Observer-based neuro-adaptive prescribed performance control of nonstrict feedback systems and its application

Abstract

This paper investigates an observer-based neuro-adaptive prescribed performance control scheme for nonstrict feedback systems with consideration of unmeasurable states and expected output tracking performance constraint. Firstly, a novel simplified barrier Lyapunov function involving the prescribed performance function is proposed. Then, via combing the backstepping technique and neural network based approximation, an adaptive prescribed performance controller is developed along with a neural network based observer. To conquer the problem “explosion of complexity” inherent in backstepping technique, dynamic surface control technique is employed, wherein, an adaptive nonlinear filter is devised to compensate for the boundary layer errors. Compared with the existing works, the logarithmic function-based transformation for the prescribed performance constraint is avoided, which renders the developed scheme computationally simple. Finally, the application to a one-link manipulator is organized to verify the effectiveness of the proposed control scheme.