Pedestrian Swarm Control Based on Temporal/Spatial Frequency of Crossing Flows

Abstract

In the densely-populated urban cities, pedestrian flows often cross each other and congestion occurs. Due to the congestion, we feel discomfort and accidents may occur. In order to reduce the congestion or the risk of accidents, it is required to control swarm behavior of pedestrian flows so that the flows become smooth. This paper proposes the control method of the crossing pedestrian flows. First, we propose the continuum model of the crossing flows. In the actual pedestrian flows, it is known that people formulate the diagonal stripe pattern in the crossing area. The continuum model enables us to quantify such dynamical change of the congestion degree. Then, we propose an implicit control method of the crossing flows. Utilizing the dynamical characteristics of the flows, swarm behavior is controlled by moving a few guides without explicit guidance. From analysis on the crossing flows, we derive a control algorithm to improve the average flow velocity. The proposed control method is also applied to the particle model, assuming the actual pedestrian flows. The validity is verified with simulations in both the continuum and particle models.