Wind Power Prediction Based on Cloud Model and GMDH Two-stage Optimization Approach

Currently, large scale of wind power grid-tied has brought great security risk on the grid. Accurate forecast of wind power is of vital importance. In this paper, a forecast method based on cloud model and GMDH two-stage optimization approach is proposed. Firstly, association rules between wind speed and different atmospheric pressure are mined, cloud model inference method is applied to get the data of wind power speed at next time. Then, many forecast methods were applied to acquire power forecast data at next time, furthermore, combination forecast model based on GMDH is acquired. Through analyzing historical data, different atmospheric pressure and generation power of some wind farm in Jibei, it is verified that this model, having high practical value, can improve the accuracy of power prediction.