Using microscopic pedestrian simulation statistics to find clogging regions

A cellular automata model allowed pedestrian flows to be investigated at a microscopic level. Hence, this microscopic model has been proposed to simulate pedestrian movement. In this paper, a four diretions pedestrian flow statistics produced from hill climbing, simulated annealing, and genetic algorithm operators techniques is examined. We further investigated pedestrian clogging regions using cell attractiveness statistics from our heat map agent. The aim of the experiment is to use the heat map agent to compute and to identify the clogging area in the layout generated. Results from the experiment have shown promising solutions in identifying a `bad' region using heat map agents. It is found that viable layouts with higher fitness values produced improved layouts with less blockage effect. Also, the blockages appear to be near exits, T-shape junctions, wider lanes without obstacles and at the biggest density of the grid layouts.