Robust Fault-Tolerant Control for Underwater Vehicles With Saturation-Triggered Thruster Health Estimation

This letter presents a saturation-triggered adaptive fault-tolerant control framework for underwater vehicles to handle concurrent thruster failures and environmental disturbances. Unlike existing methods that suffer from performance degradation when disturbances exceed predefined bounds, this paper proposes a two-stage architecture: a projection-based adaptive robust controller for nominal operation and a recursive least squares (RLS) estimator activated only upon observer saturation. This design enables effective fault compensation without conservative disturbance bound selection while maintaining computational efficiency through saturation-based activation. This work establishes uniform ultimate boundedness (UUB) of the closed-loop system under both normal operation and parameter bound violations through Lyapunov analysis. Experimental validation on a BlueROV2 Heavy platform across three scenarios—station-keeping, circular trajectory tracking, and figure-eight maneuvering—demonstrates that the proposed method maintains precise control under multiple thruster failures that preserve system controllability.