Simulation of pedestrian contra-flow by multi-agent DEM model with self-evasive action model
A pedestrian contra-flow can be seen frequently in urban spaces such as a pedestrian crossing, an open space and a concourse. A pedestrian simulation with taking a pedestrian contra-flow into account will be a rational tool for designing urban pedestrian facilities. The objective of our research is to develop a microscopic model for a pedestrian contra-flow. Although we have been developing a DEM (DistinctElement Method)-based microscopic model of crowd behavior, this conventional crowd behavior model does not necessarily provide satisfactory agreement with an observation in a pedestrian crossing with heavy commute traffic. In this paper, we newly develop a self-evasive action model, which can describe either a collision avoidance or an alignment behavior between adjacent pedestrians. The multi-agent DEM model with the self-evasive action model shows good predictions of the characteristics of pedestrian contra-flow.
Keywords: Pedestrian Traffic, Self-Evasive Action Model, DEM, Microscopic Model, Multi-Agent Model