基于行为分析的突发事件下城轨站间客流分布预测  

Forecasting Passenger Flow Distribution between Urban Rail Transit Stations Based on Behavior Analysis under Emergent Events

作　　者：刘莎莎[1] 姚恩建[1] 李斌斌[1] 唐英 LIU Shasha;YAO Enjian;LI Binbin;TANG Ying(MOE Key Laboratory for Urban Transportation Complex Systems Theory and Technology,Beijing Jiaotong University,Beijing 100044,China;Guangzhou Urban Planning&Design Survey Research Institute,Guangzhou 510060,China)

机构地区：[1]北京交通大学城市交通复杂系统理论与技术教育部重点实验室,北京100044 [2]广州市城市规划勘测设计研究院,广东广州510060

出　　处：《铁道学报》2018年第9期22-29,共8页Journal of the China Railway Society

基　　金：中央高校基本科研业务费(2017YJS116);北京市自然科学基金(8171003)

摘　　要：提出突发事件下城市轨道交通(城轨)站间客流分布预测方法,旨在为制定突发事件下的应急预案和运营管理措施提供决策支持。首先,提出突发事件下城轨受影响客流界定算法,界定受影响的客流;然后,针对受影响客流,基于非集计理论构建突发事件下城轨乘客的出行选择行为模型,捕捉受影响乘客的出行选择偏好;其次,建立突发事件下的多方式备选出行方案集合,结合乘客出行选择行为建模结果,预测突发事件下城轨站间客流的重分布。最后,基于某突发事件下某城市轨道交通系统的历史客流数据,对该预测方法进行验证。结果表明,该方法能够捕捉到突发事件下城轨乘客的出行行为特征,分时站间客流预测的平均绝对误差为2.05人,预测效果较好。An approach was proposed to forecast passenger flow distribution between urban rail transit(URT)stations based on behavior analysis under emergent events,designed to provide decision support to contingency plan making and operational management measure establishment.First,an algorithm was proposed to define the URT passenger flow affected by an emergent event in an URT system.Second,using disaggregate theory,a travel choice model was developed to capture travel preference of passengers who are influenced by emergent events.Then,a multimode alternative travel plan set under emergent events was established and passenger spatiotemporal distribution was forecasted.Finally,based on the historical passenger flow data in an URT system under an emergent event,the proposed approach was validated.The results show that this approach can reflect the travel preference of URT passengers under emergent events,with the mean absolute error of forecasted origin-destination(OD)passenger flow of 2.05 persons,indicating good forecasting performance.

关 键 词：城市轨道交通 突发事件 站间客流分布 行为分析 

分 类 号：U231.92[交通运输工程—道路与铁道工程]