Composite Learning Adaptive Intelligent Self-Triggered Fault-Tolerant Control with Improved Performance Assurance for Autonomous Surface Vehicle

Aiming at the trajectory tracking control issue of the autonomous surface vehicle (ASV) subject to unknown actuator failures, a composite learning adaptive intelligent self-triggered fault-tolerant control design with improved performance assurance is proposed in this paper. Initially, an enhanced fixed-time performance function is introduced to construct an expected tight feasible area such that the fixed-time convergence of the tracking errors can be achieved without satisfying the specific form of fixed-time stability. Then, with benefits from the outstanding fuzzy modelling and detail analysis capabilities of fuzzy wavelet neural networks (FWNNs), a nonlinear disturbance observer-based composite neural learning strategy is proposed for handling the unknown dynamics and compound disturbance, which provides a practicable method to improve approximate precision and robustness against the unknown disturbances. Furthermore, by constructing the self-triggered mechanism and fault-tolerant mechanism, an adaptive fault-tolerant trajectory tracking controller with the self-triggered feature is developed, which ensures that entire signals in the closed-loop system (CLS) are semi-globally uniformly ultimately bounded, and the tracking errors can converge to a predefined neighbourhood of the zero satisfying a pre-specified tracking accuracy even if actuator failures occur suddenly. Finally, the validity and superiority of the developed approach are verified through simulation results.