Short-term Load Forecasting Based on Embedded Support Vector Machine Combined with DB4 Wavelet

A method based on DB4 wavelet and support vector machine was proposed. Adapted wavelet soft-thresholding method was used to preprocessing load data according to the principle of wavelet denosing. Denoised series was decomposed using DB4 wavelet, each sub-series was embedded a matching support vector machines. The result which was modeled using support vector machine directly and the above result was compared simulation show the effectiveness of the proposed method