Application Of Improved Elman＿Adaboost Algorithm In Wind Power Prediction

In recent years,with the continuous attention of the state to energy saving and emission reduction,wind power generation technology is more and more strongly supported by the state.However,the randomness and volatility of wind power generation is the fundamental reason that restricts wind power grid connection.Accurate short-term wind power prediction can assist power grid dispatching department,adjust reasonable and accurate dispatching plan in real time,configure reserve electricity quantity,so as to minimize operation cost.How to achieve more accurate wind power prediction is the focus of this paper.In order to improve the prediction accuracy of wind power and make wind power more accurate access to the power grid,the dispatching of the power grid is more economical and accurate.This paper mainly studies the effect and function of combined algorithm in wind power prediction according to wind power characteristics.The main elements are as follows:(1)Learn and study the mainstream algorithms of wind power prediction,and then predict wind power according to its characteristics.The algorithm can be used to predict the accuracy of wind power BP a certain degree,but the dynamic characteristics of the algorithm are relatively weak.Compared with the BP network,the Elman network has better dynamic characteristics.Finally,the wind power prediction accuracy is higher than the BP algorithm.(2)Using genetic algorithm(GA)to optimize the initial weight and threshold of Elman algorithm and BP algorithm,so that the weight and threshold of wind power prediction can be optimized,which avoids the situation that the algorithm is easy to fall into local optimal solution,and then predicts the wind power accurately.Compared with a single BP algorithm and Elman algorithm,the genetic algorithm is optimized,The prediction accuracy of the algorithm has been greatly improved,but the prediction accuracy of the optimized Elman algorithm is still better than that of the optimized BP algorithm.Although genetic algorithm can improve the prediction accuracy of the algorithm to a great extent,the optimized performance can not exceed the characteristics of the algorithm itself.(3)The Adaboost algorithm can integrate the weak predictor and synthesize the strong predictor to improve the prediction accuracy.Adaboost algorithm is proposed to integrate and improve the GA＿Elman and GA＿BP algorithms and then predict the wind power respectively.Through error analysis and comparison,it can be obtained that the ability of the Adaboost algorithm to integrate the weak predictor makes the prediction accuracy of the final wind power further improved.