A SVM and variable structure neural network method for short-term load forecasting
暂无分享,去创建一个
This paper put forward a new method of the SVM and variable structure artificial neural network model for short-term load forecasting. The neural call function is basis of nonlinear wavelets. We overcome the shortcoming of single train set of SVM. It can be seen from the example this method can improve effectively the forecast accuracy and speed. The forecast model was tested and the result showed that it was an effective way to forecast short-term electric load.
[1] Dirk Tomandl,et al. A Modified General Regression Neural Network (MGRNN) with new, efficient training algorithms as a robust 'black box'-tool for data analysis , 2001, Neural Networks.
[2] A. C. Liew,et al. Demand forecasting using fuzzy neural computation, with special emphasis on weekend and public holiday forecasting , 1995 .