Distributed coordinated tracking of multiple autonomous underwater vehicles

This paper considers the distributed coordinated tracking problem of multiple autonomous underwater vehicles with a time-varying reference trajectory. Each vehicle is subject to model uncertainty and time-varying ocean disturbances. A novel predictor-based neural dynamic surface control design approach is proposed to develop the node controllers, under which synchronization between vehicles can be reached on condition that the augmented graph induced by the vehicles and the reference trajectory contains a spanning tree. The prediction errors are used to update the neural adaptive laws, which enable fast identifying the vehicle dynamics without excessive knowledge of their dynamical models. Further, this result is extended to the output-feedback case where only position-yaw information can be measured. A local predictor, based on its own position-yaw information, is constructed, not only to recover the unmeasured velocity information, but also to identify the unknown dynamics for each vehicle. A linear matrix inequality-based analysis is performed for the stability of the predictor. Then, distributed output-feedback tracking controllers are developed to achieve synchronization between vehicles in the presence of unknown dynamics and unmeasured velocities. For both cases, the stability properties of the closed-loop network are established via Lyapunov analysis. Simulation results demonstrate the effectiveness of the proposed methods.