Coordinated pattern tracking of multiple marine surface vehicles with uncertain kinematics and kinetics

This paper considers the coordinated pattern tracking of multiple marine surface vehicles in the presence of uncertain kinematics and kinetics. Distributed pattern tracking controllers depending on the information of neighboring vehicles are derived based on a backstepping technique, neural networks and an identifier. Specifically, the identifier is devised to precisely estimate the time-varying ocean currents at the kinematic level. Neural networks together with adaptive filtering methods are employed to extract the low frequency content of the model uncertainty and ocean disturbances at the kinetic level. The benefit of the proposed design results in adaptive pattern tracking controllers over any undirected connected graphs with guaranteed low frequency control signals, which facilitates practical implementations. The stability properties of the multi-vehicle systems are established via Lyapunov analysis, and the pattern tracking errors converge to an adjustable neighborhood of origin. An example is given to show the performance of the proposed approach.