基于风速分频和权值匹配的RBF超短期风电功率预测方法  

RBF ultra short term wind power forecasting method based on wind speed frequency division and weight matching

作　　者：杨茂[1] 董昊 Yang Mao;Dong Hao(Key Laboratory of Modern Power System Simulation Control and Green Power New Technology Ministry of Education(Northeast Electric Power University),Jilin 132012,China)

机构地区：[1]现代电力系统仿真控制与绿色电能新技术教育部重点实验室(东北电力大学),吉林吉林132012

出　　处：《可再生能源》2020年第11期1483-1488,共6页Renewable Energy Resources

基　　金：国家重点研发计划项目(2018YFB0904200)。

摘　　要：为了提高风电功率超短期预测的准确率,文章提出了一种基于风速分频特征和权值匹配的RBF超短期风电功率预测方法。该方法的总体结构包括风速分频特征的提取、风速区间的划分、径向基神经网络的训练、分频最优权值决策、仿真计算和误差分析。与现有的仅考虑历史风电功率数据的方法相比,该方法能够跟踪未来的功率趋势,物理意义清晰,并考虑了风速不同频率特征对于功率的影响。算例结果表明,基于风速分频特征和权值分配的RBF超短期风电功率预测方法的预测精度较高,预测结果有效,具有较强的实用性。In order to improve the accuracy of ultra short term wind power prediction,the paper develops a radial basis function(RBF)ultra short term wind power prediction method based on wind speed frequency division feature and weight matching.The overall structure of the method includes the extraction of wind speed frequency division feature,division of wind speed interval,training of RBF neural network,decision-making of optimal weight for frequency division,simulation of wind power prediction and analysis of associated error.Compared with the conventional prediction methods which considers historic wind power measurements only,the RBF method proposed here can not only capture the future power trend with a clear physical significance,but also take into account the influences that different frequency characteristics of wind speed have on the power.The results show that the RBF ultra short term wind power forecasting method based on the characteristics of wind speed frequency division and weight distribution has a higher prediction accuracy than conventional prediction methods,offering effective prediction results and strong practicability.

关 键 词：超短期功率预测 风速分频 Yalmip-Cplex权值优化 RBF神经网络 

分 类 号：TK81[动力工程及工程热物理—流体机械及工程]