Location and Capacity of Charging Station for Electric Vehicles Based on Weighted K-Means Clustering

Aiming at the uneven distribution of electric vehicles (EVs) charging load caused by the randomness of user travel, this paper proposes a location and capacity of charging station solution based on the weighted K-Means clustering algorithm by defining the "weight" of the charging load. First, analyze the EVs charging load distribution and calculate the weight according to the travel law and state of charge of EVs to form a set of EVs charging load data points to be clustered; Second establish the location and capacity model of charging station with the target of minimizing total annual cost; Then consider the influence of the service range, capacity and number of charging stations to perform weighted K-Means clustering on the charging load data points in the planning area, and select the best location and capacity of charging station for EVs based on different clustering results by calculating the contour coefficient; Finally, it is verified by a simulation example that the weighted K-Means clustering algorithm is feasible for optimizing the location and capacity of charging station for EVs.