Crowding Perception Thresholds of Passengers in Urban Rail Transit: A Study of Differences in Spatiotemporal Dimensions

This paper focuses on the differences in crowding perception among different types of passengers in trains, aiming to optimize passenger experience and improve the level of service of urban rail transit. Based on data from a passenger survey on the Beijing subway, this paper introduces the concept of Crowding Perception Threshold (CPT) and analyzes the principle of passenger spatiotemporal crowding. Considering factors such as gender, age, travel purpose, and standing density of passengers, the paper constructs a quantitative model of crowding perception using the ordered logit model and proposes a method for classifying the level of service accordingly. The study results indicate that the CPTs for all types of passengers range from 91.8% to 101.6%, with the females, elderly individuals, and noncommuters showing greater sensitivity to crowding. In the temporal dimension, all types of passengers have higher CPTs during peak hours than during off-peak hours, influenced by passengers' crowding expectations. In the spatial dimension, the level of service for most types of passengers is considered crowded at standing densities of 6-7 pax/m2 during peak hours, while the level of service for all types of passengers is deemed to be very crowded at 8 pax/m2, at which point additional passengers are not recommended.