A Simplified Fuzzy Wavelet Neural Control for Nonlinear Systems With Quantized Inputs and Deferred Constraints
Xiaohui Yue, Huaguang Zhang, Jiayue Sun, Xin Liu
Abstract
This article investigates a finite-time fuzzy quantized control problem for a class of nonlinear systems considering deferred constraints. Instead of the tracking errors themselves, the auxiliary error variables constructed via the shifting function are employed into nonlogarithm barrier Lyapunov function to perform error constraints, not only making the restrictive conditions in initial phase be removed but also ensuring tracking errors to evolve within the preassigned regions after a given time. Then, to allow for a reduced computational cost concerning fuzzy/neural approximators, a single parameter updating based fuzzy wavelet neural network is devised to approximate the unknown nonlinearity acting on every subsystem. Furthermore, by using hysteresis quantizer to convert continuous control inputs into discrete scalars, a robust fuzzy quantized controller is synthesized with the aid of a novel quantization decomposition scheme, where the problem of constrained data bandwidth is successfully handled without involving chattering in control signals. Finally, simulations confirm the benefits and efficiency of the proposed method.