Emergency evacuation dynamics based on evolutionary game theory

Abstract

Whether crowd cooperation behaviors bring benefit for pedestrians evacuating under panic conditions is still an open issue. Motivated by this, an improved cellular automaton model based on evolutionary game theory is proposed, focused on the impact of panic thresholds on evacuation efficiency in different emergency scenarios. Three distinct game paradigms, namely Snowdrift, Stag Hunt and Prisoner's Dilemma games are used to characterize these scenarios where pedestrian evacuation processes. Interestingly, in Snowdrift and Stag Hunt games, pedestrians tend to choose cooperation, therefore queuing orderly; whereas in the Prisoner's Dilemma, they are more likely to defect, exhibiting extremely aggression. The study also explores the impact of varying levels of cooperative behavior and panic thresholds on evacuation efficiency under different conditions. Overall, higher panic thresholds promote evacuation efficiency. The findings provide new insights into pedestrian evacuation behavior and hold significant implications for improving evacuation efficiency and management in real-life scenarios.