计及出力水平影响与自相关性的风电预测误差模拟方法  

Wind-power forecasting error simulation considering output level and self-correlation

作　　者：杨建[1] 张利[1] 王明强[1] 韩学山[1] 

机构地区：[1]山东大学电网智能化调度与控制教育部重点实验室,山东济南250061

出　　处：《电力自动化设备》2017年第9期96-102,共7页Electric Power Automation Equipment

基　　金：国家自然科学基金资助项目(51407106)~~

摘　　要：风电的预测误差不仅受预测出力水平的影响,而且时序上也存在显著的自相关性,基于此,提出一种兼顾两方面特征的风电预测误差模拟方法。充分利用Copula函数在描述变量间相关性方面的优势,通过分析风电出力的预测值与实际值间的联合概率分布,确定不同预测出力水平下预测误差的条件分布函数,对相邻时刻预测误差的相关性进行建模,并结合预测误差的条件分布函数,生成具有特定相关关系的备选数据列。顺次以前一时刻的预测误差为依据,从备选数据列中选取相应数据组成预测误差的模拟序列。算例仿真验证了所提方法的有效性。The forecasting error of wind power is affected by the forecasted output level and presents obvious self-correlation in time sequence,based on which,a method of wind-power forecasting error simulation considering both aspects is proposed. The advantages of Copula function in describing the correlation among variables are fully explored. The joint probability distribution of forecasted and actual wind-power outputs is analyzed to determine the conditional distribution function of forecasting error at different output levels. The correlation of forecasting error between adjacent instants is modelled and,combined with the conditional distribution function,a candidate data list is generated. Corresponding data are picked up in turn from the candidate data list according to the forecasting error of previous instant to form a simulation sequence of wind-power forecasting error. The effectiveness of the proposed method is verified by case simulation.

关 键 词：风电 预测误差 出力水平 自相关性 COPULA函数 

分 类 号：TM614[电气工程—电力系统及自动化] TM761