基于遗传算法优化BP神经网络的风电功率预测  

Wind power forecast based on genetic algorithm optimization BP neural network

作　　者：王德民[1] 

机构地区：[1]上海海事大学信息工程学院,上海201306

出　　处：《电子设计工程》2013年第22期95-98,共4页Electronic Design Engineering

基　　金：上海市教委科研创新项目(11YZ140);教育部博士点基金项目(20113121110002)

摘　　要：随着大量风电开始并入电网,风电场输出功率预测对接入大量风电的电力系统的运行有重要意义,针对神经网络在风电功率预测中结构和权值参数难以确定,预测精度不高等问题,提出利用遗传算法对神经网络的拓扑结构和网络权值进行优化,并将其应用于风电场功率预测,研究表明预测精度有一定程度的提高。With the large number of wind power incorporated into the power grid, wind farm output power prediction is important for the wind power system operation. For the neural network structure and weight parameters in the wind power prediction is difficult to determine, the accuracy of prediction is not high. Proposed use of the genetic algorithm neural network structure and weight parameters of the network to be optimized ,and applied it into the prediction of wind farm power, studies have shown that the prediction accuracy has improved to some extent.

关 键 词：功率预测 遗传算法实数编码 BP神经网络 优化 

分 类 号：TN919[电子电信—通信与信息系统]