Experimental model of pedestrian behavior in avoidance of small moving obstacles

Abstract

This study modeled the behavior of pedestrians who avoids moving obstacles, especially other pedestrians or bicycles are approaching, which are common in daily life. In our previous research, we modeled the behavior of a pedestrian when they avoided a static obstacle, such as a parked car, parked bicycle, or poles on a roadside. However, pedestrian motion might be affected by movement of an obstacle or its size. Accurate behavior modeling of pedestrians should include these factors. Typically, previous studies have treated pedestrian motion as a deterministic model. However, human behavior should be considered as stochastic motion even though deviations, such as walking speed, are not very large when considering local motion. When pedestrians avoid an obstacle, their motion or behavior is determined by their own individual factors, such as age, gender, etc. In this paper, we focus on the effect of obstacle speed and relative distance. To model pedestrian behavior during avoidance of a moving obstacle, walking loci were measured at first. Then, we derived and evaluated differences in avoidance behavior with respect to the obstacle speed. One result showed that the speed of a moving obstacle is one of the most important factors as the pedestrian begins avoidance, that is, the time factor that is determined by the relative speed and the relative distance affects the timing of avoidance behavior of a pedestrian.

Keywords

• safety
• pedestrian protection
• collision prediction
• moving obstacle
• avoidance