Grouping Methods for Facilitating Emergency Evacuations

Abstract

Abstract It is widely believed that the motion of crowds can be described by the “social force” model, and small group behaviors have great impacts on evacuation efficiency as a common phenomenon in pedestrian evacuations. Some previous works have modeled group behaviors in social force models by depicting group members’ interaction forces. However, the impacts of different grouping schemes have not been considered and evaluated. This paper proposes four different grouping methods to explore the evacuation dynamics of different group structured pedestrians. Results show that group evacuations based on the spatial partitioning outperform those clustering-based schemes with the larger group number. More importantly, simulations indicate that the number of groups also highly affects the evacuation efficiency, and the larger the number of groups, the more efficient the evacuation.