A combined prediction method of wind farm power

Wind power forecasting has great significance to the connection of wind farms to the electric power system. This paper analyzes individual forecast models, such as the time series forecasting, Elman network forecasting that based on the chaos theory, grey neural network forecasting, and generalized regression neural network forecasting, etc., then puts forward an entropy weight combination prediction model, and an optimal combination forecasting model for the wind power forecasting that based on vector angle cosine. The forecasting results indicate that due to the different forecast precisions of different methods, the methods with high precisions may bring great variation in some points, and the combination forecast can reduce the forecasting variation in several points, which improve the forecasting precision.