Adaptive Finite-Time Neural Network Control for Non-strict Feedback Systems

Abstract

In this paper, an adaptive finite-time neural network tracking control problem for uncertain non-strict feedback systems is studied. For unknown nonlinear functions, they are approximated using neural networks. Under the framework of adaptive backstepping, a finite-time tracking controller based on a non-strict feedback system is designed. Unlike existing finite-time results, the proposed method can guarantee that the output of the system tracks the reference signal in a shorter time, and further, the tracking error is guaranteed to be confined to a small origin domain, while all signals in the closed-loop system are bounded and fast practical finite-time stablility. Finally, simulation example is given to exhibit the effectiveness of the presented technique.