is too high for the motor's speed), theinverter becomes in saturation mode (the desired currentcannot flow anymore into motor). After this point, therewill be a loss of current tracking which will decrease themotor current. This block is used to reduce the referencetorque as a function of the motor’s speed and the torquespeedcharacteristic in order to never operate in invertersaturation mode.II. COMPARISON BETWEEN INDUCTIONMOTOR AND PMSMA summary of the comparison of the inverter fedim and pmsm is given below.1. Harmonic current: Harmonic current for a givenharmonic voltage is higher in the case of IM. Asresult, torque pulsations will be higher.2. Switching frequency: For a given hysteresis band,switching frequency is high in case of IM compareto PMSM. This results in higher inverter losses andlow overall drive efficiency in the case of IM.3. Power factor: PMSM operates at almost UPF forall loads, whereas in IM the power factor variesfrom 0.23-0.72.4. Torque: PMSM operates in true synchronous mode(as long as torque generated is more than the loadtorque), the torque ratio of PMSM and IM are foundto be 2. Higher ratios are possible for higher statorcurrents. The magnitude of current is limited onlyby the thermal limitation of the motor in the case ofPMSM. This limit is much less in IM due to rotorcopper losses.5. Windage losses: Since PM rotor has a smoothsurface which is better than the IM, the windagelosses are less or equal.6. Efficiency: Higher in case of PMSM compared toIM.7. Experimental setup: Almost the same for IM andPMSM.8. Speed encoder: In the case of IM, since therequirement is only to measure the speed, a simplespeed encoder can be used. Whereas in PMSM, asophisticated encoder can be used, as the encoderpulses are needed to generate the reference signals.9. Stress on the switching devices: Since the IMalways operate at lagging PF it draws more currentfor a given load compared to PMSM. As a result,stress on the switching device is more. PMSM canbe operated in leading PF.10. Software complexity: Since slip speed has to bedetermined in the case of IM, software structure iscomplex and the time taken by the controller toexecute the software program once is higher in IMcompared to PMSM.11. Response time: Better in PMSM.12. Starting capability: Good in case of IM. PMSMstarts only at low frequency when operated in openloop. This capability can be improved significantlyby operating in self-controlled mode.13. Stability: IM is more stable compared to PMSMwhen operated in open loop.14. Reliability: Aluminum and iron have differentcoefficients of thermal expansion. This leads toreliability problem over the life of IM if it alwaysoperated on full load continuously, or when it isused in applications which require load changesover the cycle. Absence of these windings makePMSM more reliable.15. Cost: Since magnet cost is high, PMSM is veryexpensive.16. Fabrication complexity: Special jigs are requiredto fabricate the PM rotor. Assembly of PMSM ismore complicated compared to IM.17. General comment: IM are more sensitive to bothsupply voltage and frequency variations. Torquecapability of these motors varies with the square ofthe voltage, inversely with the square of thefrequency. But their speed is less dependent onvoltage and directly related to frequency. Whereasin PMSM, the speed is independent of voltage.III. MATHEMATICAL MODEL OF PMSMVector control is one of the control technologieswith high performance for PMSM [3]. For the controlsystem, to decide control strategy should be based onspecific applications and control purposes. Whenstarting the engine in HEV, the motor should provideconstant torque to start engine quickly. In this paper, themaximum torque control strategy is used. To minimizestator current while meet torque demand, this reducesInternational Conference on Software Technology and Computer Engineering (STACE-2012), ISBN : 978-93-81693-68-1, 22 nd July 2012,Vijayawada43

PMSM Electric Drive System simulation for Parallel Hybrid Electric Vehiclethe copper loss of the motor, and is in favor of inverterswitching device to work, but also to reduce costs of thePMSM drivingsystem.Fig 1: PMSM Control system BlockUsing maximum torque/current control, the motorCurrent vectormust meet:IV. DESIGN OF CONTROLLERSThecharacteristics of the each of proportional (P),the integral (I), and thederivative (D) controls, andhowto use them to obtain a desired response. In this tutorial,we will consider the following unity feedback system:Fig 3: Control system Block diagramPlant: A systemto be controlledController: Provides the excitation for the plant;Designed to control theoverall systembehavior.The three-term controller:Thetransfer function of the PID controller lookslike the following:• Kp = Proportionalgain• KI = Integral gain• Kd = Derivative gainWorking of the PID controller in a closed-loopsystem using the schematic shown above. The variable(e) represents the tracking error, the difference betweenthe desired input value (R) and theactual outputt (Y).This error signal (e) will be sent to the PID controller,and the controller computes both thederivative and theintegral of this error signal. The signal (u) just past theInternational Conference on Software Technology and Computer Engineering (STACE-2012), ISBN : 978-93-81693-68-1, 22 nd July 2012,Vijayawada44

