In order to improve the accuracy of wind power forecasting

a nonlinear combined model for wind power forecasting based on different optimization criteria and generalized regression neural network（GRNN） is proposed. The grey correlation analysis method is applied to select the forecasting models in which the comprehensive grey correlation values are greater than zero. Based on the three selected models

different linear combined forecasting models are built with the minimum average relative error

minimum mean absolute error

and minimum root mean square error as the optimization criterion

respectively. The GRNN is used to establish the nonlinear combined forecasting model by the three linear combined models and then obtain the optimized model. The verification with the real measured data of a wind farm shows that

compared with the single forecasting models

linear combined forecasting models

the proposed optimized model improves the forecasting accuracy effectively

which proves its effectiveness and practicability.