Wind power prediction has been the main research subject of wind power generation.In order to improve the rapidity and accuracy of wind power prediction,based on least absolute shrinkage and selection operator(Lasso),and variational mode decom-position(VMD)algorithm,feature selection and feature extraction of historical wind power time series are carried out,and wind power is predicted by deep learning method.The results show that the deep belief network(DBN)prediction method combined with Lasso and VMD algorithm reduces the modeling time by 11.83 seconds and improves the accuracy by 4.62％compared with that simply using the original wind power time series or other models with the same data.
维普
万方数据