PMSM Electric Drive System simulation for Parallel Hybrid Electric Vehiclecontroller is now equal to the proportional gain (Kp)times the magnitude of the error plus the integral gain(Ki) times the integral of the error plus the derivativegain (Kd) times the derivative of the error.This signal (u) will be sent to the plant, and the newoutput (Y) will be obtained. This new output (Y) will besent back to the sensor again to find the new error signal(e). The controller takes this new error signal andcomputes its derivative and its integral again. Thisprocess repeats.CLresponse Rise time Overshoot Settling S-s errortimeKp Decrease IncreaseSmallChangeDecreaseKi Decrease Increase Increase EliminateKdSmallChangeDecreaseDecreaseSmallChangeTable 1: Comparison between controllersEffects of each of controllers Kp, Kd and Ki on aclosed-loop system are summarized in the table shownabove. A PI controller responds to an error signal in aclosed control loop and attempts to adjust the controlledquantity to achieve the desired system response. Thecontrolled parameter can be any measurable systemquantity such as speed, torque, or flux.The benefit of the PI controller is that it can beadjusted empirically by adjusting one or more gainvalues and observing the change in system response.A digital PI controller is executed at a periodicsampling interval. It is assumed that the controller isexecuted frequently enough so that the system can beproperly controlled. The error signal is formed bysubtracting the desired setting of the parameter to becontrolled from the actual measured value of thatparameter. The sign of the error indicates the directionof change required by the control input. TheProportional (P) term of the controller is formed bymultiplying the error signal by a P gain, causing the PIcontroller to produce a control response that is afunction of the error magnitude.As the error signal becomes larger, the P term of thecontroller becomes larger to provide more correction.The effect of the P term tends to reduce the overall erroras time elapses. However, the effect of the P termreduces as the error approaches zero. In most systems,the error of the controlled parameter gets very close tozero but does not converge. The result is a smallremaining steady state error. The Integral (I) term of thecontroller is used to eliminate small steady state errors.The I term calculates a continuous running total of theerror signal. Therefore, a small steady state erroraccumulates into a large error value over time. Thisaccumulated error signal is multiplied by an I gain factorand becomes the I output term of the PI controller.V. TUNING OF PI CONTROLLERSProportional-integral (PI) controllers have beenintroduced in process control industries. Hence varioustechniques using PI controllers to achieve certainperformance index for system response are presented.The technique to be adapted for determining theproportional integral constants of the controller, calledTuning, depends upon the dynamic response of theplant.In presenting the various tuning techniques we shallassume the basic control configuration wherein thecontroller input is the error between the desired output(command set point input) and the actual output. Thiserror is manipulated by the controller (PI) to produce acommand signal for the plant according to therelationship.U(s) = K p (1+1/ τ i s)(12)Or in time domainU(t) = K p [e(t) + (1/τ i ) ∫ edt] (13)whereK p = proportional gainτ i = integral time constantIf this response is S-shaped as in, Ziegler-Nichols tuningmethod is applicable.Fig 4: S shaped response of plantZeigler- Nichols Rules for tuning PI controllers:First Rule: The S-shaped response is characterizedby two constants, the dead time L and the time constantT as shown. These constants can be determined bydrawing a tangent to the S-shaped curve at the inflectionpoint and state value of the output. From the response ofthis nature the plant can be mathematically modeled asInternational Conference on Software Technology and Computer Engineering (STACE-2012), ISBN : 978-93-81693-68-1, 22 nd July 2012,Vijayawada45

PMSM Electric Drive System simulation for Parallel Hybrid Electric Vehiclefirst order system with a time constant T and delay timeL as shown in block diagram.The gain K corresponds to the steady state value ofthe output C ss . The value of K p , T i and T d of thecontrollers can then be calculated as below:KP=1.2(T/L)(14)τi = 2L(15)Fig 6: Neural Network based controllerVII. SIMULATION RESULTSTable 1: Specifications:J=0.0176 B=0.0003818P=6 Ld-Lq=0.156ContinuousFig 5: Mathematical ModelVI. NEURAL NETWORKS BASEDCONTROLLERNeural networks can perform massively paralleloperations. The exhibit fault tolerance since theinformation is distributed in the connections throughoutthe network. By using neural PI controller the peakovershoot is reduced and the system reaches the steadystate quickly when compared to a conventional PIcontroller.Program for creating the Neural Network:load nk1=max(i');k2=max(o1');P=i'/k1;T=o1'/k2;n=157128;net = newff(minmax(P),[5 1],{'tansig' 'purelin'});net.trainParam.epochs = 200;net = train(net,P,T);Y = sim(net,P);plot (P,T,P,Y,'o')gensim(net,-1)pow erguiiabci_abc (A)TmrpmLoad1800 -Krad/secwe-K-PIiq*VaVaRef. Speed (rpm)IabcTe0 id irabcVbVbiaid=0thetaIr_abcRef.I_abcVcVcthetarpm, Nm, Aidq0-i_abcPWM inverterPMSM ModelFig 7: Simulation model of PMSM3costheta2idProduct7sinProduct11iq*cos2*pi/3Product2Mux1si nirabcProduct3cosProduct4si nProduct5International Conference on Software Technology and Computer Engineering (STACE-2012), ISBN : 978-93-81693-68-1, 22 nd July 2012,Vijayawada46

