Nonsingular terminal sliding mode controller-based path tracking control for autonomous vehicles considering multiple-uncertainties disturbances

Autonomous vehicle path tracking is faced with the influence of multiple-uncertainties disturbances such as modeling inaccuracy, parameter perturbation, network delays and external environmental disturbances, which can easily lead to the degradation of tracking accuracy and driving safety. For this kind of problem, this paper proposes a path tracking fusion control strategy combining nonsingular terminal sliding mode controller (NTSMC) and radial basis function neural network (RBFNN) to effectively improve the influence of multiple-uncertainties disturbances. Firstly, a tracking error model considering multiple-uncertainties disturbances is established, on the basis of which a nonsingular terminal sliding mode surface is designed. Then, the adaptive law of RBFNN is derived to realize the real-time estimation and compensation of the multiple-uncertainties disturbances by adopting Lyapunov stability theorem, which constitutes a composite NTSMC-RBFNN controller. The complex path tracking problem is simplified to a convergence problem of the tracking error in the finite time domain and the closed-loop system stability is proved through Lyapunov theorem. Finally, experiments are implemented based on Carsim-Simulink and hardware-in-loop (HiL) platform, respectively. The results show that the proposed controller can make the system error converge in a finite time with good real-time performance and tracking ability.