Wind Speed Prediction Modeling Based on the Wavelet Neural Network

Wind speed prediction s an important part of the wind farm management and wind power grid connection. Having accurate prediction of short-term wild speed s the basis for predicting wind power. This paper proposes a short-term wild speed prediction strategy based on the wavelet analysis and the muti-layer perceptual neural network for the Dabancheng area, in China. Four wavelet neural network models using the Morlet function as the wavelet basis function were developed to forecast short-term wind speed in January, April, July, and October. Predicted wild speed was compared across the four models using the mean square error and regression. Prediction a:curacy of model 4 was high, satisfying the forecasting wind power industry requirements. Therefore, the proposed algorithm could be applied for practical short-term wild speed predictions.