Influence of waiting behavior on passenger efficiency during boarding and alighting process on the metro platform

The attributes of high efficiency and punctuality attract a substantial number of passengers to metro systems, frequently resulting in overcrowding on the confined station platforms during peak hours. To minimize pedestrian crowding and improve passenger efficiency, passengers’ movement on the platform is detailed in the present study, including the boarding and alighting process and their waiting behavior (hang about on the platform). First of all, a field survey was conducted in Chengdu Metro, where the characteristics of boarding and alighting process were recorded. The survey data showed that lining up on one side of train doors reduces the average boarding and alighting time by 0.066 s, compared with unrestricted queuing, while the “alight first” rule shortens the average alighting time by 0.83 s relative to the “board first” rule. Then, considering the anisotropic nature of the pedestrian motion, a modified Social Force Model (SFM) was proposed to simulate passengers’ boarding and alighting process. Simulation results under a single-train-door scenario quantified the impact of in-carriage crowd density on boarding and alighting efficiency: a critical crowding density threshold of 2.97 persons/m² (corresponding to the level D in Level of Service (LOS)) was identified. When carriage density exceeds this threshold (LOS level D), the average boarding and alighting efficiency exhibits an abrupt deterioration: compared with LOS A-C (density ≤ 2.97 persons/m²), the total alighting/boarding time increases by 126.3 %, the average alighting time by 130.7 %, and the average total boarding time increases by 111.8 %, while the average boarding time by 118.0 %. After leaving the carriage, the influence of waiting behavior on the platform was analyzed: when the crowd density on the platform reaches 1.36 persons/m², the maximum dwelling time of alighting passengers increases by 65 % (107.26 s) compared with scenario without waiting pedestrians. These quantified findings provide reliable references for optimizing metro station passenger flow organization, particularly for single-door scenario management and crowding threshold control. © 2026 Elsevier B.V.