Publication detail

Possibility of Using Neural Network Based Identification in Adaptive Controllers

VAŇKOVÁ, M.

Original Title

Possibility of Using Neural Network Based Identification in Adaptive Controllers

Type

conference paper

Language

English

Original Abstract

This paper deals with possibility of a neural network identification applicable in adaptive (self-tuning) controllers. The reason is a long sample time necessary for a proper function of the classical identification methods (recursive least square especially) that results improper behaviour of the whole control system. Typical neural networks used to process modelling and applicable in this adaptive control are introduced. Regular identification requirements are discussed. Properties and special cases of the various neural types are introduced. The most extended neural networks are compared in conclusion.

Key words in English

adaptive controller, neural network, identification

Authors

VAŇKOVÁ, M.

RIV year

2005

Released

1. 1. 2005

Publisher

VUT v Brně, Fakulta elektrotechniky a komunikačních technologií a Fakulta informačních technologií

Location

Brno

ISBN

80-214-2889-9

Book

Proceedings of the 11th Conference STUDENT EEICT 2005

Pages from

66

Pages to

70

Pages count

5

BibTex

@inproceedings{BUT14574,
  author="Markéta {Šulová}",
  title="Possibility of Using Neural Network Based Identification in Adaptive Controllers",
  booktitle="Proceedings of the 11th Conference STUDENT EEICT 2005",
  year="2005",
  pages="5",
  publisher="VUT v Brně, Fakulta elektrotechniky a komunikačních technologií a Fakulta informačních technologií",
  address="Brno",
  isbn="80-214-2889-9"
}