Short-Term Load Forecasting Based on Bayes and SVR

Abstract

The short-term load forecasting making use of Support Vectors Regression to be in progress forecasts, when the vector dimension is excessive if importing, not only the training time of Support Vector Machine increases by, speed unexpected turn of events is slow, whose generalization performance also will be affected. And adaptively selects input features making use of Bayes method, the characteristic choosing a group of the inputting expecting that output changes cause’s can report most out incorporates the forecasting model training SVR , make use of the SVR model completing to carry out a forecast, and use an example to have testified the validity being method’s turn.