基于分层模型预测控制的源荷协调控制策略研究

RenewableEnergyResources

可再生能源

Vol.37No.10Oct.2019

基于分层模型预测控制的源荷协调控制策略研究

渊摘要院袁袁遥袁曰尧袁袁朱广伟袁蔺红

袁曰830047冤袁遥袁尧袁遥袁遥袁遥

关键词院曰曰中图分类号院TK81文献标志码院A文章编号院1671-5292渊2019冤10-1453-06

0

引言-冶SmartAC袁袁袁袁[1]遥遥[2]遥袁袁遥[4]尧遥[5]遥[6]袁遥[7]袁遥袁遥

袁[3]袁袁袁-野袁-袁袁遥渊ModelPredictiveControl袁MPC冤尧尧[8]袁遥[9]MPC袁MPC袁遥[10]袁MPC遥

袁MPC袁MPC袁遥袁袁尧袁遥袁尧袁遥MPC袁收稿日期院2018-09-05遥

基金项目院渊51667019冤曰渊2017D01C029冤遥通讯作者院渊1969-冤袁袁袁袁遥E-mail院Tseagle@163.com

窑1453窑

可再生能源

2019袁37渊10冤

袁袁1分层模型预测控制

袁遥MPC袁尧[8]袁遥3袁袁袁袁遥1.1分层模型预测控制结构

遥

袁野-冶MPC袁1遥

袁[9]MPC袁//渊尧冤

Fig.1图Control1分层structureMPCof的控制结构

hierarchical袁MPC

袁袁袁遥袁袁袁袁袁窑1454窑

遥袁袁袁1.2优化控制策略

袁遥

袁遥袁袁驻TM驻T袁袁遥袁尧袁驻t袁N驻t遥袁2遥

0驻T

0M驻T

驻tN驻t

2Fig.2Rolling图2optimization多时间尺度滚动优化

ofmulti-timescale

2.1基于分层预测模型

MPC的源荷协调控制策略MPC遥袁扇设设设设设设PtNat-1tNbt-1W袁fore=W袁fore-缮设设设移m=1渍mPW袁fore+着移m=1兹m着W袁fore设设设PtN设ct-1t渊1冤

设墒

设设L袁fore=移渍nPL袁fore+着L袁fore-移Nd兹n着t-1n=1n=1L袁fore院Ptt着tW袁fore袁PL袁fore尧W袁fore袁着tL袁fore曰NN尧a袁Ncb袁N曰

d渍2.2m袁兹m袁渍n兹n遥2.2.1多时间尺度滚动优化朱广伟袁等基于分层模型预测控制的源荷协调控制策略研究

袁15min袁4h袁PPw,fore袁PPG遥

HLdisW尧扇设设设设设设设设minJ1=设移t沂T1渊CG袁t+CW袁t+CHL袁t冤

设设设设设C+酌设G袁tGi设=渊PtGi冤2]

设设移NG{Ktti=1Gi[琢Gi+茁GiPGi设设缮设设+KttGi渊1-KGi冤[浊0i+浊1i渊1-e

子/子i设设设设设CNw冤]}

设W袁t设=设设移KtWm琢Wm渊Ptt渊2冤

m=1W袁fore袁m-PWm冤

设设设设设CNHLdis设HL袁t设=设移Kt=1HLdis袁j琢HLdis袁jPtjHLdis袁j设设设墒

设设+NHLconKttk=1HLcon袁k琢HLcon袁kPHLcon袁kt移扇设设u渊t冤=蓘PT缮设设设GT袁PWT袁PHLdisT墒

设设设驻u渊t冤=u渊t+1冤-u渊t冤

蓡渊3冤

院CG袁t袁Cw袁tCHL袁t尧曰琢Gi袁茁Gi袁酌Gi曰浊0i袁浊1i子i曰NG曰琢曰琢WmHLdis袁j袁琢HLcon袁k尧曰驻u渊t冤淤遥

遥袁袁扇设设移N设W设设

PtNGtttW袁fore袁n缮设设n=1+设设移i=1KGiPGi=PL设N设PtHLdis渊4冤设墒

设设L=ttj=1HLdis袁jPNL院PtLPt移PNLt曰PtHLdis袁jjt曰PtHLdis遥袁院

扇设设t-1tt-1tt-1tt-1缮设设设t蓸PW-tPtW蔀+蓸PNL-PNL蔀臆蓸PtHL-PHL蔀+蓸PG-PG蔀墒

设设设PHL=PHLdis+PHLcon于0臆tP渊5冤

tW袁plan臆PW袁fore袁mmax渊6冤

院PW袁plan遥

盂尧院

KtGi袁min臆PttGiP院

Gi臆KGiPGi袁max渊7冤扇设设蓸Kt-1tt设缮设设设Gi-KGi设设t蔀蓸TGi袁on-TGmini袁ont-1t蔀逸0

墒

设设蓸K

Gi-Gi袁off-TGmini袁off蔀逸0

渊8冤

KG院

i蔀蓸T

扇设

设Kttt-1t-1设缮设设设GiPGi-KGiPGi臆PupGi设t-1tt渊9冤墒

设设设Kt-1院PGiPGdownGi袁minPGi袁maxi-KGiPGi臆PGii曰Ktt袁KtGiiKtGi=1TGi=0tt袁

tiGi袁off曰TGi袁on曰TminGi袁onTGmini袁offi曰Ptit曰PupdownGii尧榆Gi袁PGi遥-Plmax臆Ptl臆Plmax渊10冤院Ptllt曰Plmaxl虞愚-P遥

HLdis袁min臆PHLdis袁j臆PTHLdis袁max渊11冤

舆cj逸Tcmintt-1j渊PHLdis袁j屹PHLdis袁j冤

渊12冤

窑1455窑

可再生能源

2019袁37渊10冤

扇设设t设max缮设设设0臆KHLdis袁j臆NHLdis袁j设设tt-1tt-1渊13冤墒

设设KHLdis袁j=KHLdis袁j+1渊PHLdis袁j屹PHLdis袁j冤

院PHLdisminPHLdismax曰PtjtHLdis袁曰TjcjjPt-1HLdis袁j曰Tcjminj

曰KtjtHLdis袁j曰NHLdis袁max2.2.2j遥

尧袁遥袁袁遥袁5min袁15min袁袁PHLcon袁淤遥

minJ2=Pttt沂T2W袁real-PW袁plan渊14冤

院Pt移W袁real遥于-PHLcon袁min臆PHLcon袁j臆PHLcon袁max尧渊15冤

2.2.3袁渊4冤~渊6冤遥袁[11]MPC袁minNe-1s=0渊t+s冤椰2R渊16冤

移椰驻u窑1456窑

嗓umin驻umin臆3算例分析

院s=0袁1袁窑窑窑袁Ne臆u渊-1袁驻ut+N渊s冤臆et+s冤臆umax驻umax渊17冤

遥袁3遥

750kV220kVFig.3Diageam图3of新疆某地区电网示意图

Xinjiangcertainareapowergrid

MPC袁野冶遥3195MW袁2697.75MW袁1200MW遥30%耀50%袁袁300MW1%袁60%40min遥尧MW袁210袁390袁1204袁5袁袁110袁180袁704h袁/渊MW窑h冤袁遥

