Urban Power Grids Dynamic Control Model with Photovoltaic and Electric Vehicles

《Urban Power Grids Dynamic Control Model with Photovoltaic and Electric Vehicles》由会员上传分享，免费在线阅读，更多相关内容在学术论文-天天文库。

2018ChinaInternationalConferenceonElectricityDistributionTianjin,17-19Sep.2018UrbanPowerGridsDynamicControlModelwithPhotovoltaicandElectricVehicles122213QianXiong,FangfangLiu,LinLv,YouboLiu,YongLi,ChengzhiZhu1.TianfuPowerSupplyCompanyoftheStateGridSichuanElectricPowerCompany,HuayangRoadNo.11,Chengdu,China.2.SichuanUniversity,YihuanRoadNo.24,Chengdu,China.3.StateGridZhejiangElectricPowerCompany,HuanglongRoadNo.8,Hangzhou,China.topologicalstructureoftheHVDN,theyproposeddifferentAbstract—Thegrowingpenetrationlevelofphotovoltaic(PV)transmissioncongestionmanagementmodels.However,theyandelectricvehicles(EV)increasetheoperationalriskofthefailedtoconsidertheinfluencesoftherenewableenergyurbanpowersystem.Besides,theunbalancedloaddistributionatresources(RES).Theintermittenceandtheuncertaintyofthetime-spatialscalealsoimpactsthehostingcapacityofthePVsandRESareeasytocausethefrequentmanipulationsoftheHVDNcausesseveretransmissioncongestion.Thusthispaperproposedaswitches.Reference[6-7]proposedtheactiveenergybi-leveloptimizationmodeltomanagethetransmissioncongestionandenhancethePVhostingcapacityconsideringthemanagementstrategiesinanattempttoenhancethehostingreconfigurablecapabilityofthehighvoltagedistributionnetworkcapacityofthePVswiththeassistanceoftheenergystorage(HVDN).Intheupperlevel,theoptimaltopologicalstructureofsystems(ESS)controlschemes.ButthesimpleaccommodationHVDNiscalculatedaimingatminimizingthetotaloperationalofthePVsdeterioratesthecongestionproblemsespeciallycost.Inthelowerlevel,themaximumhostingcapacityofPVsisunderthehighpenetrationleveloftheEVs.Reference[8]achievedbyasecond-orderconeprogrammingmodel.TheutilizedthereconfiguretechniqueofthemediumvoltageproposedmethodwasverifiedbyanurbanpowersysteminChina.NumericalresultsdemonstratedthatreconfiguringtheHVDNdistributionnetwork(MVDN)toenhancetheconsumptiontopologicalstructureprovidehugebenefitsofenhancingPVamountoftheRES.ButthetopologicalstructureoftheHVDNpenetrationlevelandfacilitatingtheintegrationoftheEV.andMVDNhassomedistinctdifferencesandthereconfiguretechniqueadoptedinMVDNarenotsuitableinHVDN.IndexTerms—congestionmitigation,double-layeroptimization,Thusthispaperfocusedonthedynamiccharacteristicsoftheelectricvehicle,photovoltaicoutputofPVsandthechargingloadofEVsandestablishedabi-leveltransmissioncongestionmanagementmodelbasedonI.INTRODUCTIONthereconfigurabilityoftheHVDN.Intheupperlevel,theHEincreasingurbanizationlevelinChinaenablestheToptimaltopologicalstructureoftheHVDNisdetermined.Inurbanpowergridsenjoyinganunprecedentedexpansion.thelowerlevel,thebestloadcurtailmentamountandPVThe220kVtransmissionnetworkshasbeengraduallysheddingstrategyareobtained.TheproposedmethodwasextendingdeepintotheloadcenterandshapingthebackboneverifiedinarealurbanpowersystemsinChinaandnumericalstructureoftheurbanpowergrids.WhilethegrowingamountresultsshowedthereconfigurationoftheHVDNwasanoftheelectricvehicles(EV)andthepenetrationleveloftheeffectivetoolforcongestionmanagementandfacilitatingthephotovoltaic(PV)augmentthedifficultyofholdingthebalanceaccommodationofthePVs.betweenthepowersupplyanddemand.Intheoperationaltime,theunbalancedloaddistributionispronetoexperiencetransmissioncongestionproblems.ThusthispaperresortstoII.MODELINGOFTHEDYNAMICLOADreconfiguringthetopologicalstructureofthehighvoltagedistributionnetwork(HVDN)toimprovetheloaddistributionA.EVChargingLoadModel[10]ateach220kVsubstationandachievingtransmissionThispapertookthePHEV60(EPRI)asanexampleandcongestionmitigation.usedthesmartcharging/dischargingmodeldescribedin[11]toSomeresearcheshadalreadyinvestigatedthisissue.establishtheEVloadmodel.ThusthechargingloadareReference[1-5]modeledHVDNbypowersupplyunitsanddescribedasequation(1).theirconnectionrelationships.BasedontheflexibleeEVnEVp(n)(n0,1,...)(1)EVEVn!ThisworkwassupportedbytheScienceandTechnologyFoundationofEVSGCC(521104170009).CICED2018PaperNo.201804270000164Page1/52462

12018ChinaInternationalConferenceonElectricityDistributionTianjin,17-19Sep.2018WhereλEVistheexpectedvalueoftheaccessingnumberof21MtttheEVsandnEVisthepossiblenumberoftheEVsthatconnectCmin{iFifi}(7)tothegrid.i1WherethefirsttermofthefunctiondenotesthecostsoftheB.OutputModelofPVloadcurtailmentandPVsheddingamount.ThesecondtermAccordingtoreference[12]and[13],theoutputofthePVsisdenotestheHVDNreconfigurationexpense.Mrepresentstherelatedtotheilluminationintensityanditsmathematicalnumberofrandomvariablesandrepresentstheprobabilityidescriptionsisshowedinformula(2).tweight.FiscalculatedbythelowerlevelmodelwhichisWt()()t()t(2)iPGFWherethebasicfunctionµPGF(t)istakenastheexpecteddescribedbyformula(8).Thetopologicalstatusconstraintsisoutputofthephotovoltaic.θ(t)istheobstructioneffectofthedetailedinreference[5].atmosphereonthesunlight.C.Lower-levelModelBasedontheoptimaltopologicalstructureobtainedbytheIII.BI-LEVELOPTIMIZATIONMATHEMATICALMODELSupperlevelmodel,thelower-levelmodelevaluatesthecontrolA.DynamicRiskAssessmentofNetworkTopologycostproducedbythenetworkconstraintssuchaspowerflowInordertoevaluatethecurrentoperationalriskoftheurbanconstraintsinanuncertaintyscenario.Thentheevaluationtransmissionsystem,thenetworkcongestionriskindex(NCRI)resultsarefeedbacktotheupper-leveloptimizationmodel.Theisdefinedas(3)and(4).optimalobjectistomaximizethePVhostingcapacitywhichisshowedin(8).GGG(3)nZNttt2Fimin(Pwg,iPg,i)(8)i1+=1(4)WhereωreflectstheenvironmentaleffectsoftheabandonedWherethefirsttermoftheequationdenotestheglobaltoutputofPV.PwgidenotestheexpectedvalueofPVoutputtransmissioncongestiondegreewhilethesecondtermreflectstpower;PgidenotestheactualoutputofPV.Therequiredthemostseriouscongestioncondition.Thelasttermoftheconstraintsareshowedinthenextfollowing.nequationdenotestheglobalPVunitaccommodation.GandPP2GX(GRBY)(9)iQiiiiijijijijji()arethebranchsecurityriskindexvectorandtheoutputnvectorofthePVs.TheirmathematicaldescriptionsaredetailedQiQQi2BXiii(BRijijGYijij)(10)in(5)and(6).ji()gw(eaL()m1)mmxik,(5)tttPsBsssBsww(11)LLLLws()m0m0()Lm(6)220LLm0VVmaxmaxX(12)iWherewxi,kistheprobabilityweight.