Dynamic data‐driven computation method for the number of waiting passengers and waiting time in the urban rail transit network

Excess passengers gathering on the urban rail transit platform in a short time brings huge security risks to passengers and the daily operation of urban rail transit. However, the real-time monitoring method of passenger flows, which relies on manual methods, cannot satisfy the requirement of daily operations at the network level anymore. This study proposes a dynamic data-driven computation and monitoring method for the number of waiting passengers on platforms to recognize the operational risk in real-time. For waiting passengers on platforms, the waiting time duration before boarding is also calculated based on a first-come-first-service basis. It can be used to provide real-time information services to passengers and evaluate service quality. Moreover, the proposed methodology relies solely on the AFC data and the train timetable, which makes it easy to implement in the daily operation of any urban rail transit system. Finally, taking the Beijing rail transit network as a case study, the real-time number of waiting passengers on each platform, and the time duration passengers should wait before boarding are calculated dynamically. Meanwhile, the spatio-temporal distribution of passengers’ waiting time and waiting passengers are detailly analyzed based on the method of cluster analysis and the complex network theory.

短期内大量乘客在城市轨道交通站台聚集，给乘客以及城市轨道交通的日常运营带来了巨大的安全风险。然而，目前依赖人工方式的客流实时监测方法，已无法满足网络层面日常运营的需求。本研究提出一种基于动态数据驱动的站台候车乘客数量计算与监测方法，以实时识别运营风险。对于站台候车乘客，还基于先到先服务原则计算其上车前的候车时长。该方法可用于为乘客提供实时信息服务并评估服务质量。此外，所提方法仅依赖自动售检票（AFC）数据和列车时刻表，便于在任何城市轨道交通系统的日常运营中实施。最后，以北京轨道交通网络为例进行研究，动态计算各站台的实时候车乘客数量以及乘客上车前需等待的时长。同时，基于聚类分析方法和复杂网络理论，详细分析乘客候车时间和候车乘客的时空分布情况。