多时间尺度协调的区域控制偏差超前控制方法

333Vol.33No.320092PowerSystemTechnologyFeb.20091000-3673200903-0015-05TM767A470⋅40541111221()1000842610041AMulti-TimescaleCoordinationMethodforAreaControlErrorAdvancedControlLIYu-zhou1WUWen-chuan1ZHANGBo-ming1JIANGMu1XIAOLan2LUYi21.StateKeyLabofControlandSimulationofPowerSystemsandGenerationEquipments(Dept.ofElectricalEngineeringTsinghuaUniversity)HaidianDistrictBeijing100084China2.SichuanElectricPowerDispatchingCenterChengdu610041SichuanProvinceChinaABSTRACT:Amulti-timescalecoordinationmethodforACE-advancedcontrolisproposed,whichcanconsiderthedifferentresponsespeedofvariousgenerators.Firstly,thedeficienciesoftheexistingareacontrolerror(ACE)-advancecontrolmethodsareanalyzedandanovelmodeltoforecasttheidealtotalpoweroutputwithinacontrolareaisdeveloped;secondly,therelationbetweentheschedulingofscheduledgeneratorsandtheexecutedresultsareanalyzed,andanalgorithm,whichisusedtopredicttheerrorbetweentheschedulingandtheactualoutputofthescheduledgenerators,isdeveloped;andthen,consideringtotaloutput,theoutputofscheduledgeneratorsandthedesiredoperationpointofAGCgenerators,thedesiredtotaloutputofcoordinatedAGCgeneratorsattargettimeiscalculated;finally,bymeansofanoptimalmodelthedesiredtotaloutputofcoordinatedAGCgeneratorsisallocatedtoeachcoordinatedAGCgenerator.Theproposedmethodisputintoon-sitetrailoperationinSichuanprovincialpowergrid,andoperationresultsprovethattheproposedmethodisfeasible.KEYWORDS:autogenerationcontrol(AGC)decompositionandcoordinationcontrolperformancestandard(CPS)loadforecastingareacontrolerror(ACE)(areacontrolerrorACE)ACE(50707013)ProjectSupportedbyNationalNaturalScienceFoundationofChina(50707013)(autogenerationcontrolAGC)AGCAGCAGC(AGC)(CPS)(ACE)0(ACE)(AGC)ACEACECPS1CPS2(CPS)[1-9][3-5]AGCACECPS[4-9]2AGCACEAGCACE[4-8]AGCACEAGCACE[9]2ACE1ACE16Vol.33No.3Tnf=49HzPD50Hz2AGCAGC5%AGCAGCAGCACE1ACE1.13AGC(coordinationAGCCAGC)AGC11Tab.1ClassificationofgeneratorsAGC15min~1dCAGCACE1~5minAGCAGCAGCACEAGCACEAGCCAGCACEAGCCAGCAGC1.2CAGCAGCCAGCACEAGCAGCAGCAGCAGCAGC2AGC2.1CPSf=50HzPTSTP*GPtn*GP*SGGTT()()()*()()nnnLnnPtPtftKPtPt=−∆+−(1)PG(tn)∆f(tn)KLSTPPTtntn+1tn+1*G1()nPt+0,...,ntt**G0G(),...,()nPtPttn+1*G1()nPt+*FGP(tn+1)tn+1STP(tn+1)(1)GGLT()()()()nnnnPtPtftKPt′=−∆−(2)(2)G0G(),...,()nPtPt′′tn+1*G1()nPt+*FG1()nPt+(3)tn+1*G1()nPt+*FG1()nPt+*F*FSG1G1T1()()()nnnPtPtPt+++=+(3)12.2AGC1.1*FG1()nPt+*FSBG11C1A1()()()()nNnnnPtPtPtPt++++=++(4)PN(tn+1)tn+1tnPN(tn+1)333171110980:006:0012:0018:0024:00P/GW1Fig.1Comparisoncurvesofidealgeneration’sforecastedpowerandloadpowerFN1()nPt+SC1()nPt+CAGCtn+1BA1()nPt+CAGCAGCCAGCS*FFBC1G1N1A1()()()()nnnnPtPtPtPt