Mesoscopic evacuation model based on adaptive grids partitioning

Abstract

Crowd evacuation stands as a critical subject in urban planning and emergency management. However, owing to the intricate and unpredictable nature of crowd behavior, coupled with the diverse urban environments, achieving effective crowd evacuation remains a formidable challenge. To delve into the nuanced variations during crowd evacuation, we introduce a mesoscopic evacuation model based on adaptive grids partitioning. This model involves performing adaptive grids partitioning according to crowd density within the designated grid density analysis range in the evacuation scenario. In this approach, we conduct evacuation simulation experiments by simulating pedestrian flow through density flow between grids. The experimental results demonstrate that adaptive grids partitioning aptly responds to microscopic changes in the crowd evacuation process. The mesoscopic evacuation model proposed in this paper notably enhances the utilization rate of evacuation exits. The comparative visualization simulation experiments using Pathfinder software attests to the smoothness and reliability of the model.