Modeling time headways between vehicles has attracted increasing interest in the traffic flow research field recently, because the corresponding statistics help to reveal the intrinsic interactions governing the vehicle dynamics. However, most previous micro-simulation models cannot yield the observed log-normal distributed headways.
This paper designs a new car-following model inspired by the Galton board to reproduce the observed time-headway distributions as well as the complex traffic phenomena.
The consistency between the empirical data and the simulation results indicates that this new car-following model provides a reasonable description of the car-following behaviors.
[Keywords: traffic flow, car-following, log-normal distribution]
