夏季短期电力负荷 ARIMA - SVR 组合预测模型.

Since summer short- term load has the characteristics of fluctuation and nonlinearity, easily affected by temperature, day type and other factors, it is difficult to get the accurate result, relying on single traditional forecasting model. This paper established the ARIMA- SVR combination forecasting model,based on the advantages of ARIMA and SVR, through which the data was linearly fitted. Then the error of ARIMA forecasting was corrected by the SVR forecasting model with the optimized parameters by particle swarm optimization. In the context of the practical cases, the combination forecasting model was used to forecast the trend and to analyze the errors of the summer short- term load. The experimental results indicate that the ARIMA- SVR combination model has higher prediction accuracy than the single model does, which also has a high application value in the forecasting of electricity load. [ABSTRACT FROM AUTHOR]