Real-time vehicle relocation and staff rebalancing problem for electric and shared vehicle systems
研究了电动共享汽车服务中实时车辆调度和员工再平衡问题，提出基于马尔可夫决策过程的调度策略，显著提升服务水平和盈利能力。
This study addresses the challenging real-time vehicle relocation and staff rebalancing (RT-VR&SR) problem in electric carsharing services. The complexity arises from ad-hoc demand, charging requirements of electric vehicles (EVs), and staff scheduling constraints. The problem aims to maximize the profit of carsharing operators by determining strategies for vehicle relocation, vehicle charging, and staff rebalancing in real-time. It considers the uncertainty of demand and the practical nonlinear charging profile of EVs. We formulate this problem as a Markov Decision Process (MDP), and propose an efficient concurrent-scheduler-based policy. Numerical experiments are conducted to demonstrate the effectiveness of the proposed policy and methodology. The results show that the proposed policy significantly improves service level and profitability compared to a benchmark policy. It is also found that ignoring staff rebalancing in decision making can lead to overestimation of service level and profitability. In conclusion, this study presents a real-time solution for vehicle relocation and staff rebalancing in one-way electric carsharing services. The proposed policy and methodology improve performance and highlight the importance of considering staff rebalancing in decision making.