Dynamic periodic event-triggered adaptive PD-like composite neural control with iterative disturbance observer for 3D tracking of UUVs

This work focuses on the 3D tracking control issue for underactuated underwater vehicles (UUVs) subject to model uncertainties, external disturbances, and actuator mechanical degradation by mechanical wear. To reconstruct the uncertain dynamics within the system, an adaptive PD-like composite neural network (NN) is constructed. In contrast with conventional adaptive NNs, the constructed PD-like composite NN accelerates the network learning process and improves the reconstruction accuracy of uncertain dynamics. Furthermore, to counteract the adverse effects of external disturbances, an NN disturbance observer is designed with a gain iteration protocol. Unlike conventional disturbance observers that employ fixed gains, the designed observer dynamically adjusts its gains through an iterative protocol to achieve accurate disturbance estimation and enhance the robustness of the UUV system. To mitigate actuator wear and avoid unnecessary control updates, a dynamic periodic event-triggered protocol is proposed, which inherently precludes Zeno behavior. Building on these developments, a novel adaptive PD-like composite neural control law that integrates the aforementioned iterative disturbance observer and the dynamic periodic event-triggered protocol is formulated. The Lyapunov theory is employed to demonstrate that all signals within the closed-loop control system are bounded. Simulation results substantiate the effectiveness and feasibility of the proposed control strategy.