Robust Adaptive Path Tracking Control Scheme for Safe Autonomous Driving via Predicted Interval Algorithm

This paper introduces a robust adaptive path-tracking control scheme via a predicted interval approach for safe autonomous driving tasks under uncertainties. Specifically, a recursive least squares-based set-membership mechanism is firstly designed to estimate a bounding set of acceptable values to depict the uncertain parameters. Based on the estimated system parameters, an interval predictor is deployed to improve the prediction accuracy in the primary control action design. Successively, the unconstrained output feedback-based robust controller is proposed to yield the closed-loop system stabilization by utilizing predicted interval output only. Meanwhile, a model predictive control technique is conceived from solving an optimization problem that is given in the interval predictor to ensure robust constraint satisfaction. The recursive feasibility of the controlled system is theoretically analyzed by applying the nonconservative Lyapunov function with a novel structure and the closed-loop system possesses the input-to-state stability criteria. Finally, simulation results are provided to verify the efficacy of the presented strategy under various intricate scenarios. The results show that the suggested controller always maintains its cross-tracking error and longitudinal velocity error at the lowest level even in the most challenging weather scenario.