中文核心期刊
CSCD来源期刊
中国科技核心期刊
RCCSE中国核心学术期刊
重庆交通大学学报（自然科学版）

重庆交通大学学报（自然科学版） ›› 2023, Vol. 42 ›› Issue (7): 128-135.DOI: 10.3969/j.issn.1674-0696.2023.07.17

• 交通基础设施工程 • 上一篇    

城市建成环境对共享单车还车量的影响分析

王振报1，李慧庆1，刘卓2   

  1. (1. 河北工程大学 建筑与艺术学院，河北 邯郸 056038； 2. 北京工业大学 城市交通学院，北京 100124)
  • 收稿日期:2022-01-26 修回日期:2022-04-23 发布日期:2023-09-08
  • 作者简介:王振报（1978—），男，黑龙江铁力人，教授，博士，主要从事多元数据与交通规划方面的研究。E-mail：wangzhenbao@hebeu.edu.cn 通信作者：刘卓（1990—），男，北京人，讲师，博士，主要从事交通智能仿真方面的研究。E-mail：liuzhuo66@bjut.edu.cn
  • 基金资助:
    国家自然科学基金青年项目(52008006)；河北省社会科学发展研究课题(20210201407)

Influence Analysis of Urban Built Environment on the Return Quantity of Shared Bikes

WANG Zhenbao1, LI Huiqing1, LIU Zhou2   

  1. (1. School of Architecture and Art, Hebei University of Engineering, Handan 056038, Hebei, China; 2. College of Metropolitan Transportation, Beijing University of Technology, Beijing 100124, China)
  • Received:2022-01-26 Revised:2022-04-23 Published:2023-09-08

PDF

摘要： 在确定共享单车客流最佳空间单元集计方法后，进行城市建成环境对共享单车还车量的影响程度分析。以厦门岛为例，将共享单车还车位置数据进行集计得出的空间单元还车量作为因变量，计算各空间单元的建成环境指标作为自变量，利用最小二乘法、地理加权回归、多尺度地理加权回归模型，针对8种空间单元划分方法进行回归分析，确定拟合优度最佳的空间单元划分方法，并分析建成环境变量影响程度的空间异质性。研究结果表明：①采用交通分区进行还车量集计的多尺度地理加权回归模型拟合优度最佳；②公司企业兴趣点密度、建筑密度、交叉口数量和共享单车电子围栏数量与还车量呈正相关；空间封闭度、科研教育兴趣点密度、至最近地铁站距离和兴趣点多样性与还车量呈负相关；最近公交站距离在西南部呈负相关，在西北、北部和东北部呈正相关。研究思路和方法为提高共享单车还车量预测模型精度提供了参考，建成环境变量影响程度的空间异质性结果为不同区位提出差异化的建成环境更新策略提供了科学依据。

关键词: 交通工程；建成环境；共享单车出行需求量；样本集计方法；空间异质性

Abstract: After determining the optimal spatial unit collection method of dockless bike-sharing passenger flow, the influence degree of urban built environment on the return amount of shared bike was analyzed. Taking Xiamen island as an example, the spatial unit return amount obtained from the collection of shared bike return location data was taken as the dependent variable, and the built environment index that calculated each spatial unit was taken as the independent variable. The regression analysis on eight spatial unit division methods was carried out by using the ordinary least square method (OLS), geographical weighted regression (GWR) and multi-scale geographical weighted regression (MGWR) models. The spatial unit division method with the best goodness of fit was determined and the spatial heterogeneity of the impact degree of built environment variables was analyzed. The results show that: ① the MGWR model using traffic zoning for shared bike return amount aggregation has the best fit; ② the density of corporate interest points, the density of buildings, the number of intersections and the number of electronic fences for shared bike are positively correlated with the number of returned bikes; the spatial closure, the density of interest points in scientific research and education, the distance to the nearest subway station and the diversity of interest points are negatively correlated with the number of returned bikes. The distance to the nearest bus stop is negatively correlated in the southwest, and positively correlated in the northwest, north and northeast. The research ideas and methods provide a reference for the accuracy of the prediction model of the shared bike return amount, and the spatial heterogeneity results of the influence degree of built environment variables provide a scientific basis for proposing differentiated built environment renewal strategies in different locations.

Key words: traffic engineering; built environment; travel demand of bike sharing; samples aggregate methods; spatial heterogeneity

中图分类号: