Lateral collision avoidance stability control study of intelligent vehicles on low adherence roads

To enhance lateral collision avoidance safety for intelligent vehicles on various road surfaces, this paper proposes a collision avoidance algorithm specifically designed for low-adhesion roads. The study begins by a fuzzy control lane change decision model that considers the relative state information between the ego vehicle and the front vehicle. Next, the environment risk potential field is modeled, with multiple lane change paths of varying parameters considered as candidate trajectories. Critical risks during vehicle steering are analyzed, leading to the design of a dynamic instability boundary risk point potential field. Trajectory optimization is then performed using the Particle Swarm Optimization (PSO) algorithm. Finally, a fuzzy PID controller is developed to track the optimal trajectory. Simulation results demonstrate that, compared to the traditional artificial potential field method, the proposed collision avoidance algorithm provides better lateral stability in the planned paths. Additionally, the fuzzy PID controller designed in this study outperforms the Quantitative Feedback Theory (QFT) controller in terms of tracking accuracy.