The wind power forecast model based on improved EMD and SVM

摘要

In order to improve the predictive accuracy of short-term wind power, a prediction model based on improved empirical mode decomposition (EMD) and support vector machine (SVM) is constructed. As to the problems of basic EMD, it is proposed to use the steady point meaning sifting method instead of spline envelope meaning sifting method, to improve the overshoots/undershoots caused by traditional cubic spline interpolation. Wind power series can be decomposed into different series by improved EMD, and then SVM is used to forecast power by each component. The total wind power prediction result is obtained through reconstructing at last. Case study shows that the predictive accuracy has significantly been improved by comparing with other models.