基于模糊神经网络的电力系统连锁故障风险评估

41620076()JournalofZhejiangUniversity(EngineeringScience)Vol.41No.6Jun.2007:2006-02-16.():www.journals.zju.edu.cn/eng:(20050908).:(1978-),,,,.E2mail:[e-mail]:,,,.E2mail:[e-mail],,(,310027):,,,.,.,,,,.IEEE1182bus.:;;;:TM712:A:1008-973X(2007)06-0973-07RiskassessmentofcascadingfailureinpowersystembasedonfuzzyneuralnetworkCHENWei2hua,JIANGQuan2yuan,CAOYi2jia(CollegeofElectricalEngineering,ZhejiangUniversity,Hangzhou310027,China)Abstract:Thebasicprocessandtheoryofpowersystemcascadingfailureconsideringprotectionsystemhiddenfailureswereanalyzedtodecreasebulkblackoutfailure.Theriskmodelwasbuilt,andtheriskas2sessmentindexandassessmentapproachofpowersystemcascadingfailurewereproposedbasedonfuzzyneuralnetwork.Somepreventivestepswererealizedbyusingthetrainedfuzzyneuralnetwork,whichcouldreducethesystemcascadingfailureriskbyupdatingthecharactersofprotectionsystem.Thisap2proachgraspstheprobabilityandseverityofpowersystemcascadingfailure,soitaccuratelydescribesthesecuritystateofpowersystem,effectivelyclassifiesthesystemrisk,andhasfasterassessmentvelocityandgoodpracticality.ThefeasibilityofthisapproachwasillustratedbyapplicationinIEEE1182bussystem.Keywords:cascadingfailure;fuzzyneuralnetwork;riskassessment;hiddenfailure,,,..,,[122].[3].Dobson[426],,.Yu[7],.,.,,.,.,,,,;,,..IEEE1182bus.11.1.,[8].,,[9].[10].1.,PlineZ,Z3,Z3Z3,PW;Z3Z3,:Pline=PW,Z3Z3;PWexp(-Z/Z3),Z3Z3.(1)1,,,.1Fig.1Hiddenfailureprobabilityinline2protectiverelay1.2.,.[11]..2(,111),3,[12].,2.3,,.,,.37,,..:1,1,23457.:2,G1.:3,2467.:4,22.:5,489.479()41:6,M.:8,49,.3,3.1:,136,.2:,12,.3:,14,.,,.22.1[13].,,.,,.R=PeventIevent.(2):R,Pevent,Ievent.2.2,,[7].3.2.2.1PLI=1NNi=1L(i);i=1,,N.(3):i,N.,L(i)1,0.ILI=1PSNNi=1PL(i);i=1,,N.(4):PL(i)i,PS.,.,RLI=PLIILI.(5)2.2.2PBI=1NNi=1B(i);i=1,,N.(6),B(i)1,0.IBI=1PSNNi=1PG(i);i=1,,N.(7):PG(i)i.,RBI=PBIIBI.(8)2.2.3PNB=1NNi=1S(i);i=1,,N.(9),S(i)1,0.INB=1PSNNi=1PN(i);i=1,,N.(10):PN(i)i.,RNB=PNBINB.(11)2.2.4.:RINT=wBIRBI+wLIRLI+wNBRNB.(12):wBIwLIwNB,wBI+wLI+wNB=1.2.3(fuzzyneuralnetwork,FNN).,,4.4FNNFig.4FlowchartofriskassessmentofcascadingfailurebasedonFNN5796,:3,.,,,.,,[14].,,,;.,.,.,,,,..3.15,.5Fig.5Structureoffuzzyneuralnetwork,PW=[PW1,PW2,,PWm],m..,,,C,.C,,[15].CPW,nx=[x1,x2,,xn],n,15.,.,x=[x1,x2,,xn].xi,x.N1=n..,5,(NB)(NS)(ZE)(PS)(PB).ji(i=1,2,,n,j=1,2,,5),.N2=5n.,,:aj=FUN(i11,i22,,inn).(13):i1,i2,,in{1,2,,5},j=1,2,,5n,FUN.N3=5n.:…j=j5ni=1i;j=1,2,,5n.(14)N4=5n..,.,y1=5nj=1w1j…j,(15).N5=1.3.2,,w1j(j=1,2,,5n)cijij(i=1,2,,n;j=1,2,,5).5qjf(q)(x(q-1)1,x(q-1)2,,x(q-1)nq-1;w(q)j1,,w(q)j2,,w(q)jnq-1),679()41x(q)j=g(q)(f(q)).,f(1)i=x(0)i=xi,x(1)i=g(1)i=f(1)i;i=1,2,,n.(16),f(2)ij=-(x(1)i-cij)2/2ij,x(2)ij=ji=g(2)ij=exp(f(2)ij)=exp[-(xi-cij)2/2ij];i=1,2,,n,j=1,2,,5.(17),f(3)j=x(2)li1x(2)2i2x(2)nin=i11i22i33,x(3)j=j=g(3)j=f(3)j;j=1,2,,5n.(18),f(4)j=x(3)j5ni=1x(3)i=j5ni=1i,x(4)j=…aj=g(4)j=f(4)j;j=1,2,,5n.(19),f(5)1=5nj=1w1jx(4)j=5nj=1w1j…j,x(5)1=y1=g(5)1=f(5)1.(20)E=12(yd1-y1)2.(21):yd1y1.,(5)1C-5E/5f(5)1=yd1-y1,(22)(4)jC-5E/5f(4)j=-5E5f(5)15f(5)15g(4)j5g(4)j5f(4)j=(5)1w1j,(23)(3)jC-5E/5f(3)j=125i=1i-2(4)j5ni=1,iji-5nk=1,kj(4)kk,(24)(2)ijC-5E/5f(2)ij=5nk=1(3)kSijexp[-(xi-cij)2/2ij].(25):g(2)ij=jik,Sij=5f(3)k/5g(2)ij=nj=1,jiijj,Sij=0.5E5w1j=5E5f(5)15f(5)15w1j=-(5)1x(4)j,(26)5E5cij=5E5f(2)ij5f(2)ij5cij=(2)ij2(xi-cij)2ij,(27)5E5ij=5E5f(2)ij5f(2)ij5ij=-(2)ij2(xi-cij)23ij.(28),,w1j(k+1)=w1j(k)-5E5w1,j;j=1,2,,5n.(29)cij(k+1)=cij(k)-5E5cij;i=1,2,,n,j=1,2,,5.(30)ij(k+1)=ij(k)-5E5ij;i=1,2,,n,j=1,2,,5.(31)4IEEE1182bus[16],.4.1IEEE1182bus186,,,186.,3.3.3;35=15,6;,53=125;1.6Fig.6Membershipfunctionbeforetraining4.2,.,.[17],.,,16348.,221,564,93.7,ns.,RCas=130.01350.0108+130.03457796,:7Fig.7SystemcascadingfailurebyMonteCarlosimulation0.0189+130.00570.1250=5.0410-4.,40%,20%,,.20,20,1.,IR.,30%,30%,,.10,10,2.1Tab.1TrainingdataIR/10-4IR/10-415.04114.9925.94126.3534.02134.1643.66144.8754.34152.5464.75165.8376.44173.1184.34184.5894.12196.88103.02203.544.3,10-5.8(a).nt.8(b).68(b),,NB,8Fig.8Trainingerrorandmembershipfunctionaftertrain2ing.,,.,,2,,,,.,,.2Tab.2NetworktestingresultsIR/10-4/%14.995.163.2924.184.241.4234.984.872.2644.704.751.0553.323.116.7565.725.396.1273.903.841.5686.006.416.4093.754.016.48104.634.793.344.4,,.,,103,S.4,,9.94,(1),(4),4(23).9,,879()41310Tab.3Top10protectionequipmentssortbysensitivityS/10-3S/10-311158.296457.632118.177597.4331788.1181516.764958.0691086.5751617.68101316.199Fig.9Resultofprevention1,.9.5,.,,,,.:1),2,;2),,,.(References):[1].[J].,2003,27(18):1-6.XUEYu2sheng.Thewayfromasimplecontingencytosystem2widedisaster[J].AutomationofElectricPowerSystems,2003,27(18):1-6.[2]TANJC,CROSSLEYPA,MCLARENPG,etal.Applicationofawideareabackupprotectionexpertsys2temtopreventcascadingoutages[J].IEEETransactionsonPowerDelivery,2002,17(2):375-380.[3],.[J].,2004,28(9):1-6.HANZhen2xiang,CAOYi2jia.Powersystemsecurityanditsprevention[J].PowerSystemTechnology,2004,28(9):1-6.[4]DOBSONI,CARRERASBA,LYNCHVE,etal.Aninitialmodelforcomplexdynamicsinelectricpowersys2temblackouts[C]HawaiiInternationalConferenceonSystemScience.Hawaii:IEEE,2001.[5]CONS