Actuator fault detection system for a remotely operated vehicle

This paper addresses the problem of actuator fault detection for a Remotely Operated Vehicle widely used in navy and marine industries. Firstly, the dynamical model for the Remotely Operated Vehicle is presented. Then this model is used to develop a diagnostic system based on a nonlinear (Thau) observer for residual generation, and on a sequential change detection algorithm for residual evaluation. The diagnostic system is tested through simulation trials in which a fault is injected into one of the actuators. The analysis of the results proves that the developed detection system can correctly detect a fault on one of the actuators by using only measurements of sway, surge and yaw (and their speeds), thus representing an effective solution to the problem of actuator fault detection for the Remotely Operated Vehicle.