Identifying Left Behind Passengers at Subway Stations from Auto Fare Collection Data

Abstract

With the rapid growth in transport demand, it has become a frequent occurrence that passengers are left behind especially during peak hours in subway, which has led to a significant reduction in the level of service. In this paper, we propose a left behind passengers identifying method based on Automatic Fare Collection (AFC) and Automated Vehicle Location (AVL) data. Firstly, we choose the passengers with the limited deterministic information as the research objects; secondly, we propose a classification-based method for identifying left behind passengers by the probabilistic model; next, the accuracy and effectiveness of the proposed method is verified by the simulation experiment and the case of Beijing Subway. Ultimately, the proposed method will support research related to the operation, management and future development of subways.

Keywords

• Automatic Fare Collection (AFC)
• left behind
• probabilistic model
• subway
• temporal distribution