Simulation method for disaster mechanism caused by panic emotion of evacuation crowds

The current state of crowd simulation technology remains inadequate in simulating stampede scenarios and predicting resultant casualties. To simulate how panic-induced crowd stampedes occur, this study develops a simulation model of crowd behaviors under panic conditions, incorporating key mechanisms such as squeezing, pushing, and falling. The model is based on an agent-based evacuation simulation approach influenced by social factors and incorporates a “panic factor” into the agents. To quantify crowd-scale features, time points, congestion structures, and other elements in panic-induced stampede disasters, we developed a benchmark simulation environment and referenced existing agent classification techniques. The study describes the mechanism and progression of stampede disasters in panicked crowds. Simulation results demonstrated strong concordance with empirical disaster data, confirming the validity and accuracy of the findings.