Initial-Rectification Neuroadaptive Finite-Time Surge Motion Tracking Control of Autonomous Underwater Vehicle With Input Saturation

Abstract

This paper proposes an initial-rectification neuroadaptive finite-time control scheme to solve the surge motion tracking control problem of an autonomous underwater vehicle with input saturation. Compared to the mass of autonomous underwater vehicles, the input gain of control system is very small, which makes the control system designed based on traditional control strategies prone to overload. To remedy this difficulty, initial rectification approach of reference trajectory and anti-windup design strategy are adopted meanwhile in this work. A finite-time adaptive neural control law is derived by using backstepping adaptive control approach, which employs an adaptive neural network to approximate uncertainties, and constructs two finite-time differentiators to estimate the derivatives of virtual control signals for lowering the design difficulty. Unlike many existing adaptive control results, the proposed adaptive neural controller guarantees that the initial-rectification tracking error converges to a small neighbour of the origin during a finite time. The control performance is verified through a comparative numerical simulation.