基于ARMA-BP神经网络模型的风电功率预测研究  

Wind Power Prediction Based on ARMA-BP Neural Network Model

作　　者：付晓舰[1] 刁述祥[1] 周鑫[1] 王飞[1] 

机构地区：[1]西华大学数学与计算机学院,四川成都610039

出　　处：《西华大学学报（自然科学版）》2012年第3期38-40,56,共4页Journal of Xihua University:Natural Science Edition

摘　　要：结合ARMA时间序列预测模型和BP神经网络预测模型的优点,构建ARMA-BP神经网络混合模型,对风电功率预测问题进行研究。采用构造性神经网络模型得出的类别值对统计时间序列模型的预测值进行修正,提高了风电预测的精度和效率。Wind is a type of renewable energy without pollution and with big and wide developing prospect.It is the first strategic choice for energy development in various countries.Based on time series prediction model and the advantage of ARMA BP neural network model,this paper builds ARMA-BP neural network mixed model for wind power prediction.The statistical category value is obtained by using the structure of the neural network model and used to revise time series prediction model to improve the prediction precision and efficiency of the wind power.

关 键 词：ARMA 时间序列 BP神经网络 风电功率 预测模型 

分 类 号：O29[理学—应用数学] TM614[理学—数学]