A Control-Oriented Thruster Management Framework for Fault-Tolerant Propulsion of Remotely Operated Vehicles

Remotely operated vehicles (ROVs) typically rely on over-actuated propulsion systems to achieve precise dynamic positioning and maneuvering in complex underwater environments. In practice, however, conventional propulsion management based on thrust allocation is often challenged by non-ideal actuator behaviors, such as cavitation-induced thrust degradation, low-speed dead-zone effects, inter-thruster coupling, and partial actuator failures. Most existing approaches treat propulsion management as a static force distribution problem and implicitly assume ideal or fast thrust execution, which limits performance when actuator dynamics and execution uncertainty become significant. To address these limitations, this paper proposes a control-oriented thruster management framework that reformulates propulsion management as a feedback regulation problem rather than a static allocation task. In the proposed framework, actuator dynamics and thrust execution uncertainty are explicitly incorporated into the control loop. At the actuator level, thrust degradation and low-speed operation are compensated through disturbance-aware feedback control, while at the system level an LQI-based controller with thrust response compensation is employed to coordinate multi-degree-of-freedom (DOF) force and moment regulation and suppress cross-axis coupling. Fault tolerance is achieved inherently through feedback regulation without relying on explicit fault detection or reallocation. Experimental results obtained from an ROV propulsion platform, including single-thruster tests, coupled multi-DOF control, and a thruster shutdown scenario, demonstrate improved thrust executability, reduced coupling-induced disturbances, and enhanced fault-tolerant performance compared with conventional direct thrust allocation strategies.