A Hybrid Model to Forecast Wind Speed Based on Wavelet and Neural Network

To solve the reliability of wind power at the small wind farm and increase forecasting accuracy of wind speed, this paper proposed a hybrid model for forecasting wind speed based on the combination of wavelet transformation and the neural network. The proposed hybrid model to forecast wind speed is a combination of loose and compact wavelet neural networks. By using this model, wind speed signal is decomposed with wavelet transform, and reconstructed to get each scale´s sub-series. Then the sub-series are predicted by compact wavelet neural network, respectively. Compared with other models, the proposed method improves wind speed forecasting accuracy.