Abstract: In order to ensure the safety of driving in foggy weather, a microscopic model of car following and lane changing traffic flow is established based on cellular automata. In view of the instability of car following state in foggy weather, dynamic random acceleration is introduced to segment modeling. The simulation accuracy is improved by refining cell size and cell step size and adding position update allowance. The relationship curves of speed and vehicle dispersion with time and the speed density diagram are obtained. The results show that: the vehicle speed fluctuates greatly in medium fog; the dispersion of vehicle speed in dense fog is large; the number of lane changing decreases with the increase of fog concentration, which is consistent with the actual traffic characteristics. The model can partly reflect the microscopic characteristics of traffic flow, and more truly reflect the car following behavior in fog.

Key words: traffic engineering, car following in fog, cellular automata, traffic flow, computer simulation