FUZZY PERCEPTIONAL SPACING FOR INTELLIGENT MULTI AGENT PEDESTRIAN SIMULATION

Multi agent pedestrian model is a very promising current research topic because it can be applied to many fields. The current multi agent pedestrian models, however, are dealing with individual movements of pedestrian under a single situation. The fact that real pedestrians behave differently under different situations, such as queuing situation and normal condition, have not been addressed by the existing models. Large numbers of pedestrians in one group behave as a platoon which has different walking speeds compared to if they walk as individuals. In this paper, we describe the idea on how to incorporate pedestrian group behavior and multi state situation such as queuing and normal state into multi agent pedestrian simulation seamlessly using fuzzy logic. As the result, by means of this simple scheme the pedestrian agents have been advanced further with capability to behave more realistically in a group or platoon behavior.