Finite-time self-structuring neural network trajectory tracking control of underactuated autonomous underwater vehicles

Abstract

A finite-time self-structuring neural network (SSNN) trajectory tracking control scheme is proposed for an input-constrained underactuated autonomous underwater vehicle (AUV) with unknown external disturbances. First, a time-varying barrier Lyapunov function is proposed to constrain and prevent excessive consistency errors and reduce computational effort. Second, a finite-time controller is presented to achieve finite-time convergence of the closed-loop system, while dynamic surface control (DSC) is used to reduce the computational complexity of the system. Third, a finite-time SSNN method is developed to obtain the optimal number of neurons with simple computation and better approximation to estimate uncertain disturbances. Moreover, the Lyapunov stability proof indicates that all signals in the closed-loop system are ultimately bounded uniformly with respect to the initial constraints, and the trajectory tracking errors can converge to near zero in a finite time. Finally, simulations not only evaluate the performance of the proposed controller-controlled system but also verify the effectiveness of the methodology in this paper.