Wind Speed Prediction Modeling Based on the Wavelet Neural Network

Abstract

Wind speed prediction is an important part of the wind farm management and wind power grid connection. Having accurate prediction of short-term wind speed is the basis for predicting wind power. This paper proposes a short-term wind speed prediction strategy based on the wavelet analysis and the multilayer 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 wind speed was compared across the four models using the mean square error and regression. Prediction accuracy of model 4 was high, satisfying the forecasting wind power industry requirements. Therefore, the proposed algorithm could be applied for practical short-term wind speed predictions.