++++=−−(5)CAGCAGC2.32.3.1122.3.2[10]tkNdsr()()()()kkkkPtytytyt=++(6)PN(tk)yd(tk)SN1()nPt+ys(tk)5~10minyr(tk)AR(p)[11-13]r1r1r()()()kkpkpytytytφφ−−=++(7)yr(t1)yr(t2),…,yr(tn)npφ1,φ2,…,φpr1r2r1r1r2r3r12r2rr1r1r()()()()()()()()()()()()ppppnpnpnpnytytytytytytytytytytytytφφφ+++−−+−⎡⎤⎡⎤⎡⎤⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥=⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥⎣⎦⎣⎦⎣⎦(8)T1T()−=AAAZφ(9)AZpr11ˆ[()()][()()]nkkirikirytEYytEYn−+==−−∑(10)0,1,2,,1kn=−r1r2r[(),(),...,()]nYytytyt=0ˆˆˆ/,0,1,2,,1kkrrkn==−…ρ(11)0111112222120kkkkkkkkkρρρφρρρρφρρρρφρ−−−−⎡⎤⎡⎤⎡⎤⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥=⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥⎢⎥⎣⎦⎣⎦⎣⎦(12)(12)Yule-Walker1,1111,11,1,11,,1,1,11kkkkjjkkkjkjjkjkjkkkkj++=++=++++−=⎧⎪⎪−⎪⎪=⎨⎪−⎪⎪=−⎪⎩∑∑φρρρφφρφφφφφ(13)Box-Jenkin[14]AR(p)kpφk,k0pφk,kp01,12,2,,,,pppppmpmφφφ++++++…(mn=),kkφ≤21/2,11(12)Ptttnφ=+∑68.3%1,2,3,,1pp′=−…p′pp′=2.3.3tk18Vol.33No.3217:00109.69.216:0016:4017:2018:00P/GW2Fig.2Resultsofnormalforecastmethodstkgn1gSSSNNNN()()(),0,1,2,...,()()()kikigkikikigkiPtPtPtinPtPtPt−−−−−−⎧=−⎪=⎨=−⎪⎩(14)2gSSSNNNN()()(),0,1,2,...,()()()kikigkikikigkiPtPtPtinPtPtPt−−−−−−⎧=+⎪=⎨=+⎪⎩(15)2.3.4324216:0016:4017:2018:00P/GW9.89.69.29.43Fig.3Resultsofmodifiedforecastmethod0:008:0016:0024:00P/GW11.59.56.57.510.58.54Fig.4Resultsofmodifiedforecastmethodforoneday3CAGC2AGCCAGC()CAGCCAGC1min{[()]}SggngCfPt+∈∑(16)S11s.t.()()()gngngnPtPtPt++=+∆SSC11S11()()()|()|||ngngCllkkggnggnggPtPtPPPPPPtPPtVT++∈++=≤≤≤≤∆≤∑,,lLkKgC∈∈∈CCAGCLKfggSgPgPgg∆PgglPtPgPgPVggTg[15]44.15288AGC5minAGCCAGC5min5min*FG1()nPt+FN1()nPt+5minAGCFA1()nPt+FA1()nPt+A1()nPt+F*FFA1G1N1()()()nnnPtPtPt+++=−(17)6ACEAGCAGCACECAGC33319AGCAGC0:008:0016:0024:00P/GW5.51.53.55AGCAGCFig.5DailycurvesofAGCgenerators’forecastedpoweroutputandAGCgenerators’realpoweroutputAGCACE8:0010:0012:0024:00P/MW200−20006AGCAGCACEFig.6DailycurvesofACEandthedifferencebetweenAGCgenerators’forecastedpoweroutputandrealpoweroutputCAGCACE4.24.2.1ACE3CAGC22CAGCTab.2ParametersofCAGCunits/(MW/min)/MW34.54004.2.2CAGCCPS313:0018:00CAGCCPS1CPS2CPS3CPSTab.3ImpactsforCPS12CPS1189.2195.2194.6CPS22.33.4−2.24.2.3CAGCAGCAGCCAGCAGCAGC30%CAGCA1AAAAAAAAAAAA()(),()30%()31%(),()30%()BnnnnPtPtPtPPPPPPPtPPP+=⎧−−⎪⎨−+−≤−⎪⎩(18)APAPAGC(AGC)7CAGCCAGCAGCCAGCAGCCAGCAGCAGC13:0014:4016:2018:00η/%90603007CAGCAGCFig.7AssistanceofCAGCtoAGCCGACAGCCAGC5ACECAGCAGCCPS[1]JaleeliNVanSlyckLSEwartDNetalUnderstandingautomaticgenerationcontrol[J]IEEETrans.onPowerSystems19927(3)1106-1122[2]WoodAJWollenbergBFPowergenerationoperationandcontrol[M]BeijingTsinghuaUniversityPress2003328-363[3]EgidoIFe