基于多源数据的城市轨道交通乘客路径选择估计方法  

Estimation Method of Urban Rail Transit Passenger Route Choices Based on Multi-source Data

作　　者：朱国[1] 章澜岚 刘家俊[2] 杨皓帆 殷立超 张宁[2] 张恒文 ZHU Guo;ZHANG Lanlan;LIU Jiajun;YANG Haofan;YIN Lichao;ZHANG Ning;ZHANG Hengwen(Nanjing Panda Information Industry Co.,Ltd.,Nanjing 210008;ITS Rail Transit Research Institute of Southeast University,Nanjing 210018;CRRC P&C Institute Co.,Ltd.,Nanjing 211800)

机构地区：[1]南京熊猫信息产业有限公司,南京210008 [2]东南大学智能运输系统研究中心轨道交通研究所,南京210018 [3]中车城市交通规划设计研究院有限公司,南京211800

出　　处：《都市快轨交通》2025年第1期99-105,共7页Urban Rapid Rail Transit

基　　金：中国城市轨道交通协会重大专项(CAMET-KY-2022007);国家重点研发计划资助项目(2020YFB1600701)。

摘　　要：城市轨道交通的成网运行,带来了客流路径选择的多样性。依据现有自动售检票系统(AFC)交易数据和概率推断方法,难以准确获知乘客的网络路径选择,完成轨道交通网络客流分配和票务清分等工作。利用网络站点信息构建城市轨道拓扑网络,搜寻OD对可行路径集合,并利用AFC交易数据、手机信令数据、列车时刻表数据等多源数据构建非线性优化模型,推断乘客出行的路径选择。基于南京地铁网络进行验证,结果表明:模型具有有效性和鲁棒性。本研究可以为城市轨道交通运营及票务清分提供指导。The network operation of urban rail transit has introduced diversity in passenger route choices.It is difficult to accurately determine passengers’network route choices based on existing Automatic Fare Collection(AFC)transaction data and probabilistic inference methods.This difficultyaffects tasks such as rail transit network passenger flow allocation and ticket clearing.This study utilizes network station information to construct an urban rail topology network.The proposed method searches for feasible path sets for Origin-Destination(OD)pairs and uses multi-source data,including AFC transaction data,mobile signaling data,and train schedule data,to build a nonlinear optimization model to infer passengers’travel route choices.Experiments based on the Nanjing Metro network show that the model is effective and robust.This study can provide guidance for urban rail transit operations and ticket clearing.

关 键 词：城市轨道交通网络 自动售检票系统 多源数据 出行路径选择 非线性优化 

分 类 号：U231[交通运输工程—道路与铁道工程]