Exploring travel patterns and static rebalancing strategies for dockless bike-sharing systems from multi-source data: a framework and case study

Abstract

This paper proposes a research framework for investigating the travel patterns of dockless bike-sharing and accomplishing the large-scale bike rebalancing at the city level. A case study involving Shanghai combines GPS-based bike-sharing usage data and road network data. First, the spatiotemporal mobility patterns are analyzed visually; then community detection is used to divide the study area into management sub-areas according to the mobility characteristics of bike-sharing users; in addition, a clustering algorithm is used to identify virtual stations. On this basis, a heuristic algorithm is used to generate a rebalancing scheme that enables multiple visits to a given station. The results show that Shanghai can be divided into 28 bike-sharing management sub-areas. Static rebalancing based on the identified management sub-areas reduces the number and driving distance of rebalancing vehicles in use, which is a better outcome than that with a method based on administrative divisions.