突发事件导致地铁单线运营中断时乘客路径选择行为研究

郭靖凡田媛媛李俊铖何建涛罗启祥

Metro Passenger Path Choice Behavior during Single-line Operation Interruption in Risk Emergencies

GUO JingfanTIAN YuanyuanLI JunchengHE JiantaoLUO Qixiang
摘要:
[目的]为了深入了解突发事件对城市轨道交通乘客出行行为的影响,进而帮助运营部门更有效地应对突发事件带来的运营中断。[方法]问卷设计涵盖了基础性问题(RP法调查)和场景性问题(SP法调查),涉及列车火灾、车站水淹、隧道异物侵入轨道等突发事件场景,以及同线短距离、同线长距离、两线间短距离、两线间长距离等多种出行距离场景。问卷通过广州地铁线网内各类线路车站现场线下发放以及官方公众号的线上渠道进行收集,共获得有效问卷3000份。基于这些问卷数据,本文建立了多项Logit(逻辑回归)模型,量化分析了出行时间、出行费用、换乘次数等乘客出行选择因素的影响程度。[结果及结论]乘客的出行选择受到突发事件类型、出行距离和出行时段等多重因素的影响。特别是在同线短距离出行场景中,共享单车作为替代出行方式的选择比例平均超过4.5%。所建立的多项Logit模型能够评估不同疏导措施对乘客选择概率的影响,为运营部门提供了有价值的决策支持。未来的研究将进一步预测OD(起讫点)客流量,进行客流重新分配,并评估疏导措施的有效性,以优化应对策略。
Abstracts:
[Objective] To gain insights into the impact of emergencies and incidents on urban rail transit passenger travel behavior, and assist operational departments in responding more effectively to operational disruptions caused by such occasions, a questionnaire is specially designed. [Method] The questionnaire design covers both fundamental questions (revealed preference—RP survey) and scenario-based questions (stated preference—SP survey), addressing emergency scenarios such as train fires, station flooding, and foreign object intrusion into tunnels, as well as various travel distance scenarios including short and long distances on the same line and between two lines. The questionnaires are collected through offline distribution at various stations within the Guangzhou Metro network and online via the official public account, gathering a total of 3,000 valid responses. Based on these questionnaire data, multinomial‌ Logit (logistic regression), abbreviated as MNL, models are established to quantitatively analyze the influence degrees of factors such as travel time, travel cost, and travel frequency on passenger travel choices. [Result & Conclusion] Passenger travel choices are influenced by multiple factors, including the type of emergency, travel distance, and travel period. Notably, in short-distance travel scenarios on the same line, shared bicycles account for an average of over 4.5% of the alternative travel choices. The established MNL models can evaluate the impact of different diversion measures on passenger choice probabilities, providing valuable decision support for operational departments. Future research will focus on further predicting OD (origin-destination) demand, redistributing passenger flow, and assessing the effectiveness of diversion measures to optimize response strategies.
论文检索