Abstract

This paper presents a cost-effective strategy to reduce train delays by controlling passenger flow at the station entry. When a scheduled dwell time delay is likely to occur, the strategy reduces the number of passengers entering the platform by slowing down the opening speed of the automated fare collection (AFC) gates. The dwell time of the next train is predicted to enable proactive passenger flow control. A linear dwell time model is developed for the control strategy using empirical data from the Gangdong-Gu Office station of Seoul Metro, South Korea. The strategy is tested using simulation in two different settings against the base case with no passenger flow control. The results show the passenger flow control reduced the frequency of scheduled dwell time delays, but it also caused severe congestion in the AFC gate area. While the strategy appears to transfer the location of the overcrowding problem, it is safer and easier to manage congestion in the gate area than on the platform. Coordinating the passenger flow control strategy across a network could reduce scheduled dwell time delays and passenger travel times and improve rail capacity and reliability more cost-effectively than infrastructure solutions.

摘要

本文提出了一种经济有效的策略，通过控制车站入口处的乘客流量来减少火车延误。当预定的停留时间延迟很可能发生时，该策略通过减慢自动收费票（AFC）闸的开启速度来减少进入平台的乘客数量。预计下一班列车的停留时间将有助于主动进行客流控制。使用来自韩国首尔地铁的江东区办公室站的经验数据，为控制策略开发了线性停留时间模型。该策略在没有乘客流量控制的情况下，针对基本情况在两种不同的设置下使用仿真进行了测试。结果表明，客流控制减少了预定的停留时间延迟的频率，但也导致了AFC登机口区域的严重拥堵。尽管该策略似乎可以转移人满为患的问题，但与平台上的情况相比，在登机口区域管理交通拥堵更为安全和容易。与基础设施解决方案相比，在整个网络上协调旅客流量控制策略可以减少计划的停留时间延迟和旅客出行时间，并以更具成本效益的方式提高铁路通行能力和可靠性。