Self-triggered three-dimensional coordinated path following of disk-type autonomous underwater gliders based on low-frequency learning fuzzy predictors

Abstract

This paper is concerned with a self-triggered three-dimensional coordinated path following problem for a fleet of under-actuated disk-type autonomous underwater gliders (AUGs) subject to input constraints as well as internal model uncertainties and external ocean disturbances. A modular backstepping design approach is employed to devise three-dimensional fuzzy coordinated path following controllers for networked AUGs. Especially, a three-dimensional kinematic control law is firstly constructed by using a line-of-sight guidance scheme. Secondly, a predictor-based self-triggered communication mechanism is proposed to guarantee that each AUG broadcasts information at its own triggering instants, and to listen to and receive the incoming information from the neighboring AUGs at their triggering instants. Finally, an adaptive kinetic control law is designed based on low-frequency learning fuzzy predictors, which are able to extract the high-frequency components of uncertainties and disturbances. The closed-loop system is proven to be input-to-state stable via a cascade stability analysis. Simulation results validate the feasibility and efficacy of the proposed self-triggered three-dimensional coordinated path following controllers for the under-actuated disk-type AUGs subject to uncertainties and disturbances.