In view of the shortcomings of the research on electric vehicle routing optimization in combination with charging station and vehicle sharing, this paper proposes several optimization strategies, including the charging station and electric vehicle sharing with multiple service periods and centralized transportation among multiple centers. In addition, it studies the electric vehicle routing optimization problem of multi-center joint distribution based on resource sharing. First, it establishes a bi-objective optimization model to minimize the operating cost, including the rental cost of electric vehicles, the cost of electricity consumption, the cost of service, the penalty cost for violating the time windows, and the number of electric vehicles. Then, it designs a 3D-K-means spatial-temporal clustering algorithm based on the characteristics of the model considering the geographic location and demand time windows of customers. Next, it proposes a hybrid algorithm consisting of the Clarke-Wright (CW) saving algorithm and the multi-objective particle swarm optimization (CW-MOPSO). It uses the CW saving algorithm to generate the initial solutions, and devises the charging station insertion strategy, the external archive update strategy, and the resource sharing strategy in MOPSO to improve the quality of Pareto optimal solutions. Afterwards, it verifies the proposed CW-MOPSO algorithm by comparing it with the non-dominated sorting genetic algorithm, the multi-objective genetic algorithm, and the multi-objective gradient evolution algorithm. Finally, it conducts a case study of the electric vehicle routing optimization problem of multi-center joint distribution based on the resource sharing modes with the actual data of a logistics enterprise in Chongqing, China. It analyzes and discusses the changes of the operating cost, the number of electric vehicles, the electricity consumption, and the number of used charging stations of the multi-center joint distribution network under the conditions of the uncertainty of the waiting time of electric vehicles in the charging station, the stepwise relationship between the electric vehicle power consumption and speed, and the different resource sharing modes. The results show that in the scenario where some charging stations queue up and the rest do not queue up, some electric vehicles will choose a charging station with a longer distance for charging in order to reduce the waiting time in the queue, which increases the driving distance of electric vehicles and has a higher penalty cost for customer delay. However, when the electric vehicles have different speed states in different service time periods, the power consumption of electric vehicles in the process of long-distance constant speed driving is less than that of short-distance driving in different speed states. Moreover, the proposed model and algorithm can achieve the sharing of charging stations, the shared scheduling of electric distribution vehicles, and the reasonable electric vehicle routing optimization. Furthermore, the proposed model and algorithm can improve the operation efficiency of the multi-center joint distribution network and reduce the operating cost, thus providing theoretical support and decision-making reference for urban logistics and distribution enterprises to realize the rational configuration of charging stations and optimal scheduling of the electric distribution vehicles.
