DRFNN-adaptive output feedback controller for depth tracking of AUV

This paper addresses the problem of autonomous underwater vehicle (AUV) depth control in the absence of full state information. Only position message permitted as input vector of controller designing process. An observer based on dynamic recurrent fuzzy neural network (DRFNN) is designed to estimate the other states of diving dynamics, where the DRFNN is adopted to evaluate the dynamic complex nonlinear part which is caused by the hard accurate estimation of the hydrodynamic coefficients and the nonlinear structure in the pitch motion of an AUV. An adaptive output feedback controller is designed based on the developed observer using the observer backstepping technique. The proposed control scheme can guarantee that all the signals in the closed-loop system satisfy to be uniformly ultimately bounded. Simulation studies illustrate the effectiveness of the proposed control scheme.