风电场风速预测模型研究  

Wind Speed Forecast Model for Wind Farms

作　　者：陶玉飞[1] 李伟宏[1] 杨喜峰[2] 

机构地区：[1]北京木联能软件技术有限公司,北京100085 [2]水电水利规划设计总院,北京100011

出　　处：《电网与清洁能源》2009年第3期53-56,共4页Power System and Clean Energy

摘　　要：介绍了两种风电场风速预测模型,分别是BP神经网络模型和小波-BP神经网络组合模型。BP神经网络模型是风速预测中常用的模型之一,小波技术和BP神经网络结合,即为组合模型。小波技术将风速时间序列按时间和频率两个方向展开,体现了各成分对预测值贡献率的不同。将BP神经网络模型和小波-BP神经网络组合模型分别应用到我国朱日和风电场的逐时风速预测中,从预测结果对比得出组合模型更适合该风电场的逐时风速预测。This paper introduces two wind speed forecast models for wind farm, BP neural network model and wavelet-clustering-neural network model. BP neural network model is usually used model for wind speed forcast, and wavelet-clustering-neural network model is from the combination of wavelet and BP neural network technology.The wind time series are divided into the time and frequency by wavelet technology, which have different contribution to the forecast results. Forecast results of BP neural network model and wavelet-clustering-neural network mode, which used to forecast hourly wind speed in Zhurihe Wind Farm, show that the latter is more suitable for hourly wind speed for wind farms.

关 键 词：风电场 风速 预测模型 BP神经网络 小波 

分 类 号：TM614[电气工程—电力系统及自动化]