Self-Constructing Adaptive Robust Fuzzy Neural Tracking Control of Surface Vehicles With Uncertainties and Unknown Disturbances
In this paper, a novel self-constructing adaptive robust fuzzy neural control (SARFNC) scheme for tracking surface vehicles, whereby a self-constructing fuzzy neural network (SCFNN) is employed to approximate system uncertainties and unknown disturbances, is proposed. The salient features of the SAR...
Saved in:
Published in | IEEE transactions on control systems technology Vol. 23; no. 3; pp. 991 - 1002 |
---|---|
Main Authors | , |
Format | Journal Article |
Language | English |
Published |
IEEE
01.05.2015
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Summary: | In this paper, a novel self-constructing adaptive robust fuzzy neural control (SARFNC) scheme for tracking surface vehicles, whereby a self-constructing fuzzy neural network (SCFNN) is employed to approximate system uncertainties and unknown disturbances, is proposed. The salient features of the SARFNC scheme are as follows: 1) unlike the predefined-structure approaches, the SCFNN is able to online self-construct dynamic-structure fuzzy neural approximator by generating and pruning fuzzy rules, and achieve accurate approximation; 2) an adaptive approximation-based controller (AAC) is designed by combining sliding-mode control with SCFNN approximation using improved projection-based adaptive laws, which avoid parameter drift and singularity in membership functions simultaneously; 3) to compensate for approximation errors, a robust supervisory controller (RSC) is presented to enhance the robustness of the overall SARFNC control system; and 4) the SARFNC consisting of AAC and RSC can achieve an excellent tracking performance, whereby tracking errors and their first derivatives are globally uniformly ultimately bounded. Simulation studies and comprehensive comparisons with traditional adaptive control schemes demonstrate remarkable performance and superiority of the SARFNC scheme in terms of tracking errors and online approximation. |
---|---|
ISSN: | 1063-6536 1558-0865 |
DOI: | 10.1109/TCST.2014.2359880 |