Application of Recursive Right Combination Model in Mid-Long Term Load Forecasting

In view of the problems that the single forecasting in power load forecasting model cannot deal with the uncertain information effectively and it has poor adaptability, exponential smoothing forecasting method, multiple linear regression and grey GM (1,1) model, three kinds of single forecasting model are used in the paper. And then a recursive right combination forecasting model is proposed to give each method variable weight. The correlation degree error analysis method is used to verify the effectiveness of combination forecasting model, the results show that the combination forecasting model based on recursive right combination forecasting model has smaller error than the single models, and its precision is the highest of all, which indicates that the combination forecasting model is better than the single models. Eventually, this paper forecasts the power consumption of Beijing for the next six years.