Page 71 - 2024年第55卷第12期
P. 71
[13] 武昭原,周明,王剑晓,等.双碳目标下提升电 力 系 统 灵 活 性 的 市 场 机 制 综 述 [J].中 国 电 机 工 程 学 报,
2022,42(21):7746 - 7764.
[14] 卢佳,李刚,吴滇宁,等.基于 IGDT的梯级水电多元市场下中期运行风险度量方法[J].中国电机工程学
报,2021,41(19):6581 - 6590.
[15] 于旭光,李刚,李亚鹏,等.耦合日合同分解及日前市场竞价的梯级水电站短期鲁棒优化调度[J].电网
技术,2021,45(10):4016 - 4025.
[16] 杨蕊,李刚,卢佳.考虑分时段合约的梯级水电站中期分段鲁棒优化调度方法 [J].中国电机工程学报,
2023,43(4):1481 - 1492.
[17] ALIPOURP,MUKHERJEE S,NATEGHIR.Assessingclimatesensitivityofpeakelectricityloadforresilient
powersystemsplanningandoperation:AstudyappliedtotheTexasregion[J].Energy,2019,185:1143 - 1153.
[18] JANGY,BYONE,JAHANIE,etal.Onthelong - term densitypredictionofpeakelectricityloadwithdemand
sidemanagementinbuildings [J].EnergyandBuildings,2020,228:110450.
[19] 曲凯,李湃,黄越辉,等.面向新能源消纳能力评估的年负荷序列建模及场景生成方法[J].电力系统自
动化,2021,45(1):123 - 131.
[20] 魏聪颖,汪,徐浩,等.计及广义负荷不确定性和激励型需求响应的电力现货市场竞价方法[J].电力
自动化设备,2022,42(7):76 - 83.
[21] XUW,WANGZ,SUNJ,etal.Researchonloadcharacteristicdataclusteringbasedontransferlearning[J].
JournalofPhysics :ConferenceSeries,2022,2166(1):12008.
[22] 刘莉,王刚,翟登辉.k - means聚类算法在负荷曲线分类中的应用 [J].电力系统保护与控制,2011,39
(23):65 - 68.
[23] 陈烨,吴浩,史俊,等.奇异值分解方法在日 负 荷 曲 线 降 维 聚 类 分 析 中 的 应 用 [J].电 力 系 统 自 动 化,
2018,42(3):105 - 111.
[24] 张巍峰,车延博,刘 阳 升.电 力 系 统 可 靠 性 评 估 中 的 改 进 拉 丁 超 立 方 抽 样 方 法 [J].电 力 系 统 自 动 化,
2015,39(4):52 - 57.
Acoordinateddecompositionmethodforpowertradedincascadehydropower
stationsinresponsetomulti - scaledloads
1
1
1
2
2
HANXu,SHENJianjian,CHENGChuntian,SHUWeimin,RUANYanyun
(1.InstituteofHydropowerandHydroinformatics,DalianUniversityofTechnology,Dalian 116000,China;
2.ChinaYangtzePowerCo.,Ltd.,Wuhan 430014,China)
Abstract:Runoffandloaduncertaintyexacerbatethedifficultyofschedulingandoperationofmulti - receiver - end
gianthydropowerplantclustersinSouthwestChina ,whichgreatlyaffectstheperformanceofhydropowerplants’
medium -andlong - termtradedpower.Inthispaper,apowertradingcoordinationmethodisproposedformulti -
receiver - terminalandmulti - scaledifferentialloadsofthecascadehydropowerplants ,andarunoffscenariogener
ationmethodbasedonLatinhypercubicsamplingandK - meansclusteringisconstructedconsideringthelong - term
runoffpredictionerror.Byintroducingloadcharacteristicindexes ,atypicalloadfeatureextractionmethodbased
onnonparametricestimationisproposedtoaccuratelycharacterizethedifferentialloadsofthepowergridsofdiffer
entreceiverterminals.Meanwhile,inordertoavoidtheunbalancedinfluenceonpowerallocationandcurvede
composition ,theloadcurveisreconstructedbyusingthepeak - to - valleydifferencerateandtheproportionof
powerdelivery ,andamulti - objectivemodelisconstructedtomaximizetheexpectedrevenuefrom thepeakingat
thereceivingendofthegrid,andtheplannedpowerperformanceoftheterracedhydropowerplants.Avalidation
analysisiscarriedoutontheexampleoftheactualprojectofsendingpowertosevenprovinces(municipalities)
fromfourgiantterracehydropowerstationsinthemainstream ofamega - basininSouthwestChina.Theresults
showthattheproposedmethodcanrealizethereasonabledistributionofannualplannedpowerinthefloodanddry
periodsandamongprovinces.Atthesametime ,itisfoundthattheinterplaybetweenthegridpeakdemandandthe
powerstation’spowergenerationrevenueisconducivetothecoordinationofthehydropowerconsumptionprogram.
Keywords:electricitymarket;powerdecomposition;plannedelectricity;cascadehydropowerplants
(责任编辑:耿庆斋)
4
— 1 8 3 —