A CA-based model of dyads in pedestrian crowds: The case of counter flow

The calibration and validation of pedestrian dynamics simulation require the acquisition of empirical evidences of human behaviour. In this framework, this paper firstly presents the results of an experimental study focused on the negative impact of counter flow and grouping on pedestrian speed. In particular, we focused on two member groups (dyads) as the most frequently observed and basic interacting element of crowds. On the basis of the behavioural effects observed with the experiment, a novel cellular automaton is proposed to represent the different behaviour of individuals and dyads, with particular reference to the group spatial alignment and the dynamic leader-follower structure. This has been demonstrated to modulate the speed of dyads by maintaining the spatial cohesion among the two members. In addition, the model is also able to reproduce the significant impact of flow ratio observed in the experiment results.