An improved clustering algorithm for EV battery recycling network design

This paper addresses the facility location optimization problem in electric vehicle (EV) battery recycling networks, which faces challenges of heterogeneous spatial distribution and dynamic demand patterns. We propose an improved density-based clustering algorithm that integrates a novel density-demand evaluation system with an adaptive service radius mechanism. The proposed algorithm not only effectively captures both spatial clustering patterns and temporal demand characteristics but also dynamically adjusts facility coverage based on regional demand intensity. To optimize the network design, a multi-objective optimization model is developed to balance economic benefits, environmental impact, and service quality. Computational experiments demonstrate that our approach achieves superior performance in both load balancing (0.162) and service coverage (54.4%) compared to traditional methods such as K-means (0.452). Furthermore, the proposed framework provides a practical solution for sustainable recycling network design, thereby contributing to the efficient management of retired EV batteries.