2024 年 89 巻 819 号 p. 808-817
This study aims to assess the applicability of multi-agent simulation for crowd control after large-scale events. It focuses on pedestrian pathways around a stadium in urban area. By utilizing real-world pedestrian flow data, we investigated the parameters essential for simulation and analyzed crowd control. Consequently, we identified critical simulation parameters, such as speed-density relationships and personal distance, which produced highly accurate results across different days. In terms of congestion control, we established an acceptable congestion range based on data and illustrated that averaging peak bottleneck flows effectively alleviates congestion using the simulation.