Adaptive trajectory tracking for underactuated AUVs in time-varying currents: An RBFNN observer-based backstepping approach

This paper addresses the planar trajectory tracking problem of underactuated autonomous underwater vehicles (AUVs) in the presence of time-varying ocean currents, model uncertainties, and external disturbances. In contrast to many existing studies, the effects of ocean currents are explicitly incorporated into both the kinematic and dynamic models. To estimate the unknown current-induced perturbations online, an ocean current observer based on a radial basis function neural network (RBFNN) is developed. On this basis, a backstepping trajectory-tracking controller is constructed, in which a novel virtual-control design is introduced and the yaw rate is employed as a virtual control variable to avoid the singularity issue encountered in conventional designs. Lyapunov-based analysis is carried out to establish the boundedness of all closed-loop signals and the convergence of the tracking errors to a small neighborhood of the origin. Numerical examples show that the proposed method achieves improved tracking accuracy and faster disturbance-estimation performance than benchmark schemes under time-varying ocean-current conditions.