Robust tube-based model predictive tracking control of autonomous vehicles considering lateral and longitudinal coupling

This paper investigates the tube-based robust model predictive control problem for autonomous vehicles, which can address the complexities of lateral and longitudinal coupling dynamics along with multiple system state and control input constraints. Firstly, an integrated vehicle tracking model by T-S fuzzy technology is established by combining lateral dynamics and longitudinal kinematics, where parametric uncertainties and external disturbances arising from fluctuations in vehicle speed and road conditions are considered simultaneously. Due to the strong coupling inherent in tire-road forces, a tire friction circle is constructed to encapsulate the interplay between longitudinal and lateral forces. The friction circle can approximate by octagon, which facilitates the translation of vehicle tracking dynamics related to coupled friction forces into a framework of convex constraints. Then, a linear tube-based model predictive controller and a fuzzy tube-based model predictive controller are designed for both the longitudinal and lateral movement adjustment of the vehicles. Finally, the effectiveness of the proposed controller design method is illustrated in the Carsim/Simulink co-simulation environment. The simulation results show that the method proposed in this paper has better tracking performance in both longitudinal and lateral dimensions compared to traditional linear time-varying-model predictive control approaches.