A Research of Short-Term Wind Power Prediction Based on Support Vector Regression

Abstract

The wind power prediction is a basis of dispatching automation system. But wind power is often unstable, and this is because wind power is restricted by the wind speed, climate, and other factors. The instability of wind power will influence the safety of the power grid, and the power grid scheduling needs an accurate prediction of wind power. In this paper, it realizes a short-term wind power prediction based on support vector regression. Firstly, it sets the initial clustering center based on historical fan data. Then, it establishes a model of wind power prediction. Lastly, it uses SVR algorithm to achieve the short-term wind power prediction.