Forecasting of fluctuations and turning points of power demand in China based on the maximum entropy method and ARMA model

Influenced by the economic cycle, power demand in china shows some cyclical fluctuations, which is unhealthy for the development of national economy and production efficiency of electric power industry. Correctly forecasting the fluctuation rule of power demand and the turning points in China is helpful to make the corresponding strategy complied with the cycle. With the full consideration of power demand fluctuations, the paper establishes a forecasting model based on maximum entropy method and ARMA model: firstly, the paper makes a spectrum analysis on the growth rate of power demand and the major cycle of the cyclical fluctuations can be correspondingly obtained, then a periodic function which can reflect the fluctuation features is employed through the least square method; secondly, the paper establishes an ARMA model on the residual series which can be obtained by eliminating the periodic sequence from the original series; at last, the hybrid forecasting model is obtained by combining the periodic function and ARMA model. Experimental results show that the proposed model is effective and reasonable.