Thermo-inspired model of self-propelled hard disk agents for heterogeneous bidirectional pedestrian flow

Abstract

The aim of this work is to present a hybrid pedestrian model that describes the bidirectional walking of crowds in a corridor. The self-propelled hard disk model combines both perceptual as well as decision-making characteristics of situated agents. This model considers homogeneous and heterogeneous groups with relaxed, standard, and hurried pedestrian profiles according to their walking preferences. The situated agent perception module considers one radius in the base model, while the extended model considers three radii to improve the pedestrian evasion rules. We have explored the corresponding changes in the formation of spatial walking patterns. The findings include comparative variations in the formation of following lanes, the time-based evolution of crowds in both homogeneous and heterogeneous groups, and the transition from transient traffic jams to the appearance of segregation waves emerging from self-organization mechanisms. In this point, we propose a thermodynamics-inspired model associated with the collective behavior of pedestrians. A macroscopic ideal gas-type equation of state is established to define the dependence between volume, pressure, and temperature of pedestrian crowds in counterflow. To achieve this goal, the temperature, defined as T∼vα¯ where v is the walking speed of individual pedestrians, must possess an α parameter in the interval [1,2]. The experimental simulations include binary contact interactions and collisions of pedestrians with the corridor borders. The macroscopic description of the model inspired by thermodynamics can be complemented with the usual analysis of the fundamental diagram.