Wind energy prediction with LS‐SVM based on Lorenz perturbation

Abstract

Owing to the deterioration of the global environment and the depletion of traditional resources, renewable energy has received a high degree of attention. Among them, the fastest growing wind energy has become an excellent alternative to traditional energies. But the non-linearity and volatility of wind speed have brought great challenges to the stability of power system. As a deterministic system, the motion of an aerodynamic system can be described as a set of simple differential equations – Lorenz equation. Thus a small disturbance in the system will have a great impact on the formation of the wind and the wind power prediction work. Therefore, on consideration of the atmospheric dynamical system, a least squares support vector machine (LS-SVM) wind speed prediction model based on Lorenz perturbation is proposed here. The results show that compared with the traditional prediction model (LS-SVM, RBF neural network), the model proposed in this paper effectively weakens the fluctuation of the wind speed sequence and significantly improves the accuracy of short-term wind speed prediction. The research work of this study will reduce the electric running cost and can effectively promote the large-scale development and utilization of the renewable energy.