Speaker
Description
In recent years, dockless bike-sharing (DBS) have become an important ways of transportation in many cities and have shown outstanding performance in terms of environmental friendliness and short distance travel. Compared to docked bike-sharing that require specific docks to be used, DBS can be rented and returned almost anywhere, making them more flexible and convenient. They flexibility also poses challenges for operation. We can often see piles of DBS blocking the way. Meanwhile, we couldn’t find any bikes sometimes.
Many researchers have explored the spatiotemporal travel patterns of DBS, but existing studies are mostly descriptive and qualitative. Some researchers discuss how built environment affects the travel patterns of DBS. However, these studies only focus on the total usage and lack discussion on arrival and departure behavior respectively which is closely associated with the operation of DBS. Complex networks have ability to characterize the interaction relationships between different nodes including the incoming and outgoing direction, which matches the arrival and departure behavior of bikes. However, most studies using network method focus on docked bike sharing because they have clear nodes for network.
To fill these gaps, we cluster the DBS distribution by K-means and construct a complex network using cluster centroid as nodes. Then we measure the characteristics of DBS complex networks and use OLS regression model to evaluate the impact of built environment on the characteristics of DBS complex network especially in- and out-degrees. The DBS data used in the article comes from the Shenzhen Municipal Government. We select the DBS data in the first week of Feb. 2021. Firstly, we use the K-means clustering method to cluster all the DBS origin-destination coordinates into 1089 points. We constructed a directed DBS network according to the OD flow, and calculated network indicators such as weighted in- and out-degrees, eigenvector centrality, and resilience centrality of each node in the network. Meanwhile, we calculate the built environment metrics within 500m of each node. Finally, we estimate a set of OLS models to explore how the built environment affects DBS network indicators.
The results show that: 1) There is a strong correlation between the weighted out-degree and in-degree indicators in the DBS network and the built environment factors such as job density, residential density, land use mix, and distance to metro. It is worth noting that the length of bicycle lanes and the proximity of bus stops only have a positive promoting effect on the weighted in-degree. In addition, the ratio of weighted indegree to outdegree shows a strong positive correlation with the job density, indicating that people tend to take a DBS to work instead of from work. High job density can lead to DBS accumulation. 2) Road density and length of dedicated bicycle lane are positively related to eigenvector centrality. This means small block and dense road network design and dedicated bicycle lane can promote the clustering of high impact bicycle parking locations. 3) High road density and proximity to shopping centers are associated with high node resilience.
Using network analysis method, our findings provide empirical evidence demonstrating the significant influence of built environment on arrival and departure patterns and other node characteristics of DBS network, which can assist policymakers in planning for DBS parking and help DBS operator to dispatch their DBS more rationally.
References
Diao, Mi; Song, Ke; Shi, Shuai; Zhu, Yi; Liu, Bing (2023): The environmental benefits of dockless bike sharing systems for commuting trips. In Transportation Research Part D: Transport and Environment 124, p. 103959.
Guo, Y., & He, S. Y. (2020). Built environment effects on the integration of dockless bike-sharing and the metro. Transportation Research Part D: Transport and Environment, 83, 102335.
Wu, C., Chung, H., Liu, Z., & Kim, I. (2021). Examining the effects of the built environment on topological properties of the bike-sharing network in Suzhou, China. International Journal of Sustainable Transportation, 15(5), 338-350.
Song, J., Zhang, L., Qin, Z., & Ramli, M. A. (2021). A spatiotemporal dynamic analyses approach for dockless bike-share system. Computers, Environment and Urban Systems, 85, 101566.
| Keywords | Dockless bike-sharing; Built environment; Complex network; K-means |
|---|---|
| Best Congress Paper Award | Yes |
