Journal of Transportation Systems Engineering and Information Technology ›› 2022, Vol. 22 ›› Issue (4): 106-112.DOI: 10.16097/j.cnki.1009-6744.2022.04.012

Previous Articles     Next Articles

Refined Classification of Urban Rail Transit Stations Based on Clustered Station's Passenger Traffic Flow Features

JIANG Yang-sheng a, b, YU Gao-shang a, b, HU Lu* a, b, LI Yan a, b   

  1. a. School of Transportation and Logistics; b. National Engineering Laboratory of Integrated Transportation Big Data Application Technology, Southwest Jiaotong University, Chengdu 611756, China
  • Received:2022-03-20 Revised:2022-04-16 Accepted:2022-04-27 Online:2022-08-25 Published:2022-08-23
  • Supported by:
    National Natural Science Foundation of China(71901183);Applied Basic Research Programs of Science and Technology Department of Sichuan Province(2021YJ0066)。

基于聚类站点客流公共特征的轨道交通车站精细分类

蒋阳升a, b,俞高赏a, b,胡路* a, b,李衍a, b   

  1. 西南交通大学,a. 交通运输与物流学院,b. 综合交通大数据应用技术国家工程实验室,成都 611756
  • 作者简介:蒋阳升(1976- ),男,湖南衡阳人,教授,博士。
  • 基金资助:
    国家自然科学基金;四川省科学技术厅应用基础研究项目

Abstract: The existing classification of urban rail transit stations is mostly based on qualitative analysis, which cannot meet the needs of refined design and operation. This paper proposes a refined station classification method based on clustering station public features. First of all, the entry flow data from AFC (Automatic Fare Collection) is processed into time series data, and each station is clustered by the data based on K-Means++ algorithm; A fitting equation between the passenger flow clustering and the multi-dimensional parameters of land use characteristics is established to calculate the public characteristics of stations in five major categories, such as residential station, working station, and regional center station. On this basis, considering the segmentation characteristics of different stations belonging to the same broad category of stations, this paper proposes a refined description of specific station types using a combination of the public features of five types of passenger flow. The result of case study shows that the mean absolute percentage error between the real passenger flow value and the fitted passenger flow value calculated using the proposed refined classification method for each station falls within 14% , indicating a good feasibility of the proposed classification method.

Key words: urban traffic, time series clustering, land use features, station fine classification, passenger flow features

摘要: 已有城市轨道交通车站分类多基于定性分析,不能满足精细化设计和运营的需要。本文提出一种基于聚类站点公共特征的站点精细分类方法。首先,将来源于AFC(Automatic Fare Collection)的进站客流量数据处理为时间序列数据,并基于K-Means++算法对各个站点的客流量进行聚类;其次,建立客流量聚类结果与土地利用特征多维参数的拟合方程,计算获得居住密集型、工作就业型以及区域中心型等5种大类站点的客流量公共特征。在此基础上,充分考虑属于同一大类站点不同站点的细分特性,使用5类客流量公共特征比重组合精细描述具体站点类型。 实例结果表明,使用本文提出的精细分类方法计算得到的每个站的客流量拟合值与真实客流值间的平均绝对百分比误差控制在14%以内,说明该分类方法具有可行性。

关键词: 城市交通, 时间序列聚类, 土地利用特征, 站点精细分类, 客流特征

CLC Number: