基于局部加权线性回归模型的风速短期预测研究

Short-term Wind Speed Forecasting Based on Local Weighted Linear Regression

为提高风速预测的准确性,引入局部加权线性回归算法,建立风速局部加权线性回归预测模型。该模型首先通过核函数对每一历史风速设置权重,产生新的风速数据集。其次,对新的数据集进行最小二乘线性拟合回归,建立风速局部加权线性回归模型。最后,根据该模型对风速进行短期预测。以某风电场的实测数据进行仿真验证,并与传统的ARIMA模型和BP神经网络预测模型对比,实验结果表明本文所提方法的有效性,为风速预测提供一种新方法。

In this paper,a short-term wind speed forecasting method based on local weighted linear regression is proposed,which is produced new wind data set by kernel function.Then a linear regression model is build based on the new set and apply the model to forcesting the short-term wind speed.Short-term wind speed forecasting based on local weighted linear regression model is established and simulated with the measured data of a wind farm,and compared with the traditional ARIMA model and BP neural network prediction model.The result shows the model proposed in this paper is efficacy,and provide a new method for the wind speed forecasting.