Fault diagnosis and reconfigurable control of autonomous underwater vehicles with actuator and sensor faults

Actuator and sensor faults pose significant threats to the operational safety of autonomous underwater vehicles (AUVs). To address this issue, a fault-diagnosis-based control reconfiguration strategy is proposed. The proposed strategy addresses mixed fault scenarios involving simultaneous actuator and sensor faults. It explicitly accounts for bounded external disturbances and convergence rate requirements, while balancing robustness and control conservativeness. A fault diagnosis scheme is developed to accurately detect and estimate mixed faults in the presence of external disturbances. Furthermore, a reconfiguration mechanism is introduced to enhance the transient performance during the controller switching process. Subsequently, based on the fault estimation results, a gain-adaptive integral sliding mode fault-tolerant controller is designed. Without requiring explicit disturbance information, the proposed controller allows the control gain to adapt to real-time disturbance variations, thereby ensuring robustness while significantly reducing control conservativeness. Simulation studies under various conditions validate the effectiveness of the proposed reconfiguration strategy.