Car-Following Behavior Based on LiDAR Trajectory Data at Urban Intersections

The characteristic analysis of car-following behavior in different scenarios is important for studying microscopic traffic flow models. This study first used roadside light detecting and ranging (LiDAR) to collect traffic 3-D point cloud data to reveal the driving behavior characteristics of localized urban intersections. The background subtraction method is used to filter out the background objects. The density-based spatial clustering of applications with noise algorithm detects and identifies vehicle objects. The traffic target tracking method based on historical frame data fusion obtained the vehicle trajectory. Second, this article designed a method to extract high-resolution microscopic traffic information. This method constructs the boundary box model of traffic objects by finding fixed point features. Then, the speed, distance, and other information are calculated through the coordinate relationship of vehicle position. Finally, the velocity and gap distance characteristics are analyzed in car-following behavior. This study found a significant correlation between the speed of the front and rear vehicles in the car-following process. A linear function can effectively fit the relationship between speed and gap distance. This article also verifies the asymmetric characteristics of driving behavior under acceleration and deceleration conditions at urban intersections.