地球信息科学学报

地球信息科学学报 ›› 2020, Vol. 22 ›› Issue (6): 1282-1293.doi: 10.12082/dqxxkx.2020.190623

• 大数据与智慧交通 • 上一篇    下一篇

基于大规模手机位置数据的城市潜在自行车出行需求评估

周亚娟1,2, 赵志远1,2,3, 吴升1,2,3,*(), 方志祥4, 陈佐旗1,2,3   

  1. 1. 福州大学数字中国研究院（福建）,福州 350003
    2. 空间数据挖掘与信息共享教育部重点实验室,福州 350003
    3. 海西政务大数据应用协同创新中心,福州 350002
    4. 武汉大学测绘遥感信息工程国家重点实验室,武汉 430079
  • 收稿日期:2019-10-24 修回日期:2020-12-12 出版日期:2020-06-25 发布日期:2020-08-25
  • 作者简介:周亚娟（1993— ）,女,安徽安庆人,硕士生,从事地理信息服务与时空数据挖掘研究。E-mail: 1184310214@qq.com
  • 基金资助:
    国家重点研发计划项目(2017YFB0503500);中国博士后科学基金(2019M652244)

Estimating the Potential Demand for Bicycle Travel based on Large-scale Mobile Phone Location Data

ZHOU Yajuan1,2, ZHAO Zhiyuan1,2,3, WU Sheng1,2,3,*(), FANG Zhixiang4, CHEN Zuoqi1,2,3   

  1. 1. Academy of Digital China ( Fujian ), Fuzhou University, Fuzhou 350003, China
    2. Key Laboratory of Spatial Data Mining and Information Sharing, Ministry of Education, Fuzhou 350003, China
    3. Fujian Collaborative Innovation Center for Big Data Applications in Governments, Fuzhou 350002, China
    4. State Key Laboratory of Information Engineering for Surveying, Mapping and Remote Sensing, Wuhan University, Wuhan 430079, China;
  • Received:2019-10-24 Revised:2020-12-12 Online:2020-06-25 Published:2020-08-25
  • Contact: WU Sheng
  • Supported by:
    National Key R & D Program of China(2017YFB0503500);Supported by China Postdoctoral Science Foundation(2019M652244)

RichHTML

PDF (PC)

赞

摘要：

潜在自行车出行是指可能会使用自行车作为交通工具的出行,评估潜在自行车出行需求能够帮助指导城市自行车资源配置方案的优化。大规模手机位置数据蕴含丰富的人群移动信息,而且具有大样本、低成本的特点,能够用于评估城市潜在自行车出行需求。本文结合自行车出行的时间和距离特征,提出一种基于大规模手机位置数据的潜在自行车出行需求评估方法。该方法以单次出行为分析单元,从手机用户的轨迹中提取出具有短距离出行特征和公共交通接驳出行特征的移动轨迹段,并根据该移动轨迹段评估潜在自行车出行需求。基于该方法,利用上海市大规模手机位置数据评估上海市潜在自行车出行需求并分析其时空分布特征,发现在空间上,潜在自行车短距离出行需求主要分布在城市中心和郊区的商业中心,而公共交通接驳的自行车需求主要分布在郊区。在时间上,上午,自行车出行需求从非中心城区向中心城区聚拢;晚上,上海市自行车出行骑车与停车需求从中心城区向非中心城区扩散。

关键词: 手机位置数据, SMoT, 潜在自行车出行需求, 短距离出行, 公共交通接驳, 停留识别, 移动轨迹段, 上海

Abstract:

The potential bicycle travel demand indicates the travel demand that could potentially be served by bicycles. Assessing the potential bicycle travel demand can help to optimize the allocation of the related infrastructure (e.g., bike parking areas and bike lanes) in cities. Mobile phone location data have the advantage of providing low-cost and large-scale sample sizes that contain rich human mobility information. The data can be used to estimate the potential bicycle travel demand. Based on the spatiotemporal characteristics of daily bicycle travel, we proposed a method for assessing the potential bicycle travel demand from large-scale mobile phone location data. Specifically, each individual instance of travel was taken as a sample for the analysis. First, we used the Stops and Moves of a Trajectory (SMoT) model to extract the movement trajectory segments of the users. Second, we identified a "tour" pattern for the trajectory segments, where the start location and the end location were the same. Then, the location that was at the furthest point from the start location was used to divide the movement trajectory segment into two segments. Finally, the movement trajectory segments that were characterized by short distances and those in which the "last mile" of the travels was served by the public transport system were extracted for further assessment of the potential bicycle travel demand. In this study, Shanghai was chosen as the example city. Through our proposed method, we assessed and analyzed the spatiotemporal characteristics of daily bicycle travel in Shanghai to determine the potential bicycle travel demand. From a spatial perspective, we found the following: (1) the potential bicycle travel demand in Shanghai was mainly concentrated in the downtown areas and commercial centers in the suburb areas; (2) the potential bicycle travel demand in the downtown areas and commercial central urban areas was stable, while the potential bicycle travel demand in the suburban areas tended to be variable; and (3) most of the “last mile” demands were located in the suburb areas, which showed that the characteristics of the “last mile” demands at different public transport stations varied. From a temporal perspective, several patterns could be observed during the morning and evening rush hours: (1) the potential bicycle travel demand in the central urban area continued to remain relatively high; (2) the potential bicycle travel demand in the suburbs in the Songjiang and Qingpu districts had relatively large differences; (3) the potential bicycle travel demand was concentrated in the direction of the central urban area from the noncentral urban areas in the morning, while the potential bicycle travel demand spread from the central urban areas to the noncentral urban areas in the evening; and (4) the potential bicycle travel demand of Shanghai showed a double-peak characteristic (at 11:00—12:00 and 16:00—17:00). The “last mile” type demand also had two peaks (at 7:00—9:00 and 17:00—18:00).

Key words: mobile phone location data, SMoT, potential bicycle travel demand, short distance travel, public transportation transfer travel, stop identification, move trajectory segment, Shanghai