“a”isapositivenumber.22ω(Lm)representstheoverloadamountofthebranchm.Lmisthe222S(GB)(2X2X2R)(13)loadratioofbranchm.L0isthespecifiedthreshold.Mdenotesijijijijijthetotalbranchnumber.SS22(14)ijmaxB.Upper-levelModel222XXRY(15)ijijijWhenNCRIbeyondthespecifiedrange,thereconfigurationWhere(9)and(10)arepowerflowconstraintsofHVDN.(11)oftheHVDNwillbetriggered.Thetwo-layeroptimizationisthedirectpowerflowconstraintsoftransmissionnetwork.modeldescribedinliterature[14]isusedinthispaper.The(12)~(15)aretherelaxedsecond-orderconeconstraints.upperlevelmodelmainlyoptimizesthetopologystructureofHVDN,andmakethenetworkstructureadaptabletosupplyD.SolvingProcessmoreEVsandincreasePVaccommodation.TheobjectiveTheparticleswarmoptimizationalgorithm(PSO)isusedtofunctionisdescribedin(7).findtheoptimaltopologicalstateandthesecond-orderconeprogrammingmodelisusedtocalculatetheoptimalPVCICED2018PaperNo.201804270000164Page2/52463

22018ChinaInternationalConferenceonElectricityDistributionTianjin,17-19Sep.2018curtailmentamount.TheflowchartoftheproposedmodelisshowedasFig.1S6StartL64L16S5InputnetworkdataL45S4L15L35S1S3InitializetheparametersofthePSOalgorithmL12S2L03L0L02L041InitializethepositionsoftheparticlesEvaluatetheparticlesusinglower-levelmodelBSFig.3Topologicalstructureof220kVtransmissionnetworkAsshowedinFig.3,thetransmissionsystemhassixConverge?Updatethepositionssubstationsandtentransmissionlines.TheHVDNhasthirty-fourpowerunitswhichcanbedividedinto6unitgroups.WhereG1,G3andG6aremainlycomposedbythemunicipalEndandresidentialelectricityload.G2andG4aremainlyFig.1Flowchartoftheproposedmethodcomposedbythecommercialload.G5isusedtopowertheheavyindustrialload.ThefeasibletopologicalinformationoftheunitgroupisshowninTABLEI.IV.CASESTUDYA.ParametersSpecificationTABLEITRANSFORMABLECELLGROUPFEASIBLETOPOLOGICALNUMBERThetopologicalstructureofHVDNisshowedinFig.2whileTotaltopologyFeasibletopologyUnitGroupitssuperiorsystem—220kVtransmissionnetwork—isnumbernumberdetailedinFig.3.ThesimulationplatformisMatlab2016aandG125613thefrequencyofthecomputeris2.5GHz.TheparametersoftheG2325PSOalgorithmisreferredfromliterature[15].G3325U27U28U29U21U16U15G4102421P21P16P15S1P27P28P29G5325UnitGroupG4U24UnitGroupG6U23U30G6102423P22P17U17P24P23P30U22UnitGroupG5U18P18P34P33P31P32P25P26B.NumericalResultU34U33U31U32U20U25U26P20P19U19Fig.4showsthePVunitsaccommodationintheearlypeakS2loadtime.Duringthisperiod,theEVsaremainlycharginginU6thecommercialareas.TheoutputofthePVsinresidentialareasS6P6S3S5remainlow.ThetransmissioncongestionisveryseriousandU5UnitGroupG1U4U8U9U11theriskofthesystemhitsnearly62.73%asshowedinFig.5.P5P4P8P9P11UnitGroupG3U1U2U3U7UnitGroupG2U13P1P2P3P7P10P12P13U10U12S4P14220kVtransformationunitU14transformationunitPhotovoltaicunitChargingunitFig.2TopologicalstructureofHVDNCICED2018PaperNo.201804270000164Page3/52464

