A cutting-edge approach to observer-driven finite-time adaptive fault-tolerant control for unmanned surface vessels leveraging advanced neural architectures

This study delves into fault-tolerant control strategies for underactuated surface vessels (USVs) in the context of global tracking control. Targeting the challenge of lateral drive vector alignment in USVs, this paper successfully constructs an effective lateral drive strategy by introducing supplementary control means. At the same time, a robust neural network (NN) adaptive technology is proposed to accurately reconstruct the unknown dynamic behaviour of the ship. A finite-time disturbance observer (FTDO) is developed to real-time reconstruct centralized uncertainties, encompassing unknown external disturbances and approximation errors. This method achieves fast and accurate online reconstruction of centralized uncertainties, reduces the dependence on accurate multi-system vehicle motion models, and thus expands the application scope of online disturbance estimation technology. The study conducts a rigorous stability analysis of the closed-loop system using Lyapunov stability theory. The effectiveness of the proposed control scheme in achieving global tracking is validated through comprehensive simulations.