基于地铁的乘客出行群组模式发现与可视化  

Discovery and visualization of passenger travel group patterns based on subway systems

作　　者：何伟 张彤[1] 黄靖 HE Wei;ZHANG Tong;HUANG Jing(State Key Laboratory of Information Engineering in Surveying,Mapping and Remote Sensing,Wuhan University,Wuhan 430079,China)

机构地区：[1]武汉大学测绘遥感信息工程国家重点实验室,武汉430079

出　　处：《测绘科学》2021年第8期156-164,共9页Science of Surveying and Mapping

摘　　要：拥挤现象是困扰公共交通出行者和管理者的一个重要问题。公共交通出行者需要了解公共交通载具的负载情况从而尽可能地避免拥挤,公共交通管理者需要了解公共交通的客运情况来指挥调度。为了服务于以上需求,该文参考自由轨迹的群组定义方式定义了一种在公共交通系统中与拥挤现象强相关的群组模式,并重点给出了包含原始群组提取和融合原始群组在内的群组模式挖掘算法。基于提取到的群组进一步设计了群组的可视化方法。最后基于深圳地铁智能卡数据对地铁群组进行了模式挖掘与可视化。基于提取到的群组,乘客可以错峰出行,提高出行舒适度;交通规划者可以对拥挤易发生的路段进行调度。Crowding is a major problem for public transport travelers and managers. Public transport travelers need to know the load of public transport vehicles to avoid crowding as much as possible, and public transport managers need to understand the passenger situation of public transport to dispatch the traffic. In order to serve the above needs, this paper defines a group pattern that is strongly related to congestion in the public transportation system with reference to the group definition method of free trajectories, focusing on the group pattern mining algorithm including the original group extraction and fusion of the original group. Based on the extracted group, a group visualization method is further designed. Finally, based on the data of Shenzhen Metro Smart Card, the pattern mining and visualization of the metro group was carried out. Based on the extracted groups, passengers can avoid peak travel thus improving travel comfort;traffic planners can manage crowded road sections.

关 键 词：智能卡数据 出行链 群组模式 可视化 

分 类 号：P208[天文地球—地图制图学与地理信息工程]