The Effect of Grouping in Human Crowd Movement

Human crowd modelling has gained importance for floor plan designs and crowd movement management. It increases the need for more realistic modelling of human-like behaviours. The state of the art in modelling studies often focuses on homogeneous crowds. Grouping behaviour has been rarely addressed. Social studies show that crowds are often inhomogeneous, which are containing different social groups. How groups in crowds interact with each other has been rarely investigated. Exploring the interactions of groups in such inhomogeneous crowds becomes important to understand its impact on crowd movement. Therefore, this study aims to model the dynamics of inhomogeneous crowds including group interactions.
 
   This study starts by building a model that captures the grouping behaviour of individuals in an inhomogeneous crowd. The formation of grouping and non-grouping behaviours was investigated by simulating the model in a narrowing corridor, a turning corridor, a T-intersection corridor, and a corridor with an obstacle. The results of the simulations of grouping and non-grouping behaviours were analysed to explore which behaviour was effective for crowd movement.
 
   This study found that grouping behaviour helped to achieve potential flow rates by preventing turbulence. In contrast, non-grouping behaviour created turbulent flows of individuals and inefficient movement. This study suggests that letting groups stay together is potentially effective for crowd management when a large number of individuals are egressing the venue.