Design of A Terminal Sliding Mode Controller Based on Extended State Observer for Vehicle Control

This paper aims to present a robust approach for controlling autonomous vehicles with nonlinear dynamics, addressing the key challenges of navigating in unknown environments and performing tasks under uncertain dynamics. Traditional approaches often exhibit limitations in handling uncertainty and convergence speed. To overcome these issues, this paper presents work on integrating a Terminal Sliding Mode Control (TSMC) with an Extended State Observer (ESO) to ensure finite-time convergence and active disturbance rejection. The proposed framework synergistically combines the TSMC controller and the ESO to achieve accurate trajectory tracking and robust disturbance rejection. The method is validated through MATLAB/Simulink simulations across multiple scenarios, confirming its effectiveness in navigating unknown environments under dynamic uncertainty. The proposed methodology demonstrates superior performance compared to traditional PID control. Key results indicate a nearly 60% reduction in settling time (to 1.3 seconds) and the complete elimination of overshoot in linear velocity tracking. Furthermore, the controller achieves approximately 78% higher tracking accuracy (RMSE 0.04 rad/s) while suppressing control vibrations by more than 75%. Meanwhile, ESO provides high-precision state estimation, converging in less than 0.6 seconds with an estimation error of less than 2.5%, and reconstructing system uncertainties with 70% higher accuracy than conventional methods. These results demonstrate that the proposed method is a superior and robust solution for high-performance autonomous vehicle applications operating under dynamic uncertainties and stringent real-time constraints.