基于BP神经网络的风电场发电功率短期预测  

Wind Power Short Term Prediction based on Back Propagation Neural Network

作　　者：刘晓楠[1] 王胜辉[1] 金月新[2] 曲可丁 

机构地区：[1]沈阳工程学院电力学院,辽宁沈阳110136 [2]沈阳工程学院产业管理处,辽宁沈阳110136 [3]东北电网有限公司,辽宁沈阳110136

出　　处：《沈阳工程学院学报（自然科学版）》2015年第1期10-13,24,共5页Journal of Shenyang Institute of Engineering:Natural Science

基　　金：国家自然基金资助项目(61372195)

摘　　要：结合我国风电发展的基本情况,分析总结了国内外风力发电功率预测的现状及方法。由于BP神经网络能以任意精度逼近任意非线性映射并且泛化能力强,所以运用BP神经网络法来进行功率预测,建立内蒙某风电场提供的数值天气预报数据与发电功率的映射模型。利用MATLAB进行仿真,验证设计预测模型的实际可行性,并且预测精度满足相关要求。最后运用VB简单设计开发了一个风电功率预测系统。This paper analyze and summarize the current situation as well as methods of forecasting wind power from home and aboard based on wind power development of China. Due to the BP neural network can approximate any nonlinear mapping with any degree of accuracy and it＇ s generalization ability is strong. This paper used BP neural net- work for power prediction, set up a model with numerical weather prediction data and wind power of a wind farm in Inner Mongolia Autonomous Region. Then used MATLAB to simulate and verify the feasibility of this prediction mod- el The precision meet the requirements.

关 键 词：风力发电 功率预测 BP神经网络 预测系统 

分 类 号：TM614[电气工程—电力系统及自动化]