Cooperative Look-ahead Lane Change System for Improving Driving Intelligence of Automated Vehicles in Critical Scenarios

Traffic accidents often result in quick bottlenecks and increase injudicious lane changes near incidents (or lane blockage), worsening collision risks, congestion, and fuel consumption. As a practical solution, this paper proposes a novel cooperative look-ahead lane change (Co-LLC) system for automated vehicles (AVs) to mitigate sudden accident-induced traffic effects by improving driving intelligence and safety in critical scenarios. The proposed Co-LLC system comprises the state prediction model, safety and impact evaluation unit, and decision system. Firstly, we analyze the immediate impacts of traffic accidents and identify that a lack of anticipation causes a lane change hot spot near the incident. Consequently, most attempts to change lanes early are unsuccessful due to uncooperative behavior from vehicles in the destination lane. Secondly, we design anticipatory and cooperative lane change systems for AVs to decide the need and feasibility of a lane change in advance. Thus, the proposed system enables AVs to change lanes smoothly and cooperate with other vehicles during lane changes. Finally, we investigate the impact of different penetration rates of AVs using the proposed system on overall traffic performance. The performance of our proposed system is compared to the traditional driving system, and the results show that our proposed system improves the lane-changing behavior of AVs, assists traditional vehicles in changing their lanes smoothly, and mitigates sudden accident-induced traffic impacts. Moreover, the proposed system improves the overall traffic performance with increased penetration rates. Our proposed system is computationally efficient and suitable for real-time driving in critical traffic scenarios.