Active fault diagnosis for AUV over-actuated systems based on UIOs and control allocation

Area of Science:

• Robotics and Control Systems
• Marine Engineering
• Fault Diagnosis and Fault Tolerant Control

Background:

• Autonomous underwater vehicles (AUVs) depend critically on their propulsion systems for navigation.
• Ensuring the safety and reliability of AUV propulsion is paramount for mission success.
• Over-actuated systems offer enhanced control but require robust fault management strategies.

Purpose of the Study:

• To develop an active fault diagnosis framework for a six-degrees-of-freedom (6-DOF) over-actuated AUV propulsion system.
• To integrate unknown input observers (UIOs) with control allocation for fault detection, estimation, and isolation.
• To demonstrate the system's ability to maintain normal operation through internal input adjustments.

Main Methods:

• Investigation of an over-actuated AUV propulsion system with 6-DOF.
• Implementation of an active fault diagnosis framework combining unknown input observers (UIOs) and control allocation.
• Simulation-based validation of the proposed integrated methodology.

Main Results:

• Successful fault detection, magnitude estimation, and isolation in the AUV over-actuated system.
• Demonstrated capability to ensure normal system operation by adjusting internal input allocation.
• Validation of the framework's effectiveness without requiring auxiliary signals.

Conclusions:

• The integrated fault diagnosis framework enhances the safety and reliability of AUV propulsion systems.
• Internal adjustment of control allocation is sufficient for fault management in over-actuated AUVs.
• The proposed method offers a practical solution for real-time fault diagnosis in AUVs.