Fault diagnosis of Underwater Robots based on recurrent neural network

Research on thruster fault diagnosis of Underwater Robots (URs) is undertaken to improve its whole system reliability. Based on the BP neural network, a recurrent neural network (RNN) is presented and the network training algorithm is deduced. The RNN is trained by voyage head and yaw turning experiments, and the well trained network is applied to model for the URs. Compared the outputs between model and sensor, the residuals can be acquired; Fault diagnosis rules can be reached from the residuals to execute thruster fault detection. The methods proposed here are used for the simulation experiments and sea trials, and plenty of results are obtained. Based on the analysis of the experiment results, the validity and feasibility of the methods can be verified, and some guidance value in practical engineering applications can be demonstrated by the results.