Research on mid-long term load forecasting based on combination forecasting mode

Mid-long term load forecasting for power system is one of the basic works of power planning for cities. Each power forecasting model has its own advantages and disadvantages and has its own application range. In this paper a combination load forecasting model with variable weight is built. In this way it can maximize the advantage of each single model in different ranges. Furthermore, the Fourier technique of the residual correction method is used to decrease the absolute error of forecasting error of combination model. Based on the sample data in a city, the experiments are performed. The results show that the forecasting precision of combination model is higher than any single model which is more than 94%. After residual correction, the forecasting precision is further improved to 95%.