Adaptive Lane-Change Trajectory Planning and MPC-Based Tracking Control for Rear-End Collision Avoidance Under Varying Road Friction Conditions

Abstract

Rear-end collision avoidance is one of the most critical active safety features in modern vehicles. This paper proposes a steering-based rear-end collision avoidance strategy designed to enhance vehicle safety and stability under various driving conditions. Collision avoidance is achieved through lateral maneuvering using a smooth and feasible trajectory generated by a quintic polynomial. To ensure dynamic feasibility and vehicle stability, the minimum feasible lane change time is determined according to allowable lateral acceleration limited by the available tire-road friction. Unlike conventional approaches that assume a fixed lane-change time, the proposed method adapts the maneuver according to road conditions, thereby reducing the risk of instability, particularly on low-friction roads. To accurately track the generated trajectory, a Model Predictive Control (MPC) controller is employed. Simulation results using MATLAB/Simulink demonstrate that the proposed approach successfully achieves collision avoidance while maintaining vehicle stability and acceptable ride comfort under different driving conditions. Furthermore, a comparative analysis with a conventional fixed lane-change strategy shows that the proposed adaptive approach significantly improves vehicle stability and path tracking performance under low-friction road conditions. Specifically, at μ = 0.4, the adaptive strategy maintains tire slip angle within ±1.5°, well below the 3°–4° linear stability limit and reduces peak acceleration by 44% compared to the fixed lane-change strategy. The results confirm that the proposed approach effectively prevents excessive lateral acceleration and vehicle instability, ensuring safe obstacle avoidance on wet roads.