Regulating multi-directional passenger flow: Impact of obstacle position and flow level on pedestrian merging process

Abstract

Mass passenger flow and crowd gathering in subway stations raise the potential of crowd accidents. To improve the safety level in subway stations, it is crucial to investigate the crowd movement characteristics within specific structures with higher risk levels and regulate the passenger flow. The merging structure is common in subway stations with a larger scale of complex multi-directional movements and always be considered a potential bottleneck. It has been proven that specific geometric constraints in merging structures can significantly reduce traffic efficiency and expedite the formation of congestion by aggravating the conflicts among the crowds. However, there is still a lack of research on measures to improve the performance of existing merging structures. Previous findings show that properly set obstacles may increase the performance of bottleneck regions, which provides an opportunity to consider whether the obstacle effect can also be observed during the merging process. From this perspective, we carried out 39 groups of controlled experiments to investigate the impact of obstacle positions on the merging process under three different flow levels. A real human was used as the obstacle, and the distribution of different obstacle positions was divided into three regions. Our results demonstrate that setting the obstacle in certain regions will increase the performance of merging structures in terms of average outflow, velocity, and local density. The average velocity and outflow are increased by up to 6.82 % and 11.26 %, while the local density within the merging area is kept below 3.5 m⁻². In contrast, some other regions will generate less efficient outcomes and cause a higher probability of jamming on the merging process. The maximum reduction of average velocity and outflow is 30.99 % and 19.45 %, and the local density can exceed 5 m⁻². The performance of different obstacle positions is also affected by the flow level, and controlling the flow at a lower level is more beneficial for the merging process. Other microscopic parameters, such as channel utilization, detour degree, and personal distance, are also introduced to compare the impact of obstacle setups on pedestrian behaviors. The results provide a feasible way to improve the efficiency and safety level of merging structures in subway stations.