An adaptive large neighborhood search heuristic for the Electric Vehicle Scheduling Problem

• The electric vehicle scheduling with partial charging is addressed.
• A mixed integer programming model based on a directed acyclic graph is presented.
• An adaptive large neighborhood search heuristic is developed and tested on generated data.
• The model can solve instances with up to 30 trips and the heuristic up to 500 trips.

This paper addresses the Electric Vehicle Scheduling Problem (E-VSP), in which a set of timetabled bus trips, each starting from and ending at specific locations and at specific times, should be carried out by a set of electric buses or vehicles based at a number of depots with limited driving ranges. The electric vehicles are allowed to be recharged fully or partially at any of the given recharging stations. The objective is to firstly minimize the number of vehicles needed to cover all the timetabled trips, and secondly to minimize the total traveling distance, which is equivalent to minimizing the total deadheading distance. A mixed integer programming formulation as well as an Adaptive Large Neighborhood Search (ALNS) heuristic for the E-VSP are presented. ALNS is tested on newly generated E-VSP benchmark instances. Result shows that the proposed heuristic can provide good solutions to large E-VSP instances and optimal or near-optimal solutions to small E-VSP instances.