摘要
地铁自动售检票系统可以采集大量乘客刷卡数据,可提供更全面的地铁乘客时空信息。对乘客的出行模式分析有利于城市轨道交通运营企业预测地铁客流和制定运营策略。提出了分析地铁乘客出行模式的数据挖掘方法:对地铁刷卡数据进行预处理,根据其时空信息生成乘客出行链;分析反映乘客时空特性的聚类变量;利用K-means聚类算法对各聚类变量进行乘客聚类;分析潜在的乘客出行模式。以深圳地铁刷卡数据为例,对提出的地铁乘客出行模式分析方法进行了试验验证。
The subway Automatic Fare Collection systems collect tremendous amount of smart card data,which provides comprehensive spatial-temporal information about subway passengers.The analysis of passengers travel patterns benefits urban rail transit operation companies in predicting subway passenger flow and formulating operational strategies.A data-mining procedure to identify travel patterns of subway passengers was introduced:after pretreatment of smart card data,passengers travel chains were generated based on the spatial-temporal information of it;clustering variables that reflect spatial-temporal characteristics of passengers were analyzed;K-means clustering algorithm was adopted to cluster the passengers;the potential passengers travel patterns were then analyzed.Taking the Shenzhen subway smart card data as example,verification experiment was conducted on the proposed subway passengers traveling patterns analysis methodology.
作者
项煜
陈晓旭
杨超
段红勇
XIANG Yu;CHEN Xiaoxu;YANG Chao;DUAN Hongyong(Henan Transport Card Co.,Ltd.,450018,Zhengzhou,China;不详)
出处
《城市轨道交通研究》
北大核心
2020年第6期63-67,共5页
Urban Mass Transit
基金
河南省交通运输科技计划项目(2019G-2-2)
中央高校基本科研业务费专项资金项目(22120180241)。
