物理与数据双驱动水风光短期功率联合预报研究  

Joint forecast for short-term power of hydro-wind-photovoltaic system

作　　者：徐伟峰 徐长江 刘攀[2] 赵杨坤锳 XU Weifeng;XU Changjiang;LIU Pan;ZHAO-YANG Kunying(Bureau of Hydrology,Changjiang Water Resources Commission,Wuhan 430010,China;State Key Laboratory of Water Resources Engineering and Management,Wuhan University,Wuhan 430072,China;School of Mathematics and Statistics,Beijing Institute of Technology,Beijing 102401,China)

机构地区：[1]长江水利委员会水文局,湖北武汉430010 [2]武汉大学水资源工程与调度全国重点实验室,湖北武汉430072 [3]北京理工大学数学与统计学院,北京102401

出　　处：《人民长江》2024年第10期55-61,共7页Yangtze River

基　　金：国家重点研发计划项目(2022YFC3202804);国家杰出青年科学基金项目(52225901)。

摘　　要：短期功率预报是保障可再生能源电站稳定运行的重要举措。针对多能互补系统功率预报难以考虑不同能源间时空相关性和预报误差互补特性的问题,提出了物理与数据双驱动水风光短期功率联合预报模型。采用WRF-新安江模型驱动气象水文要素预报,通过互信息方法识别功率预报的关键因子。基于长短期记忆网络构建功率联合预报模型,并将该方法应用于雅砻江流域官地水风光互补系统。结果表明:①相较于独立预报模型,物理与数据双驱动的短期联合预报模型总功率预报平均绝对误差和均方根误差分别降低了9.62%和8.31%;②由于预报误差存在互补特性,系统总功率预报误差小于各电站功率预报误差之和;③相对于单独预报模型,水风光预报误差互补率提高了10.96%。所提方法可为水风光系统协同运行提供技术支撑。Short-term power forecast is an important way to guide operation of renewable energy stations.A joint short-term power forecast model for hydro-wind-photovoltaic system is proposed for problem of hardly considering the spatial-temporal correlation of different energies and the complementary characteristics of forecast errors in power forecast of hybrid energy system.The WRF-Xin'anjiang coupling model is used to drive joint meteorological and hydrological elements forecast.The joint forecast factors are identified by the mutual information method.The joint power forecast schemes are constructed based on long short-term memory network.The complementary rate of forecast errors is proposed to evaluate the power forecast accuracy.Guandi hydro-wind-photovoltaic complementary system in Yalong River basin is considered as a case study.Results indicate that compared to independent forecast model,the joint forecast model driven by physics and data reduced the average absolute error and root mean square error of total power forecast by 9.62%and 8.31%,respectively.The total system power forecast error is lower than the summation of forecast errors for each power station due to the complementary characteristics of forecast errors.Compared with the independent forecast model,the complementary rate of forecast errors increases by 10.96%.The proposed method can provide technical support for the coordination of hydro-wind-photovoltaic system.

关 键 词：水风光系统 功率预报 联合预报 预报误差互补 官地水电站 雅砻江流域 

分 类 号：TV213[水利工程—水文学及水资源]