A novel hybrid approach for wind power forecasting

Abstract

This chapter proposes a hybrid approach for wind power forecasting based on autoregressive integrated moving average and back propagation neural network, which employs a new combination mode. The model took into account the wind speed, wind direction, and the physical limitations of the wind plant. Compared with the real data of a wind plant, the analyzing result shows that the proposed method improves the accuracy of wind power prediction effectively, and it also has good practical value. © 2014 Springer Science+Business Media New York.