SVM with improved grid search and its application to wind power prediction

Abstract

Wind power prediction is of great significance to the safe and stable operation of the power system. The key factor of wind power prediction is the selection of prediction model. This paper chooses support vector machine (SVM) as the wind power prediction model and applies an improved grid search method to optimize the parameters of C and g in SVM model. The model is able to predict the real-time (15 minutes) wind power, and several evaluation indicators are used to analyze the accuracy of prediction results. The simulation results show that the model has good accuracy which reaches up to 78.49%. An experiment is used to compare the performance of the SVM model based on improved grid search with that of the SVM model only, and results show that the former performs better. For comparative analysis, time series and Back Propagation (BP) neural network were also used for power prediction in the paper, and results show that the SVM model based on improved grid search gets the highest accuracy and is a useful tool in wind power prediction.