Research on the Layout Algorithm of Automobile Charging Piles Based on Density Clustering

Abstract

To address major industry challenges in China’s urban electric vehicle charging infrastructure, including unbalanced spatial layout of electric vehicle charging piles, mismatch between supply and demand, strong subjectivity in site selection, and limited distribution resources in older urban districts, this paper proposes a two-layer layout planning algorithm for charging piles based on improved DBSCAN density clustering + dual-objective constrained optimization. First, multi-source data, including urban POI locations, residents’ travel GPS stay-point data, road network data, and distribution network capacity data are integrated to quantify the spatial distribution of charging demand, and a mathematical model for charging-demand density is constructed; Second, the density-based clustering process is improved, and the centroids of the identified clusters are used as candidate locations for EV charging facilities; Finally, with the optimization objective of minimizing total construction cost and minimizing the average charging detour distance for users, a mixed-integer programming mathematical constraint model is established to optimize the number of chargers allocated to each candidate site. The proposed method verifies the scientific validity and practical applicability of the improved density-clustering algorithm in EV charger layout planning.