Research On Short-term Wind Power Prediction Optimization Algorithm

Wind power prediction plays an important role in reducing wind power grid connection and improving the stability of wind power generation.In this paper,a wind turbine as the research object,the acquisition unit actual operation data and historical information data base,the wind turbine history information after data preprocessing,using BP neural network algorithm and support vector machine based wind power prediction power model,and aiming at the disadvantages of quantity model in prediction,respectively.The introduction of numerical weather prediction method,using clustering analysis method and artificial fish swarm algorithm to optimize BP neural network prediction model and support vector machine prediction model.This paper Summarizes the wind power prediction related factors and classification methods,detailed analysis of the distribution characteristics and variation of the wind,the wind in the analysis of the characteristics of wind speed measurement method and prediction method,and an overview of wind power prediction modeling process,through the different angle to the wind power prediction method for classification.Introduces the implementation of wind power prediction process prediction method using the BP neural network prediction method of the unit of numerical weather prediction BP neural network short-term wind power based on research,and based on the BP neural network modeling in the process of training data processing method is single,lack of historical data and numerical weather prediction of wind power in history the data,presented by the cluster analysis method to classify the training data,BP neural network method will input the classified data as.Taking the historical data and the actual output power of a wind turbine as the object,through the simulation experiment in Matlab,the prediction results and the optimization results of the two algorithms are verified.Reseaches the basic method of wind power prediction model of support vector machine,support vector machine regression analysis principle and the kernel function SVM algorithm,parameter optimization of support vector model,analyzes the basic principle of artificial fish swarm algorithm,and uses artificial fish swarm algorithm to optimize the model of support vector machine.Finally through the simulation experiment.The optimization effect of the proposed algorithm.