An Agent-Based Crowd Dynamics Simulation that Considers Idling and Time-and-Distance-Conscious Optimising Behaviour

Abstract

This agent-based simulation study investigates pedestrian dynamics with a focus on the impacts of behaviour idling on pedestrian flows. It also examines the influence of psychological, social, and environmental factors on pedestrian flows. Our research categorises pedestrian behaviour into three types: time-sensitive (Type A), mobility-constrained (Type B), and 'wandering' type (Type C), defined as pedestrians moving without a specific destination, which includes tourists, shoppers, and leisure walkers. We demonstrate how behaviour heterogeneity influences flow and movement patterns through simulations in unidirectional, bi-directional, and multi-directional pedestrian facilities. We find that Type C pedestrians significantly slow down Type A pedestrians, leading to a speed reduction of up to 30% in high-density tourist scenarios, and cause prolonged stationary periods for Type B pedestrians, particularly in less crowded settings where Type C's tendency to idle is more pronounced. Our results show a linear relationship between density and speed reduction, with tourist behaviour notably exacerbating congestion in high-density environments. Key insights highlight the critical role of wandering (Type C) behaviours in affecting pedestrian flow, emphasising the necessity for urban planning and infrastructure design to accommodate this variability. Future research aims to apply these findings to real-world contexts, further refining urban design strategies to accommodate the full spectrum of pedestrian behaviours.

Keywords

• agent-based
• heterogeneous
• pedestrian dynamic
• pedestrian behaviour
• microsimulation
• crowds