Modeling pedestrian crossing activities in an urban environment using microscopic traffic simulation. (February 2013)

• Record Type: Journal Article
• Authors: Suh, Wonho; Henclewood, Dwayne; Greenwood, Aaron; Guin, Angshuman; Guensler, Randall; Hunter, Michael P; Fujimoto, Richard
• Is Part Of: Simulation. Volume 89:Number 2(2013)
• Page Start: 213
• Page End: 224
• Publication Date: 2013-02
• Subjects: traffic simulation -- pedestrian simulation -- multiagent simulation -- Social Force Model; Computer simulation -- Periodicals; 003.3
• DOI: 10.1177/0037549712469843
• Languages: English
• ISSNs: 0037-5497

Abstract

Microscopic traffic simulation tools are increasingly being employed as an integral part of modeling vehicular traffic and pedestrian activity. However, the complexity of pedestrians' behaviors and their interactions with the various components of the traffic network is commonly under-represented in simulation models, resulting in potentially misleading analyses. This paper explores modeling pedestrians at the microscopic level, attempting to replicate observed pedestrian behavior at a crosswalk in Midtown Atlanta, GA. The VISSIM® simulation tool, with pedestrian movement based on the Social Force Model by Helbing and Molnár (Social Force Model for pedestrian dynamics. Phys Rev E 1995; 51: 4282-4286), is utilized. Comparative results of field observations and the microscopic modeling of the same vehicle and pedestrian activity are presented. A primary observation is that, for the intersection studied, only a small percentage of pedestrians appear to comply with the pedestrian signal indication, with the vast majority of pedestrians exhibiting gap-seeking behavior, crossing when a gap is available, regardless of the signal indication. This results in potentially significant over-estimates of wait time if high pedestrian signal compliance rates are assumed. A second notable observation is that the pedestrian crossing behavior is strongly related to the cross-street traffic queue clearance time and subsequent traffic flow. Capturing this interaction significantly enhances the models' ability to reflect the observed field performance.