An Improved Wavelet Neural Network Method for Wind Speed Forecasting

The randomness and intermittency of wind speed have a great influence on grid security, system stability and economic benefits. Wind speed forecasting plays a key role in tackling these challenges. In order to improve the prediction accuracy, a novel hybrid forecasting model is proposed, which is based on a combination of two types of traditional wavelet neural networks. The proposed hybrid model consists of two parts: the preprocessing module based on wavelet transform and the prediction module based on a kind of wavelet neural network. By wavelet transform, the preprocessing module discomposes and reconstructs an actual wind speed data into an approximation and some details. These subseries obtained are forecasted by the prediction module, respectively. The efficiency of the proposed approach has been evaluated by using four sets of season data randomly selected from a wind farm in North China. Experimental results show that the proposed method can improve the prediction precision of wind speed compared with other approaches according to the root mean squared error (RMSE) and the mean absolute percentage error (MAPE) results.