Models of pedestrian evacuation based on cellular automata

Abstract

We study two models of pedestrian evacuation based on cellular automata and intelligent agents. We use the static floor field approach and augment it with additional transition rules (random movement and preservation of personal space) in order to model various aspects of human behaviour. Through numerical simulations we investigate pedestrian evacuation from rooms with various geometries (e.g. empty room, classroom). Using heatmaps (density diagrams of pedestrians), average evacuation times, velocity and other parameters we study the effectiveness of evacuation and potential dangers pedestrians may encounter.