光伏发电系统最大功率点跟踪算法研究及实现

苏州大学硕士学位论文光伏发电系统最大功率点跟踪算法研究及实现姓名：黄克亚申请学位级别：硕士专业：集成电路工程指导教师：李文石2010-11IMPPTMatlabSimulinkI-VP-VBoostMPPT3V32MPPT35.6%1.1%MPPTMPPTMPPT(MPPT)BoostAbstractTheAlgorithmResearchofPVSystemMPPTanditsImplementationIITheAlgorithmResearchofPhotovoltaicSystemMaximumPowerPointTrackinganditsImplementationAbstractSolarphotovoltaicsystemhasbecomeoneofthehotfieldsinnewenergyresearch.Inthispaper,inordertoobtainthemaximumpowerofsolarcells,photovoltaicsystemswerediscussedandanalyzed.Bycomparingtheknownmaximumpowerpointtracking(MPPT)algorithms,micro-controller-unitsystemforverifyingwasbuilt.Consequently,wehavedeepenedtheunderstandingtheeffectofphotovoltaicandappreciatedtheexcellentalgorithmtechnologyforthecontributionofgreendesign.ThemodelofphotovoltaiccellswasestablishedbyusingSimulinkembeddedfunctioninMatlab,theI-VandtheP-Vcharacteristicsofphotovoltaiccellswereanalyzed.Besides,thecircuitmodelofBoostforresearchMPPTwasabstracted.BytheCombinationoftwomodels,photovoltaicsystem'sbasicMPPTprinciplewassimulated.ThethreeclassicalMPPTalgorithms,suchasPerturbationandObservation(P&O),IncrementalConductanceAlgorithm(ICA)andFuzzyControl(FC),werecompared.Forthefirstandthesecondalgorithms,wescannedvoltagevariablewiththreedifferentdutycyclesteps(small,medium,large),andthenobservedthetwokeydependentvariables(tracedelayandpowerfluctuation).ThesimulationresultsshowICAisslightlybetterthanP&O,butthesizeofthetrackingstepwillaffectthedynamicandstaticperformances,andthetrade-offdesignwasmoredifficulttofind.TheMPPTalgorithmbasedonFCwasemphasesfocusedanddiscussed.Thekeytechnologyadoptedartificialneuralnetworktogeneratefuzzycontrolrulebyusingthemeasureddata.ComparingwithP&OandICAinTheAlgorithmResearchofPVSystemMPPTanditsImplementationAbstractIIIsimulation,wecanseethatFChasbetterdynamicandstaticperformances(tracedelayandthepowerfluctuationsarereducedby35.6%and1.1%).Bytheaidofself-establishedmicro-controller-unitsystem,theP&Omethodwasverifiedsuccessfully,andthemaximumpowerpointwasfoundaccurately.Whatismore,wealsoanalyzedthetechnicalfeasibilityoftheMPPTalgorithmbasedonFCwithhardwareandsoftware.SimulationresultsshowthattheperformanceofthecomplexFCMPPTalgorithmisindeedbetterthanP&OandICAmethods.Bytheachievementsofsoftwareandhardwaredesign,asimplealgorithmsuchasP&Omethodisconductedeasily.ThisattemptedworkmaybenefittoacceleratethephotovoltaicMPPTresearchformicroelectronicsystemintegrationandbasicdataaccumulation.Keywords:PhotovoltaicSystem,Maximumpowerpointtracing,Boost,Fuzzylogiccontrol,MCUWrittenbyHuangKeyaSupervisedbyLiWenshi11.121200010.1%58.8%3.4%1.3%36.5%/1-1[1]1-1[2]2[3]211.2MPPT1.2.1[4]1-21-22050208030%~40%213195819712000[5]1.2.2MPPT20/12%~17%MPPT1-11-1Vmax[22,23,24]Vm[6,14,15][16,17,21]dI/dU-I/U[26,27,28][32,34,35,36,38]MPPTMPPT1.31.3.1MatlabMSimulinkI-VP-V4DC-DCBuckBoostBoostBoostMPPTMPPTMPPT1.3.21SimulinkBoost235MatlabDC-DC2.12.1.1[6][7]2.1.2Stand-alonePVSystem[8]Grid-connectedPVSystem[9]2.2Matlab2.2.1[10]2-1PNPNNPILPNPNPNIF2-1PNILIF62-12.2.22-2[10]2-2IphDP-NRshRsIV2-2P-N2-1[10]0()exp1ssphshqVRIVRIIIInKTR++=−−−(2-1)I(A)V(V)q(l.6x10-19C)K(l.381023J/K)T(K)n(n=1-5)I02-2[11]73011expgaprrrarqETIITKTT=−(2-2)TrIrr(0K)EgapsshUIRR+2-10()exp1sphqVRIIIInKT+=−−(2-3)2-12-2IphI0nRshRs[11]2.2.3Matlab1Matlab[12]MATLABMATLABSimulink2Simulink[13][14]Matlab/SinulinkSimulinkMSolarexMSx6060W1000W/m2252-182-1SolarexMSx6060wPm59.9WIm3.5AVm17.1VUoc3.74AIsc21.0VKt2.06(mA/)KT-0.07(V/)2-3IphI0RSVII-V2-3IVI2-3MatlabMSolarFun(Ua,G,Tac)Ia2-32-3MatlabMI-VP-V2-492-42-412380%4(MPPMaximumPowerPoint)2.3DC-DCDC-DC10DC/DC[15]---[16](Buck)(Boost)(Buck-Boost)(BuckConverter)(BoostConverter)DC-DC2.3.1DC-DC1BuckConverter2-5VoVin2-4DonoinintVVDVT==(2-4)C1C2DLRVS+-VinVoisilVlS--+++-2-5BuckPWM2-5C112BoostConverter2-6SDLCSCVl=VinSVl=Vin-Vo11CDLRVS+-VinVoisilVlS--+++--2-6Boost13[17]2-6LCViVoiLBoostTD0-DTS2-7(a)ViLCDTTS2-7bLC2-70DTLLidiivLLdtDT∆==(2-5)DTT12(1)LLoidiiVVLLdtDT∆−==−(2-6)2-52-61oViVD=−(2-7)2.3.2BuckBoostBuckMPPTBuckMPPTBoostBoost50V100V[18](BoostConverter)BoostDC-DC2.3.3BoostMatlab1BoostMatlabsimulinksimulink(simpowersystemtoolbox)PspicesimulinkBoost2-8132-8BoostMatlab210V100kHz50%R100L2e-4HC2e-5F2-92-700.511.522.533.544.55x10-3051015202530time2-9Boost2.4DC-DCDC-DC(Buck)(Boost)14BoostBoostDC-DCMatlabBoostMPPT15MPPTBoostMPPT3.1MPPT3.1.1MPPT3-1abI-VI-V3-1I-VDC-DCDC-DCDC-DCDMPPT163.1.2BoostMPPT1BoostBoostDC-DC3-2Boost100%Boost[19][20]3-2Boost2-7BoostV’=VL*(1-D)I’=IL/(1-D)22*(1)''*(1)*(1)'/(1)LLLLLVDVVRDRDIIDI−===−=−−(3-1)RBOOSTDRLBOOST3-1BoostBoostBOOST2BoostBoostMatlabsimulinkEmbeddedMatlabFunctionfunctionR2=Rchang(RLD)R2=RL*(1-D)^2;3-3100301000W/m225MPPT17D[01]plotP-D3-43-3BoostMPPT3-4BoostP-DD0-1RL100D0.783-1R’=100*10.7824.84RL=30D0.63-1R’=30*10.624.82-1Um17.1VIm=3.5ARmRm=Um/Im=4.89RmRL10030BoostBoostMPPT18MPPT3BoostMPPT[21]Boost?3-1R’=RL*(1-D)2(l-D)21Boost3-5BoostRLABBoost05101520253000.511.522.533.54UBARL1R'RL23-5Boost3.2MPPTMPPTMPPTMPPTMPPT193.2.13-6P-V()MPPTVmaxVmaxMPPTCVTConstantVoltageTrackingMPPTCVTCVT20[22]3-6Um10.35~0.45[23]3-7-300+