Abstract:
In order to study the influence of large vehicles' lane-changing and sensitive driving behaviors on the highway traffic flow, a two-lane mixed traffic cellular automate (CA) model was built by introducing the lane-changing rule into the sensitive driving Nash (SDNS) model, on the assumption that the vehicles with different lengths and maximum velocities appear on the two lanes simultaneously. Under the periodic boundary condition, the fundamental diagrams of the mixed traffic velocity, flow, and density were obtained through numerical simulation by setting different values for parameters such as the lane-changing probability, mixing ratio, braking probability, vehicle length, and maximum velocity. The results show that the system critical density, the maximum average velocity, and the traffic flow decreases with the braking probability increasing. When the proportion of long and slow vehicles is more than 50%, the mixed traffic is more prone to congestion, and the lane-changing rate is less than 5%; when the proportion of long vehicles reaches 50%, the traffic flow peak is merely 0.42, which is 20% less than that of all small/short vehicles. The major factor causing lane changing difficult is the proportion of long vehicles. When the lane-changing probability is 50% and the proportion of long vehicles in the system increases from 25% to 75%, the maximum lane-changing rate decreases from 6.32% to 2.78%.