Adaptive NNs‐based 3D trajectory tracking control by state error PCH method for AUVs

ABSTRACT This article focuses on the trajectory tracking control (TTC) of autonomous underwater vehicles (AUVs) operating in three‐dimensional (3D) underwater environments with concurrent ocean disturbances, parametric uncertainties and actuator saturation constraints. A tracking control architecture integrating the state‐error port‐controlled Hamiltonian (SEPCH) energy shaping approach with radial basis function neural networks (RBFNNs) is developed to guarantee the uniform ultimate boundedness (UUB) of closed‐loop dynamics. Specifically, the Hamiltonian energy‐shaping framework is systematically synthesized to establish the intrinsic stability property, while adaptive RBFNNs compensators are designed to estimate the compounded perturbations arising from the ocean current disturbances and parameter uncertainties effects. Furthermore, the challenge caused by input saturation is tackled by means of an auxiliary system by confining the control force to certain predefined saturation bounds. Simulation examples demonstrate the validity of the proposed control method.