A multistage simulation-optimization-integrated methodology framework for user-oriented electric vehicle carsharing reallocation under dynamic price subsidy

Abstract

Electric vehicle (EV) carsharing operation is facing major challenge to balance the increasing user demand with insufficient vehicle availability and station capacity. Previous researchers mainly focus on reallocating EVs in limited stations and discrete periods through operator-oriented strategies, while containing scant study of dynamic pricing scheme that can benefit supply-demand balance and service profitability in continuous operation periods. Therefore, this paper proposes a multistage simulation-optimization-integrated methodology framework for adaptively incentivizing users based on dynamic price subsidy (DPS). In stage I, Simulation of Urban Mobility (SUMO) is used to determine the real-time supply-demand distributions from user historical data. In stage II, a novel user-oriented reallocation optimization model is established to maximize the operation profit based on the obtained SUMO results. To obtian feasibile solutions, an improved immune genetic algorithm is developed to effectively solve the model. The real-world case study in Shanghai shows that, compared with other benchmark methods, the proposed methodology can significantly increase the daily profit from 78981 RMB to 89506 RMB and improve the system balance from 78.1 % to 94.5 %. Sensitivity analysis and cost-benefit analysis show that the user-oriented DPS can reduce operation cost and order waiting time, while promoting EV utilization within different user populations.