Detail publikace

Neural Networks Applied to Real-Time Identification

Originální název

Neural Networks Applied to Real-Time Identification

Anglický název

Neural Networks Applied to Real-Time Identification

Jazyk

en

Originální abstrakt

Identification of dynamic systems is essential for adaptive control. Properly identified parameters of a dynamic system make it possible to design a suitable controller. One of the possible approaches is employing neural networks. The neural networks must be correctly train to reach our requirements. In this paper the on-line identification by using neural networks (with Levenberg-Marquardt training algorithm) is discussed.

Anglický abstrakt

Identification of dynamic systems is essential for adaptive control. Properly identified parameters of a dynamic system make it possible to design a suitable controller. One of the possible approaches is employing neural networks. The neural networks must be correctly train to reach our requirements. In this paper the on-line identification by using neural networks (with Levenberg-Marquardt training algorithm) is discussed.

BibTex


@inproceedings{BUT11767,
  author="Jiří {Dohnal} and Petr {Pivoňka}",
  title="Neural Networks Applied to Real-Time Identification",
  annote="Identification of dynamic systems is essential for adaptive control. Properly identified parameters of a dynamic system make it possible to design a suitable controller. One of the possible approaches is employing neural networks. The neural networks must be correctly train to reach our requirements. In this paper the on-line identification by using neural networks (with Levenberg-Marquardt training algorithm) is discussed.",
  address="DAAAM International Vienna",
  booktitle="Annals of DAAAM for 2004 & Proceedings",
  chapter="11767",
  institution="DAAAM International Vienna",
  year="2004",
  month="november",
  pages="95",
  publisher="DAAAM International Vienna",
  type="conference paper"
}