Short-term wind power prediction based on intrinsic time-scale decomposition and LS-SVM

Abstract

This paper proposes a wind power prediction method based on intrinsic time-scale decomposition (ITD) and least square support vector machine (LS-SVM) to improve the accuracy of wind power forecast. The proposed method employs ITD as a preprocessing method to decompose wind power data into a set of proper rotation components and a monotonous baseline signal. Afterwards, the backward difference of each component is used as input of the LS-SVM model for training and prediction. Simulation studies are carried out on wind power data to evaluate the performance of the proposed method, and the results have shown that, by introducing the ITD, the proposed method outperforms the original LS-SVM method.