Proposing a cognitive multi-agent model for the panic in crowds phenomenon

The panic in crowds phenomenon is studied by different approaches. Multi-agent system (MAS) simulations can be an excellent tool for understanding and analyzing some real-life panic situations, such as fires and earthquakes. This paper proposes a conceptual model that follows a social-cognitive approach for this kind of phenomenon. By this approach, new structures and variations for collective panic situations can be proposed. Also, this sort of research could bring a better understanding of the principles of the social and psychological fields related to panic in crowds behavior.