Autômatos celulares aplicados na modelagem da dinâmica de pedestres em ambientes com obstáculos transponíveis e fogo

Abstract

This work presents a model based on probabilistic cellular automata used to represent the movement dynamics of people (pedestrians) in indoor environments during fire events. The developed model extends a study from the literature by adopting alternative calculations for the static floor field and introducing new features, such as the inclusion of both non-traversable and traversable obstacles, as well as the propagation of fire sources and their influence on pedestrian movement throughout the environment. The goal is to increase the realism of evacuation simulations by incorporating these characteristics into the original model. The results indicate that the adoption of alternative calculations for the static floor field enabled the successful introduction of obstacles into the model. Moreover, the presence of traversable obstacles contributed to improved simulation efficiency in certain environment configurations, especially in the presence of fire, when pedestrians tend to avoid regions near the flames. Overall, the implemented extensions enhanced the model's capability to represent complex pedestrian behaviors while preserving the inherent simplicity of cellular automata.