Simulation of Pedestrian Evacuation Behaviour in Fire Based on Smoke Field Dynamics and Metacellular Automata Models
摘要
Urban public transport hubs are tasked with facilitating rapid interchanges, providing passengers with a convenient and efficient transfer experience; however, they also face significant fire risks. This paper establishes a virtual evacuation environment utilizing the Unity platform to simulate pedestrian behavioural choices under fire conditions, taking into account factors such as line-of-sight obstruction, smoke fields, fire source locations, and obstacles. It comprehensively analyses specific pedestrian behavioural characteristics in these scenarios. A metacellular automata model of pedestrian evacuation in smoke conditions is developed by integrating the classical smoke diffusion model, with evacuation simulations conducted using MATLAB. The results indicate that smoke not only restricts pedestrians’ fields of view but also produces a notable dispersion effect. Interestingly, when smoke is located far from the exit, it may assist pedestrians in locating the escape route. Consequently, when designing escape routes and exits, it is crucial to consider the impact of visibility changes, ensuring that the area around exits remains unobstructed. Additionally, it is advisable to minimize the placement of electrical equipment near exits to create effective control measures that promote the safe evacuation of individuals.