Fault Diagnosis of an Underwater Remotely Operated Vehicle Through Structural Analysis

This paper introduces a model-based approach for diagnosing actuator faults in an underwater remote operated vehicle (ROV). The method employs structural analysis techniques to perform diagnosability analysis and to generate residuals, aiming to detect actuator faults at an early stage and issue timely warnings. To achieve this objective, we construct the mathematical model of the underwater robot, which facilitates the development of the system's structural model. This approach yields parity equations that serve as residual generators. A key advantage of the proposed method is its ability to provide practical solutions for residual generation in nonlinear systems. The CUSUM algorithm is used to detect changes in residual signals. The underwater remote operated vehicle BlueROV2 serves as the robotic platform for experimentation.