Modeling impacts of driving automation system on mixed traffic flow at off-ramp freeway facilities

Abstract This study aims to model a mixed traffic flow consisting of vehicles controlled by a driving automation system in the highway off-ramp system. The driving automation system includes no automation (NA), driver assistance (DA), partial automation (PA), and conditional automation (CA). Developing this model requires specifying the longitudinal and lateral driving characteristics of vehicles with different driving automation systems and then selecting appropriate car-following and lane-changing models. The proposed cooperative lane change model applied by CA considers the lane utility and car-following models of the following vehicles in adjacent lanes that can be obtained by vehicle-to-vehicle communication. The impacts of the market penetrations of PA and CA on traffic flow are investigated, and it is found that both PA and CA can improve traffic capability. Compared to PA, CA is more effective in improving the highway capacity near the off-ramp. The impact of CA on the headway indicates that CA enables vehicles that require changing lanes to find suitable gaps and changing lanes in a timely manner, which improves the capacity of the off-ramp and the overall freeway. By adjusting the lane change adventure factors, the contribution of CA to capacity can be further enhanced without compromising driving comfort and safety.