Abstract: In order to solve the problem of low accuracy of wind power prediction caused by the volatility and randomness in wind power prediction,an integrated model combined with the least squares support vector machine（LSSVM）and an sparrow search algorithm based on variational mode decomposition（VMD）. Tent chaotic mapping and random walkis proposed. First,the whale optimization algorithm（WOA）is used to automatically optimize the core parameters（K value and penalty coefficient α）of VMD. After original wind power time series is decomposed by WOA-VMD,the improved sparrow search algorithm SSA is introduced to optimize the learning parameters of LSSVM,and then the SSA-LSSVM prediction model is established for each subsequence obtained by the decomposition.Finally,the prediction value of each subsequence is superimposed to get the final predicted value. Compared with the existing single prediction models and the general integrated models in simulation experiment,the proposed integrated model has a great improvement in the prediction accuracy.