风电功率概率预测方法及展望  

Methods and Prospects for Probabilistic Forecasting of Wind Power

作　　者：吴问足 乔颖[1,2] 鲁宗相[1,2] 汪宁渤[3] 周强[3] 

机构地区：[1]清华大学电机工程与应用电子技术系,北京市100084 [2]电力系统及发电设备控制和仿真国家重点实验室,清华大学,北京市100084 [3]甘肃省电力公司风电技术中心,甘肃省兰州市730050

出　　处：《电力系统自动化》2017年第18期167-175,共9页Automation of Electric Power Systems

基　　金：国家自然科学基金重大项目(51190101);国家重点研发计划资助项目(2016YFB0900101);国家电网公司科技项目(522727160002)~~

摘　　要：风电功率的概率预测能提供风电功率的预测区间或分布函数,国内相关的研究和应用尚处于起步阶段。文中对风电功率概率预测的基本框架、主要模式、难点和热点进行了综述。首先,明确了概率预测的概念及其适用问题。然后,对概率预测的建模方法提出了两种不同的分类方式:按照是否进行条件化假设或参数化假设进行分类,并介绍了概率预测中涉及的新型算法和概率预测的评价指标。最后,结合概率预测发展现状,针对误差分析不精细、概率预测与电力系统结合不充分等不足,总结了今后的发展方向和需要进一步探索的研究内容。Prediction intervals or distribution functions of wind power in the future can be provided through probabilistic forecasting of wind power.Relevant research in China is still at the early stage.This paper gives a comprehensive review on the basic approaches,typical patterns and key problems in probabilistic forecasting of wind power.Firstly,the definition of probabilistic forecasting is presented and its applicable problems are summarized.Secondly,two different classification methods are introduced：conditional classification method and parametric classification method.New algorithms and evaluation indices used in probabilistic forecasting are also described.Finally,according to the-state-of-the-art of probabilistic forecasting,the shortcomings of error analysis and the insufficiency of probabilistic forecasting in combining with the power system,the future key issues and the research content which needs further exploration are summarized.

关 键 词：风电功率预测 概率预测 不确定性 预测误差建模 

分 类 号：TM614[电气工程—电力系统及自动化]