4遥

18001600140012001000

800

60000院00

04院0008院00

12院00/h

16院0020院0024院00

Fig.4Output图4curve风电预测出力曲线

ofwindpowerforecast

4袁1699.994

朱广伟袁等基于分层模型预测控制的源荷协调控制策略研究

MW袁袁00院00-08院00袁1遥5遥遥

Table表C11Cost系统运行成本

ofCsystemoperating

W/CJ2533.4G/2379.6

11.2

HL/0

1.82544.61/41.9

2423.3

1院2544.6袁CC袁G2533.4袁曰W11.2

121.3C袁袁2379.6袁1.8J12423.3CG袁C41.9WHL袁袁袁遥

30002800260024002200200018001600

1400

1200

1000

00院0004院0008院00渊a冤12院00/h16院0020院0024院00

500450400350300250200150100

00院0004院00

08院0012院0016院0020院00200渊b冤/h

24院00

1801601401201008060402000院000

04院0008院00渊c冤12院00/h

16院0020院0024院00

Fig.5图Power5可调设备的功率曲线curveofadjustabledevice

5袁00院00-08院00袁袁袁袁袁袁23149.46MW窑h袁2911.18MW窑h遥6遥

1800淤淤1600于1400盂12001000于800600

盂00院0004院00

08院0012院00/h

16院0020院0024院00

Fig.6图6Output风电出力曲线curveofwind

10h袁3905.8976MW院窑h遥657.145MW袁袁7.5h袁283.034MW袁3626.827MW窑h曰270.319MW窑h遥袁袁袁袁袁4结论遥袁尧遥袁袁遥袁袁袁遥袁袁袁遥

窑1457窑

可再生能源

参考文献院[1][2][3]

袁袁袁[J].袁袁袁..袁2014袁34渊12冤院13-野--冶[6][7][8][9][10]袁袁袁.袁.[J].2019袁37渊10冤18袁25.袁2018袁42渊5冤院80-85袁119.袁.袁袁袁[J]...-2012袁36渊21冤院1-6.

[J].袁

9.

袁2017袁32渊3冤院1-要impactevaluationforpacificgasandelectriccompany爷ElectricCompany袁2013.渊7冤院74-81.1238.袁袁袁袁.PerryM袁HartmannC袁HartmannL袁etal.2012Load

袁袁sSmartACprogram[R].SanFrancisco院PacificGasand[4][5]

85.[J].袁袁2013袁33渊5冤院79--[J].袁2018袁38-渊9冤院6-14.袁[J].袁2014袁38袁袁[J]..袁2017袁41渊4冤院1230-

[11]ValverdeG袁CutsemTV.Modelpredictivecontrolof

voltages

TransactionsonSmartGrid袁2013袁4渊4冤院2152-2161援

in

active

distribution

networks[J].IEEE

[J].袁2016袁31渊6冤院195-205.

Studyonsource-loadcoordinationcontrolstrategybasedon

hierarchicalmodelpredictivecontrol

渊SchoolofElectricEngineering袁XinjiangUniversity袁Urumqi830047袁China冤

ZhuGuangwei袁LinHong

Abstract院Withthecapacityofwindpowerintegrationincreases袁itsstochasticandvolatilityincreases

theregulationpressureofsystem袁andtheregulationcapacityofconventionalpowersourcesislimited.Itisnecessarytoexcavateadjustableresourcesfrombothsidesofthesourceandtheload.Asource-loadcoordinatedcontrolstrategybasedonhierarchicalmodelpredictivecontrolisproposed.Thecontrolprocessismulti-timescalerollingoptimizationcontrolforlongandshorttimescales.Thelong-timescaleoptimizationcontrolstrategytakestheoptimalsystemoperatingcostbasedonminimumabandonmentastheobjective袁thewindpowerforecastingoutputoflong-timeasthestatevariable袁theconventionalunitpower袁highloadcapacitypowerandwindpoweroutputasthecontrolvariables.Andthequadraticprogrammingmethodisusedtomulti-steprollingsolvethemode.Theshort-timerollingoptimizationcontrolmodeltakesthedeviationthesmallestofbetweentheactualoutputofwindpower

andtheplannedoutputastheobjective袁thewindpowerforecastingoutputofshort-timeasthestatevariable袁andthecontinuouslyadjustedhighenergyloadasthecontrolvariable.Andthecontrolsequenceisobtainedbyrollingoptimizationandmakecorrectionstotheimpactofpredictionerrorserror.Anditusesthequadraticprogrammingmethodtosolvetheoptimalcontrolmodel..SimulationsontheactualpowergridinaXinjiangcertainareademonstratetheeffectivenessandfeasibilityoftheproposedcontrolstrategy.

Keywords院source-loadcoordination曰hierarchicalmodelpredictivecontrol曰highenergyload曰multi-timescalecoordination曰quadraticprogramming

窑1458窑