Hybrid fault-tolerant adaptive controller for robust trajectory tracking of mobile robots

Abstract

Autonomous differential-drive robots face significant challenges in trajectory tracking under disturbances, model uncertainties, and actuator degradations. This paper proposes a hierarchically decoupled dual-observer–based fault–disturbance separation architecture for robust trajectory tracking of wheeled mobile robots. An improved fault observer (IFO) explicitly reconstructs actuator faults with adaptive initialization for smooth compensation, while a high-order integral-chain differentiator (HICD) performs state differentiation and disturbance estimation with noise attenuation. The reconstructed fault is incorporated into both the kinematic fuzzy backstepping layer and the dynamic adaptive sliding mode control (ASMC) layer, forming a cross-layer fault compensation mechanism. A unified Lyapunov analysis establishes uniform ultimate boundedness of the observer and convergence of the tracking error without requiring prior disturbance bounds. Simulation results under multiple actuator fault scenarios demonstrate significant reductions in tracking error (up to 82% root mean squared error (RMSE) reduction vs. backstepping sliding mode control (B-SMC)) and improved robustness against disturbances and parameter variations. The proposed framework provides a structurally integrated solution for fault-tolerant trajectory tracking of nonholonomic mobile robots.