This paper analyzes the importance of the wind farm wind speed prediction, as well as the different forecasting methods in various fields. And established the RBF neural network forecasting model can forecast one hour ahead of the wind farm wind speed, and the results meet the actual forecast requirements. By reconstructing the wind speed time series, wind speed can be predicted one day in advance, the prediction accuracy and one hour ahead of forecast accuracy is similar. The method can predict a longer period of time the wind speed, and provide important reference for the wind farm power generation scheduling.

No abstract