In order to improve the accuracy of ultra-short-term wind power prediction,a chaotic ergodic PSO optimized LS-SVM model for ultra-short-term wind power prediction is proposed.The method uses chaos algorithm to optimize particle swarm op-timization,improves the ability of particles to jump out of the local optimum,and uses chaos traversal particle swarm optimiza-tion LS-SVM kernel function kernel deviation coefficient,gets the desired output.The LS-SVM algorithm based on chaotic er-godic particle swarm optimization is used to establish the wind power ultra-short power prediction model.LS-SVM and LS-SVM optimized by chaotic traversal PSO are used to perform regression tests on UCI functions,which verify that the proposed method has higher prediction accuracy.The proposed method,GA-LS-SVM,LS-SVM,BP and SVM are used to establish wind power prediction models,respectively.The comparison results show that the proposed method has better prediction effect and is more suitable for solving wind power prediction,which provides a theoretical basis for power grid development strategy.
维普
万方数据