Skip to Main Content
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.