Modeling the interrelationship forces in pedestrian dynamics during evacuation processes using bond graph [EMBARGOED]

Abstract

Pedestrian simulation is significant in transport station management, building evacuation, and massive public event safety management. The social force model is one of the most widely used continuous pedestrian flow models, and it supports all the use cases mentioned above. However, an important difficulty arises from the limited representation of these forces, particularly in emergency scenarios, which impedes appropriate safety solutions for evacuation operations. Current modelling methodologies have problems in capturing the complex nature of interrelationships within pedestrian dynamics during evacuations, preventing a thorough understanding of energy transfers and interactions. This study aims to bridge the gap between complexity and usability by introducing a bond graph model to represent interrelationship forces in pedestrian dynamics during evacuations and analyzing the developed model using relevant variables to understand the system's dynamics, including aspects related to motion and the movement of variables. We propose transforming the bond graph model into a translation mechanical submodel for performing a thorough analysis of momentum and effort variable rate changes. The simulations are conducted using 20SIM software, examining diverse scenarios. The research analyzes several various evacuation scenarios with varying masses, locations, and closeness to exits to understand how people behavior in different situations. Results obtained from the proposed model demonstrate that the pedestrians who are closer to the wall experience greater spatial constraints, which can impede their initial movement and result in lower initial values of momentum. For instance, under the same conditions, the initial rate of change of momentum (dp/dt) for a pedestrian near a wall was found to be significantly lower compared to a pedestrian in the middle of the hall, showing a difference of approximately 29%, and took 29% longer to reach a steady state compared to those in the middle of the hall. Furthermore, a greater number of pedestrians, especially those who are close to walls, experience larger changes in momentum and take longer to stabilize, indicating heightened anxiety. Additionally, the layered view which is structured based on proximity to the chosen pedestrian shows that pedestrians in the closest layer exhibit the highest values of effort exerted at the steady state. The research also establishes a direct relationship between the pedestrian’s mass and the initial effort exerted, When the chosen pedestrian is heavier than others, the effort exerted tends to be positive, increasing as the mass of others decreases. Conversely, encountering heavier individuals may hinder movement, resulting in negative initial effort values. This research can play a vital role in future developments of interrelationship forces between pedestrians during emergencies.