Proactive Longitudinal Control of Connected and Autonomous Vehicles with Lane-Change Assistance for Human-Driven Vehicles

Vehicle automation and connectivity enable the cooperative platooning control of connected and autonomous vehicles (CAVs), which can enhance traffic safety and alleviate traffic oscillations. However, CAVs and human-driven vehicles (HDVs) will coexist on roads prior to the pure CAV environment, creating a mixed-flow traffic environment. Mixed-flow traffic introduces challenges for CAV operations in terms of lane-change maneuvers of HDVs in adjacent lanes, which can generate oscillations, jeopardizing the performance of platooning control. Hence, there is a need to explore the interactions between the CAV and the HDV in the lane-change process, and analyze how CAVs can proactively respond to HDV lane change to dampen the generated disturbances. This study proposes a deep reinforcement learning-based proactive longitudinal control strategy for the CAV to assist the HDV lane change, and minimize the adverse impact on the smooth operation of the mixed platoon. Results from numerical experiments illustrate the effectiveness of the proposed control strategy. Further, the generalizability of the proposed control strategy for different HDV driver types is demonstrated.