Prescribed performance cooperative fault-tolerant control for gripper-based underwater salvage robot based on a Sliding Mode Extended State Observer

To address composite actuator faults and environmental disturbances during cooperative underwater salvage operations involving a Gripper-type Submersible Recovery Robot (GSR) and an Unmanned Surface Vehicle (USV), this paper proposes a prescribed-performance cooperative fault-tolerant control strategy based on a Sliding Mode Extended State Observer (SMESO). First, a cooperative control framework with relative position constraints is established by considering the fixed offset required in practical GSR–USV formation. At the kinematic level, prescribed performance control (PPC) is introduced through performance functions and error transformation to confine cooperative tracking errors within predefined transient and steady-state bounds, thereby improving system safety in the presence of actuator faults. At the dynamic level, an SMESO is designed to estimate and compensate for lumped uncertainties caused by external disturbances and thruster faults in real time. In addition, a logarithmic mapping-based projection algorithm is developed to address the coupling between thruster saturation and fault effects, thereby enhancing system robustness and fault-tolerant capability. Comparative simulation results show that, under composite fault conditions, the proposed method achieves better tracking accuracy, faster convergence, and stronger disturbance rejection than the benchmark controllers. Additional simulations under sinusoidal and elliptical trajectories further verify the robustness and generalization capability of the proposed framework. These results demonstrate the effectiveness of the proposed method for high-precision underwater cooperative salvage missions.