哈尔滨工程大学学报2025，Vol.46Issue(4)：755-763,9.DOI:10.11990/jheu.202306007
代理模型辅助进化算法求解大规模电动车辆路径问题
Surrogate-assisted evolutionary algorithm for large-scale electric vehicle routing problems
摘要
Abstract
Aiming to solve the large-scale electric vehicle routing problem,a surrogate-assisted evolutionary algo-rithm is proposed.This algorithm is based on a general two-stage optimization framework:routing and charging.By introducing a surrogate model into the charging optimization stage,the algorithm can partially replace the time-con-suming real charging optimization process,thus enhancing search efficiency.Specifically,the improved max-min ant colony system algorithm is used in the routing optimization stage to generate high-quality routes for customer vis-its.In the charging optimization stage,the surrogate model is then introduced to build the corresponding relation-ship between the routes and total distance using historical data,enabling fast predictions of the total distance for the complete solution after adding charging stations.This approach reduces the required time for real charging optimiza-tion in large-scale customer routing.The proposed algorithm is compared with heuristic and evolutionary algorithms on EVRP test sets with different sizes.Experimental results show that the proposed algorithm achieves comparable solution quality while improving computational efficiency by an average of nearly 14 percentage points compared to existing algorithms.This provides an efficient and practical solution to the electric vehicle routing problem,with strong potential for future application.关键词
电动车辆路径问题/充电优化问题/进化算法/大规模优化/代理模型/蚁群算法/两阶段优化/计算效率Key words
electric vehicle routing problem/charging optimization problem/evolutionary algorithm/large-scale op-timization/surrogate model/colony system algorithm/two-stage optimization/computational efficiency分类
计算机与自动化引用本文复制引用
王朝,查帮政,秦芳..代理模型辅助进化算法求解大规模电动车辆路径问题[J].哈尔滨工程大学学报,2025,46(4):755-763,9.基金项目
国家自然科学基金青年项目(62106002). （62106002）
