Vehicle lateral tracking control optimization based on fuzzy preview time and ant lion algorithm

To enhance the path tracking performance of intelligent vehicles, this paper conducts optimization research on the classical Linear Quadratic Regulator (LQR) controller based on a 2-degrees-of-freedom (2-DOF) vehicle dynamics lateral tracking error model. Aiming at the insufficient adaptability of the LQR controller with fixed weight coefficients at varying vehicle speeds, the Ant Lion Optimizer (ALO) is introduced to dynamically adjust the matrix weight coefficients, and a preview feed-forward steering angle compensation strategy is integrated to improve the lateral path-tracking capability. Furthermore, to address the reduced steering stability of the feed-forward LQR controller caused by model linearization, an adaptive prediction mechanism based on fuzzy control is designed. This mechanism integrates parameters such as vehicle speed, path curvature, and its rate of change. By utilizing a dual-fuzzy controller, a hybrid control strategy that combines dynamic prediction time and fixed preview time is constructed. Simulation verification is conducted via MATLAB/Simulink and CarSim co-simulation. Results show the proposed lateral control method balances tracking accuracy and system stability, with good robustness across speeds—simulation at double lane change 95.66% lower than traditional LQR at 15 m/s , and only 39.74% of traditional LQR's average deviation at 25 m/s. This study offers an efficient solution for intelligent vehicle lateral tracking, addressing fixed-weight LQR and fixed preview time limitations in complex roads.