Depth-keeping control for autonomous underwater vehicle in vertical plane using adaptive output feedback controller

To keep the depth of an autonomous underwater vehicle(AUV) in a vertical plane,we design an adaptive output feedback controller based on the backstepping technique,using the information of measurable depth and the pitch angle.To achieve this objective,an observer is built to implement the feedback of the immeasurable velocity of the pitch angle.A radial basis neural network is adopted to compensate the nonlinear effects produced by the uncertain hydrodynamic coefficients as well as the coupling velocities of surge,dive,and pitch angle.An adaptive strategy is used to inhibit the disturbance from the velocity of surge and dive.A first-order incomplete model of AUV is employed without linearization;this relaxes the restriction on the variation range of the pitch angle.Theoretical analysis and simulation experiment results show that the proposed method implements the control of depth and altitude of the AUV with high adaptability and strong robustness for the unordered nonlinear dynamics and bounded disturbances.