A Simulation-optimization framework of vehicle relocation for one-way electric carsharing systems

Abstract

In one-way electric carsharing systems (ECSs), customers can pick up an electric vehicle (EV) at one location and return it to any location, causing an imbalance of EVs and revenue loss for ECSs. In order to address this problem, a simulation-optimization framework is introduced to determine optimal vehicle relocation strategies. In this framework, the discrete event simulation model considers stochasticity and EV-dependent road congestion to simulate the operation of ECSs. And the optimization module is triggered by the simulation model to provide relocation strategies to the operator based on the genetic algorithm with the optimal computing budget allocation. The results of numerical experiments demonstrate that the simulation model can reflect the operation of ECSs in the real world, and the joint relocation strategies achieve improvements in both the profit and the vehicle utilization rate and reduce the number of relocations concerning the single relocation strategies used in the ECS.

KEYWORDS:

• Carsharing
• electric vehicle
• discrete event simulation
• genetic algorithm
• optimal computing budget allocation

Acknowledgments

The authors would like to thank Dr. Haobin Li of the National University of Singapore and Ph.D. candidate Shaowen Zhu of the Southeast University for their valuable comments on an earlier draft.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [grant number 71901183].