Efficient Extended Ant Colony Optimization for Capacitated Electric Vehicle Routing

Bo-Cheng Lin,Xiao-Fang Liu,Yi Mei

DOI: https://doi.org/10.1109/ssci51031.2022.10022179

2022-01-01

Abstract:With the rapid development of electric vehicle (EV) freight industry, the capacitated electric vehicle routing problem (CEVRP) becomes a challenging and significant prob-lem. CEVRP considers both the cargo capacity of EVs and the battery capacity of EVs, resulting in the difficulty of the construction of feasible solutions. In this paper, we propose an extended ant colony optimization (EACO) algorithm, which uses a pheromone matrix to record the routing information of customers and charging stations for solution construction. Two route construction methods: coarse route construction method and fine route construction method, are developed to generate a customer service sequence with recharging stations. In the coarse route construction method, cargo capacity constraints are ignored and the selection of recharging stations depends on the remaining battery power after the routing of customers only. In the fine routing construction method, the cargo capacities are considered to better evaluate the remaining battery power of vehicles if a return to depots is necessary in the current construction step. After the construction of routing, the giant tour is split and local search is applied to improve solution quality. Experiments show that the proposed EACO is moderate effective. The fine routing construction shows higher exploration ability and performs better than coarse routine construction.