Flow capacities from cellular automata modeling of proportional splits of pedestrians by direction

This paper describes the design and application of an Artificial Life cellular automata-based (CA) microsimulation that yields speed-flow-density characteristics, or fundamental properties, for various proportional splits of pedestrians by direction. It is shown that a CA model using a relatively small rule set is capable of effectively capturing collective behaviors of pedestrians who are autonomous at the micro-level. The model captures emergent pedestrian stream characteristics for one, two, cross-, and four directional flows and reveals how much stream formation and directional splits are important factors in determining walkway capacities. For all flows studied capacities occur at about 30 percent occupancy, a guide for channeling maximum flows. The results agree with and extend well-established fundamental properties of pedestrian flows according to the directional splits.