A Decision Support System for User-Based Vehicle Relocation in Car Sharing Systems
提出一种决策支持系统，通过阈值政策的经济激励解决用户参与的车辆调度问题，利用离散事件仿真和粒子群优化确定最优阈值，案例表明能显著提升系统性能。
Car sharing (CS) services are promising solutions complementary to the classic public transport forms. In order to make CS effectively competitive, suitable planning and management strategies are required. This paper presents a decision support system (DSS) for handling the user-based vehicle relocation problem by applying economic incentives ruled by a threshold policy. Unlike the existing approaches, a methodology is proposed for determining the optimal threshold, which considers explicitly the stochastic reactions of the customers to the incentives. To this aim, the CS system is described in detail by unified modeling language diagrams and is modeled in a discrete event system framework. Moreover, a closed-loop control strategy is introduced to implement the vehicle relocation policy on the basis of the system state and the best threshold values, evaluated by discrete event simulation and particle swarm optimization. A case study simulation analysis shows that the proposed DSS management strategy can significantly improve the system performance.