基于小波变换与支持向量机的短期电力负荷预测

Short-Term Load Forecasting Based on Wavelet Transform and Support Vector Machine

利用多分辨分析的小波变换对短期电力负荷序列进行分解处理,将负荷序列投影到不同的尺度上,对各子负荷序列根据其特性采用不同的支持向量机进行训练和预测,最后把各预测的结果叠加得到完整的负荷预测结果。算例结果表明该方法同支持向量机的方法相比较具有较高的预测精度和较强的适应能力。

In this paper, the wavelet transform(WT) based on Multi-resolution Analysis(MRA) is applied to decompose short-term power load into wavelet component. By the WT, the different load sequence components are projects to the different scales. According the different character of the load components, the different support vector ma-chine(SVM) is used for forecasting. The forecast results is obtained by adding up all the forecast components. Simulation results demonstrate that the proposed method can offer higher forecast precision.