32018ChinaInternationalConferenceonElectricityDistributionTianjin,17-19Sep.2018PVsaccommodation(100%)80%70%60%50%40%29.79%30%20%17.2%10%6.5%7.42%0%PVUnitsP1P12P29P32Fig.4PVaccommodationatearlypeakloadtimebeforeHVDNreconfiguration.Fig.7NetworksafetyregionafterconductingHVDNreconfigurationFig.6manifeststhattheHVDNreconfigurationcansignificantlyenhancethePVaccommodationandenlargethesecurityregionofthetransmissionsystems.ThroughchangingtheconnectionrelationshipsoftheHVDNpowerunits,theremnantPVoutputintheresidentialareaweretransferredtotheseriouslycongestedcommercialarea.AsshowedinFig.7,thenetworksafetyriskisgreatlyreducedto5.32%.TABLEIIshowstheindexwhichrepresentstheabsorptionoftheoutputofPVsbeforeandaftertheHVDNreconfigurationfrom8:00to18:00.ThetriggeringtimeoftheHVDNreconfigurationare8:00,12:00,15:00,and18:00Fig.5NetworksafetyregionbeforeHVDNreconfigurationatearlypeakloadrespectively.timeThepercentageofNCRIindicatesthedegreeofrisktotheTABLEIInetwork,andwhenthevalueofNCRIexceeds0.3,itisPVSACCOMMODATIONBEFOREANDAFTERRECONFIGURATIONassumedthattheurbannetworkislikelytohaveablockingriskTimeBeforeafter(hour)ReconfigurationYes/Noreconfigurationatthemoment.TheNCRIinfigure5is0.6273beyondtheconfidencerange.Thatindicatesthetransmissionnetworks815.23%162.32%haveseriouscongestionandlowpenetrationlevelofPVunits.958.78%058.78%Afterconductingtheproposedmethodology,theoptimalresultsofthenetworkoperationalstatusareexhibitedinFig.61053.04%053.04%andFig.7.1150.49%050.49%PVsaccommodation(100%)1217.49%178.03%80%73.46%1339.76%039.76%70%68.64%60%57.96%1434.48%034.48%49.22%50%1516.7%159.83%40%30%1643.68%043.68%20%1742.09%042.09%10%0%PVUnits1810.11%163.57%P1P12P29P32Fig.6PVaccommodationafterconductingHVDNreconfigurationCICED2018PaperNo.201804270000164Page4/52465

42018ChinaInternationalConferenceonElectricityDistributionTianjin,17-19Sep.2018ItcanbeconcludedfromTABLEIIthatafterenforcingtheREFERENCESHVDNreconstruction,theaccommodationofPVscanbe[1]F.A.J.Yong,S.B.L.Junyong,T.C.L.HongweiandF.D.Z.Xi,"Theincreasedbyatleast30%.Take12:00asanexample,theloadbi-levelprogrammingmodelofloadtransferringstrategybasedonofEVsisdischargedwithahighprobability,andtheleveloftopologicalunitsofhigh-voltagedistributionnetwork,"2016ChinaInternationalConferenceonElectricityDistribution(CICED),Xi'an,thePVaccommodationislessthan20%.Thesystemnetwork2016,pp.1-7.congestionareaismainlyconcentratedinresidentialareasand[2]JINYong,LIUJunyong,ZHANGXi,etal.Bi-levelModelBasedonindustrialareas,andtheriskofthesystemisnearly51.49%.ByFunctionUnitsforOptimizingHigh-levelVoltageDistributionNetwork[J].AdvancedEngineeringSciences.2017,49(3):179-190.triggeringthenetworktopologyreconfiguration,theloadofthe[3]ZhangXi,LvLin,JinYong,etal.CongestionMitigationModelandindustrialareasinheavierresidentialareasweretransferredtoAlgorithrmforUrbanPowerGridsConsideringReconfigurabilityofthelightloadedpartthusmitigatesthetransmissioncongestion.High-voltageDistributionTransformerUnitGroups[J].ProceedingsoftheCSEE.2016,36(20):5403-5414.Besides,thedegreeofPVaccommodationwereincreasedto[4]JINYong,LIUYoubo,LIUJunyong,etal.OperationCongestion78%.ThecomparisonofthenetworkriskisshowedinFig.8.ManagementModelforUrbanPowerGridsBasedonLoadTransferLogicalConstraints[J].AutomationofElectricPowerSystems.2016,40(23):77-85.