Fairly accurate forecast of wind power is an effective means for improving power system security and economy. Based on an analysis of the principle of the ridgelet neural network， the network is applied to the forecast of wind speed， wind direction and wind power. Two forecast models are developed to predict wind speed and wind direction， respectively， and the nonlinear neural network is applied to the approximation of an actual power curve. Finally， the wind power is calculated according to the forecasted wind speed， wind direction and the power fitting curve. Simulation results show that， compared with the wavelet neural network， BP neural network and RBF neural network， the ridgelet neural network is found to yield a higher accuracy of wind power forecast than all the other three.