An Agent-Based Model of Crowd Evacuation Integrating Agent Perception and Proximity Pressure

Abstract

Many crowd evacuation models have been developed in isolation with respect to the aspects of interest. In this paper, we propose an agent-based model of crowd evacuation, pertaining to symmetric utilization of resources in terms of exit usage and evacuation time. The model is based on a game-theoretic decision support, resolving the consequences of proximity pressure on agents of two types; emotional and rational. This model is combined with the model of routing decision, in the presence of two possible information dispersion modalities; without and with inter-agent information sharing; to update agents' perception map. An agent-based simulation using a hypothetical space is performed to evaluate the performance of the model. The simulation results reveal that panic is minimal when agents adopt nearest exit strategy, a mechanized strategy with no human aspects and interactions. However, it takes more time and results in uneven distribution of agents across the exits. In spite of utilizing human aspects, the strategies involving information sharing between agents results in even distribution of agents across the exits with acceptable population panic.