Dynamic Vehicle Routing for Battery Swapping in an Electric Bike-sharing System

In the daily operation of the emerging electric bike-sharing system, timely and efficient battery swapping is essential as there are no charging facilities at e-bike stations. This paper deals with the path planning of the trucks for battery swapping, considering dynamic battery demands at different stations. This problem is modeled as a vehicle routing problem with intermediate stops and soft time windows. A new solution method combining a dynamic planning strategy and a multiple neighborhood search algorithm is proposed to solve the problem. Numerical results on benchmark instances and a real-world case prove the effectiveness of our model and method.

I. Introduction

Electric bike-sharing (EBS) is a new mode for short-distance urban travel, which operates similarly to a station-based bike-sharing system but with battery-powered motors. A user can only pick up and return an e-bike at fixed stations. There are no charging facilities at e-bike stations, and the operator needs to swap the batteries of e-bikes at different stations [1], [2]. The current strategy for battery swapping utilized by commercial companies is to divide a large EBS service area into smaller zones, each of which includes multiple stations and a battery warehouse. Inside each zone, operators drive trucks to deliver fully charged batteries from the warehouse to the stations and replace the low-power batteries on the e-bikes at the station there. This paper studies the path planning of these trucks for battery swapping.