基于小波变换和AWLS-SVM的短期负荷预测

short-term load forecasting based on wavelet transform and adaptive weighed least squares support vector machine

提出了一种基于小波变换和自适应加权最小二乘支持向量机(AWLS-SVM)的电力系统短期负荷预测方法。针对负荷变化具有拟周期性和随机性的特点,本方法先将负荷值利用小波变换分解为几个低频段的拟周期量和一个高频段随机量,然后根据各分量特点应用AWLS-SVM模型进行预测,最后小波重构各分量获得预测结果。实例预测结果表明该方法具有较高的预测精度。

A new short-term load forecasting approach based on wavelet transform and adaptive weighed least squares support vector machines（AWLS-SVM）is presented.Considering the load having trait of approximate period and random,firstly the load series is decomposed into several lower frequency approximately periodic components and a higher frequency random components with wavelet transform,then different AWLS-SVM models with advantage of automatic adjusting parameter are constructed to forecast the components,finally the forecasted signals of the components are reconstructed to obtain the ultimate forecasting result.The result of load forecasting shows that this method has high accuracy.