Neural network based sliding mode controller for a class of linear systems with unmatched uncertainties

Author

Baric, Miroslav ; Petrovic, Ivan ; Peri, N.

Author_Institution

Fac. of Electr. Eng., Zagreb Univ., Croatia

Volume

1

fYear

2002

fDate

10-13 Dec. 2002

Firstpage

967

Abstract

This paper considers the application of a neural network for the performance improvement of the sliding mode controller for a class of linear systems with unmatched uncertainties/disturbances. A neural network is employed for the online estimation of the uncertainties using the simple gradient descent learning algorithm. The combination of the sliding mode and backstepping-like recursive control design is used to achieve the desired tracking performance. The algorithm is verified through computer simulations.

Keywords

gradient methods; learning (artificial intelligence); linear systems; multidimensional systems; neurocontrollers; tracking; uncertain systems; variable structure systems; backstepping like recursive control; gradient descent learning algorithm; linear systems; multidimensional systems; neural network; neurocontrol; sliding mode controller; tracking; uncertain systems; Backstepping; Control design; Control systems; Electromechanical systems; Linear systems; Neural networks; Robust control; Sliding mode control; State-space methods; Uncertainty;

fLanguage

English

Publisher

ieee

Conference_Titel

Decision and Control, 2002, Proceedings of the 41st IEEE Conference on

ISSN

0191-2216

Print_ISBN

0-7803-7516-5

Type

conf

DOI

10.1109/CDC.2002.1184634

Filename

1184634