基于D-S证据理论的短期风速预测模型  

Short-term wind speed forecasting model based on D-S evidence theory

作　　者：刘亚南[1] 卫志农[1] 朱艳[2] 孙国强[1] 孙永辉[1] 杨友情[3] 钱瑛[3] 周军[3] 

机构地区：[1]河海大学可再生能源发电技术教育部工程研究中心,江苏南京210098 [2]国电南瑞科技股份有限公司,江苏南京210061 [3]安徽省电力公司池州供电公司,安徽池州247000

出　　处：《电力自动化设备》2013年第8期131-136,共6页Electric Power Automation Equipment

基　　金：国家自然科学基金资助项目(50877024;51107032;61104045)~~

摘　　要：提出一种基于D-S证据理论的短期风速组合预测模型。分别采用时间序列、BP神经网络和支持向量机预测模型对风速进行预测,通过对预测误差的分析,借助D-S证据理论对3种模型进行融合。选取待测日前几日的风速数据作为融合样本,计算出相应的基本信任分配函数,同时将函数进行融合,并将融合结果作为风速预测模型的权重,得到待预测日的风速预测结果。仿真结果表明,所提组合预测模型的预测误差更小,效果更好。A combined short-term wind speed forecasting model based on D-S evidence theory is proposed. The forecasting models of time series,BP neural network and support vector machine are adopted to respectively forecast the wind speed. Based on the analysis of forecast errors,D-S evidence theory is applied to fuse these three models. The wind speed data for several days before are taken as the fusion samples to calculate the corresponding basic trust distribution functions,which are then fused. The results of fusion are taken as the weights of the wind speed forecasting model and the wind speed of the day to be forecasted is calculated. Simulative results show that,the proposed combined forecasting model has smaller forecasting error and better effect.

关 键 词：风电 时间序列 BP神经网络 支持向量机 D-S证据理论 预测 模型 

分 类 号：TM715[电气工程—电力系统及自动化] TM614