Data–Driven Fault Detection and Isolation of the Actuators of an Autonomous Underwater Vehicle

This work proposes the development of a scheme for the fault diagnosis of the actuators of a simulated model accurately representing the behaviour of an autonomous under-water vehicle. The Fossen model usually adopted to describe the dynamics of the underwater vehicle has been generalised in this paper to take into account time–varying sea currents. The proposed fault detection and isolation strategy uses a data–driven approach relying on multi–layer perceptron neural networks that include auto–regressive exogenous prototypes. These tools are thus exploited to design a bank of dynamic neural networks for residual generation that are trained on the basis of the input and output measurements acquired from the simulator. The neural network bank is able to provide the detection of the faults affecting the actuators jointly with their isolation in case of simultaneous and concurrent faults The paper firstly describes the steps performed for deriving the proposed fault diagnosis solution. Secondly, the effectiveness of the scheme is demonstrated by means of high–fidelity simulations, in presence of faults and marine current.