An integrated framework for electric vehicle rebalancing and staff relocation in one-way carsharing systems: Model formulation and Lagrangian relaxation-based solution approach. (November 2018)

• Record Type: Journal Article
• Authors: Zhao, Meng; Li, Xiaopeng; Yin, Jiateng; Cui, Jianxun; Yang, Lixing; An, Shi
• Is Part Of: Transportation research. Volume 117(2018)Part A
• Page Start: 542
• Page End: 572
• Publication Date: 2018-11
• Subjects: One-way carsharing system -- Electric vehicle rebalancing -- Staff relocation -- Space-time network -- Lagrangian relaxation; Transportation -- Research -- Periodicals; Transportation -- Mathematical models -- Periodicals
• DOI: 10.1016/j.trb.2018.09.014
• Languages: English
• ISSNs: 0191-2615

Abstract

Highlights: An MILP model is formulated to strategically design the allocation plans of EVs and staff members to parking stations This model considers the EV rebalancing and staff relocation operations with spatial-time-dependent user reservations. A Lagrangian relaxation framework is developed to obtain the lower bound of the model. A three-phase implementing algorithm based on DP is proposed for generating good-quality solutions. Abstract: In one-way electric vehicle (EV) carsharing systems, a practical issue that needs to be addressed is the imbalance of EVs with respect to the spatial time-dependent user reservations at different carsharing stations. In practice, appropriate EV rebalancing operations can satisfy user reservations with limited resources and effectively save system investments. This paper proposes an integrated framework that can determine the optimal allocation plan of EVs and staff on the strategic level while considering the operational EV relocation and staff relocation decisions, in order to minimize the total cost, including the EV and staff investment, EV rebalancing and staff relocation costs. In this framework, the dispatching routes of EVs and staff are represented by two sets of space-time paths in the planning time horizon by using a space-time network representation, and the considered problem is then formulated into a mixed-integer linear programming model (MILP). This model explicitly considers (1) the satisfaction of time-dependent user reservations through dynamically rebalancing EVs and relocating staff to keep the service quality of carsharing system, and (2) the EV battery capacity with limited traveling distance and the charging process of EVs at parking stations. A Lagrangian relaxation-based solution approach is developed to decompose the primal problem into several sets of computationally efficient subproblems. In order to generate good-quality solutions, we also propose a three-phase implementing algorithm based on dynamic programming according to the values of Lagrangian multipliers. An illustrative numerical example and a real-world case study (based on the operation data of Seattle, WA) are conducted to verify the applicability of the formulated model and effectiveness of the proposed approach.