gCgCEEgCgCEEgCEgCEPMSM Electric Drive System simulation for Parallel Hybrid Electric Vehicle[s1][s3][s5]FromFrom2From4pulsesIGBT/Diode1IGBT/Diode3IGBT/Diode5[s1]1Iabc2Ir_abcIabcpulsesIr_abcGoto[s2]Goto1[s3]Goto2[s4]Goto3[s5]Goto4Vdc[s2]From1IGBT/Diode4[s4]From3[s6]From5IGBT/Diode6A VaB VbC VcoutageIGBT/Diode2a Ab Bc CLP filter1Va2Vb3Vc[s6]Goto52Va3Vb4VcvavbvcthetaVqVd1TmVqTmVdidiqTewetheta3Te2weidiqthetaiabcia1iabc4iaa. PMSM simulation with NN Controllerv_abc to VdqPMSM_Crtidq0-i_abc5theta2iqid1Vqwe iqVqq_axis crt2TmIqTmweTe4we3Te3VdVdIqweidIdthetatorque _speed crt5theta1idd_axis crtb. PMSM simulation with PI Controller1Iq3IdProduct20.1546yaf0.0008Ld-Lq2Te16*(3/4)6/20.00176s+0.00038818P*(3/4)1/Js+BP/22TmFig 8: Sub blocks of PMSM1we1sIntegrator23thetaFig 9 (a & b): Simulation Results for NN and PIcontrollerVIII.CONCLUSIONSFor PMSM control and driving in HEV, accordingto PMSM mathematical model, this paper deducesPMSM vector control algorithm, studies HEV energyregeneration requirements for motors, and detailedanalyzes power generation work mechanism of PMSM.Then the control strategy of PMSM energy regenerationis proposed and the simulation results of the motorcontrol system show that the system has goodperformance.International Conference on Software Technology and Computer Engineering (STACE-2012), ISBN : 978-93-81693-68-1, 22 nd July 2012,Vijayawada47

PMSM Electric Drive System simulation for Parallel Hybrid Electric VehicleACKNOWLEDGMENTI would like to first acknowledge and express mysincere thanks to my supervisor& co-supervisor Dr.P.V.RajGopal & Dr. G.TulasiRamDas, for theopportunity that he gave me to work on this highlypromising and exciting research area. I would also liketo thank Dr.C.K.Sarma, a senior professor from theG.R.I.E.T, whose helped me in understandingmathematical formulation. Finally, a special thank yougoes to my parents M.Rama Krishna Rao & M. GirijaKumari for their moral and financial supports and alsomy husband P.Srikanth for his encouragementthroughout.REFERENCES[1] Cikanek, S.R.; Bailey, K.E. Regenerative brakingsystem for ahybrid electric vehicle. Proceedings ofthe American ControlConference 2002 , Volume: 4:3129 –3134[2] Wu Hong-xing, Cheng Shu-kang, Cui Shu-mei.Communicationon Vehicle Management Unit in theElectric Vehicle. 12th EML,May 28-31,2004,Saint-Louis,USA. IEEE TransactionsonMagnetics, v 41, n 1 II, January, 2005, p 514-517.[3] Chan, C.C.; Zhang, R.; Chau, K.T.; Jiang, J.Z.Optimal efficiencycontrol of PM hybrid motordrives for electrical vehicles. PowerElectronicsSpecialists Conference, 1997. PESC '97 Record.,28 th Annual IEEE , Volume: 1 , 1997: 363 -368vol.1[4] W. Xia,Philip. Chin. A Specially Designed EVPM Motor Drive.Energy Management and PowerDelivery, 1998.Proceedings ofEMPD '98. 1998International Conference on , VolumeInternational Conference on Software Technology and Computer Engineering (STACE-2012), ISBN : 978-93-81693-68-1, 22 nd July 2012,Vijayawada48