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) ›› 2020, Vol. 39 ›› Issue (08): 6-13.DOI: 10.3969/j.issn.1674-0696.2020.08.02

• 交通+大数据人工智能 • 上一篇    下一篇

公交通勤个体出行特征图谱构建及相似性判别

梁泉1,翁剑成1,周伟2,荣建1   

  1. (1. 交通运输部管理干部学院 道路教研部,北京 101601; 2. 北京工业大学 城市交通学院,北京100124; 3. 中华人民共和国交通运输部,北京100736)
  • 收稿日期:2018-07-22 修回日期:2019-01-03 出版日期:2020-08-18 发布日期:2020-08-25
  • 作者简介:梁泉(1989—),女,山东潍坊人,博士研究生,主要从事智能交通、交通运行监测方面的研究。E-mail: lquan0730@163.com 通信作者:翁剑成(1981—),男,浙江金华人,副教授,博士,主要从事智能交通、交通运行监测方面的研究。E-mail: youthweng@bjut.edu.cn
  • 基金资助:
    国家自然科学基金项目(51578028);北京市“科技新星”计划项目(Z171100001117100)

Specific Chromatogram Construction and Similarity Identification of Individual Travel of Public Transport Commuters

LIANG Quan1, WENG Jiancheng2, ZHOU Wei3, RONG Jian2   

  1. (1. Road Teaching and Research Department, Transport Management Institute Ministry of Transport of the Peoples Republic of China; Beijing 101601, China; 2. College of Metropolitan Transportation, Beijing University of Technology, Beijing 100124, China; 3. Ministry of Transport of the Peoples Republic of China, Beijing 100736, China)
  • Received:2018-07-22 Revised:2019-01-03 Online:2020-08-18 Published:2020-08-25

摘要: 从个体角度获取公共交通通勤乘客出行行为特征,有助于更准确地把握公交总体出行规律,更好地满足乘客的出行需求。基于公共交通刷卡数据与线站数据,针对公交通勤出行个体,借助图谱可视化表达优势,研究形成了以横轴为时间序列轴、纵轴为位置轴、节点大小为方向角度的公共交通通勤乘客出行特征图谱构建方法。在此基础上,选用出行方式、方向、时间和线路等属性指标,采用结构相似度与最长公共子序列相结合的分析方法,实现了图谱相似性判定。以个体乘客一周的公共交通出行数据为例,验证了研究结果的合理性和有效性。研究结果能够直观反映个体出行过程,为出行特征表达与提取提供了新思路。

关键词: 交通工程, 通勤乘客, 特征图谱, 相似性判别, 个体特征

Abstract: It is helpful to grasp the overall travel rules of public transport more accurately and better meet the travel needs of passengers to obtain the travel behavior characteristics of public transport commuters from the perspective of individuals. Based on the public transport card swiping data and line station data, by use of the representation advantage of graph visualization of public transport commuters, the travel graph construction method of public transport commuting trip was studied, which took the horizontal axis as the time series axis, the vertical axis as the position axis, and the node size as the direction angle. On this basis, the method of structure similarity combined with the longest common subsequence was used to realize the similarity identification of graph, which selected attribute indicators such as travel mode, direction, time and route. Taking the one-week public transport travel data of individual passengers as an example, the rationality and validity of the research results were verified. The research results can intuitively reflect the individual travel process, which provides a new idea for the expression and extraction of trip characteristics.

Key words: traffic engineering, commute passenger, travel graph, similarity identification, individual characteristics

中图分类号: