ABSTRACT
As a popular means of public transportation, shared bikes are spread all over the world and effectively serve the development of sustainable transportation. However, in the process of development, problems such as difficulty in finding bikes and slow dispatching are gradually exposed. Based on the complex network theory, this paper studies the network characteristics and scheduling problems of station-based and dock-less bike-sharing systems by selecting the actual trajectory data of bicycles in Shanghai, China and Minneapolis, USA. First, according to the characteristics of the two types of bike stations, the station-based bike-sharing system adopts the original graph construction method, and the dock-less bike-sharing system proposes a method based on manual grid division, which can reflect the accuracy of the division unit in analyzing the characteristics of the bike-sharing network from the perspective of geographic space. Then, the small-world and scale-free characteristics of the constructed network are analyzed macroscopically. On the micro-level, multiple centrality indexes are used to measure node centrality, and comprehensive indexes are obtained through dimensionality reduction algorithm to quantitatively identify important nodes and regions of shared bike network. Such spatial distribution characteristics can provide reasonable suggestions for operators to place bikes. Finally, the concept of in and out intensity in complex network is introduced to analyze the adaptive scheduling capability of the site. The research conclusion can provide valuable reference for the planning and operation of shared bikes and urban space planning.
