An Autonomous Response Method to Cluster Classification of Large Charging Piles

The accurate response of electric vehicles as a flexible load to the regulatory signals of multiple power markets requires a high degree of cooperation from charging piles, which poses a serious challenge to the response performance of the piles. In this paper, a method of classifying charging pile clusters is proposed based on the response accuracy as well as the linear fit to construct response characteristics. The charging piles will be classified into four types, the K-means clustering method is used to provide uniform regulation instructions for linear clusters and the probability distribution of power deviations to control the error range of nonlinear clusters respectively. By giving different degrees of regulation instructions to different clusters, it ensures that the charging pile resources with good responsiveness and cost-effective regulation can be utilized in priority