Study On Dynamic Game Charging Method Of Electric Vehicle Based On Typical Loads

Under the dual pressure of environmental pollution and resource shortage,electric vehicles have attracted more and more attention because of their huge advantages in energy conservation and emission reduction.With the continuous maturity of relevant technologies,large-scale application of electric vehicles has become an inevitable trend in the future automotive industry.Large-scale electric vehicles are connected to the power grid through various charging facilities for charging,which will inevitably have a certain impact on the power grid,and it is of great significance to quantitatively evaluate the impact of charging loads on the power grid and effectively eliminate this problem.Therefore,based on the study of relevant literature,the innovation work in this thesis is as follows:First,a typical load pattern extraction model based on data mining theory was proposed in view of the influence of complex power loads in the power grid on the orderly charging of electric vehicles.This model was based on similarity,and there was more similarity between the same models,so complex power load data could be classified into several typical load types and contain all the information of power load data.It could directly establish the corresponding relationship with the data information.Second,in view of the traditional k-means clustering algorithm(K-means)at the problem of electric power load data of study,the author combined the traditional k-means clustering algorithm(K-means)and power load characteristics and the composition and proposed double clustering algorithm based on weight automatically updated(AHPK-means),the method could update clustering center according to the weights ratio of load characteristic and composition,and merged all the clustering center again to determine the ultimate load mode.The accuracy of power load data clustering was improved.Thirdly,in view of the negative impact of large-scale electric vehicles disorderly charging on the power grid,a dynamic game charging method for electric vehicles based on time-sharing price is proposed.In this method,the dynamic dispatching device is used as the control center,the time-sharing price is used as the guide,and the user responds independently.Monte Carlo simulation method is used to calculate quantitatively that the method can achieve "peak shaving and valley filling" and reduce the impact of large-scale electric vehicle disorderly charging on the power grid.Fourthly,a dynamic game theory charging method based on typical loads was proposed to improve the accuracy of setting peak and valley time periods and peak and valley prices based on the influence of users on the responsiveness of the time-of-use electricity charging method.This method determined the applied typical load curve through Pearson coefficient,and divided the typical load curve into the periods of peak,flat and valley load by using membership function,so as to find out more time of valley charging,and determined the peak and valley price to complete the classification of charging price based on the user response model finally.