[5]ZHANGXi,LIUYoubo,LVLin,etal.TotalSupplyCapabilityAnalysisRisk(%)RiskchangesbeforeandafterreconfigurationofUrban220kVAreaPowerSystemConsideringLoadTransferCapabilityofHVDistributionNetwork[J].PowerSystem80Technology.2017,41(5):1612-1620.Beforereconfiguration[6]LUANWeijie,JIANGXianwei,ZHANGJietan,etal.Consumptive70AfterreconfigurationAbilityAnalysisforDistributedPhotovoltaicGenerationConsideringReconfigurationtimeActiveManagement[J].ElectricPowerConstruction.2016,31(1):137-143.60[7]LIPeng,HUAHaorui.Two-stageMultiObjectiveConsumptionModelforDistributedPhotovoltaicBasedonEnergyStorageScheduling50Mode[J].ElectricPowerScienceandEngineering.2016,32(10):1-8.[8]ZHUJunpeng,GUWei,ZHANGHandan,etal.OptimalSitingand40SizingofDistributedGeneratorsConsideringDynamicNetworkReconfiguration[J].AutomationofElectricPower30Systems.2018,42(5):111-119.[9]ZhangLibo,ChengHaozhong,ZengPingliang,etal.AThree-Point20EstimateMethodforSolvingProbabilisticLoadFlowBasedonInverseNatafTransformation[J].TransactionofChinaElectrotechnical10Society.2016,31(6):187-194.[10]AXSENJ,BURKEA,KURANIK.BatteriesforPlug-inHybridElectricTime(h)Vehicle(PHEVs):goalsandthestateoftechnologyCirca2008[R].89101112131415161718[S.l.]:UniversityofCalifornia,2008.[11]QIANK,ZHOUC,ALLANM,etal.ModelingofloaddemandduetoEVbatterychargingindistributionsystems[J].IEEETransonPowerFig.8ComparisonofthenetworkriskbeforeandafterHVDNreconfiguration.Systems,2011,26(2):802-810.AsshowedinFig.8,theriskofthenetworkcanbe[12]CHENCan,WUWenchuan,ZHANGBoming,etal.ProbabilisticLoadconsiderablydecreasedbyreconfiguringtheHVDNFlowofDistributionNetworkConsideringCorrelatedPhotovoltaictopologicalstructure.PowerOutput[J].AutomationofElectricPowerSystems.2015,39(9):41-47.[13]WUChenxi,WENFushuan,CHENYong,etal.ProbabilisticloadflowofV.CONCLUSIONpowersystemwithWFs,PVsandPEVs[J].ElectricPowerAutomationEquipment.2013,33(10):9-15.TheexistingresearchesfailtocounttheimpactsofRESand[14]ZhengEnze,GuJie,LiuBo,etal.OptimalConfigurationofEnergytheprobabilisticmobilityofEVchargingloadsthuswouldStorageforMulti-energyComplementaryMicro-gridsBasedonacausefrequenttransferactions.ThispapercombinedtheBi-levelOptimizationModel[J].PowerSystem&Automation.2017,39(3):45-58.probabilisticpowerflowwiththesecond-ordercone[15]SUHaibin,GAOMengze,CHANGHaisong.MicrogridDroopControlprogrammingmodeltoimprovethefitnessoftheHVDNofActiveandReactivePowerBasedonPSO[J].TransactionsofChinatopologicalstructureandenhancethecomputationefficiency.ElectrotechnicalSociety.2015,30(1):365-369.TheproposedNCRIindicatorproperlyreflectsthesystemXiongQian:vicegeneralmanagerofStateGridSichuanPowerCompany,securityriskandPVaccommodationwhichprovideahelpfulreceivedmasterdegreefromSchoolofElectricalEngineeringandInformationdecisionbasisfortheoperators.Besides,numericalresultsofSichuanUniversityin2008,majorinHVDCtransmission,powersystemmanifestedthatreconfiguringHVDNtopologicalstructureshasprotectionandreliability.hugebenefitsforimprovingsystemsecuritylevelandLiuFangfang1993,Kaifeng,China,female,apresentmasterinSichuanenhancingPVhostingcapacity.University,majorinplanningandoperationofthedistributednetwork.E-mail:894209600@qq.com.CICED2018PaperNo.201804270000164Page5/52466