Departure headways at signalized intersections: A log-normal distribution model approach

Abstract

Modeling departure headways at signalized intersections attracts constant research efforts in the last four decades due to its importance. But most previous approaches focus on the mean departure headways only. In this paper, two interesting findings are presented. First, the distributions of the departure headways at each position in a queue are revealed to approximately follow a certain log-normal distribution (except the first one) and the corresponding mean values level out gradually. Second, a car-following model from the driver’s perspective is proposed to explain why the departure headways follow such a log-normal distribution. The consistency between the empirical and simulation results indicates that this new car-following model provides an acceptable description of the unseen interactions between the vehicles in a discharging queue. The new model is also useful for simulation based intersection capacity analysis and traffic signal control.