Detail publikace

The Semi-Inversion Controller

PIVOŇKA, P., KRUPANSKÝ, P., VELEBA, V., DOHNAL, J.

Originální název

The Semi-Inversion Controller

Anglický název

The Semi-Inversion Controller

Jazyk

en

Originální abstrakt

When the basic approaches to control using neural networks had been studied, a new control algorithm - Semi-inversion controller was designed. Due to the high degree of freedom of neural networks in the form of circuits with a universal neural model, the adaptation behavior of the control circuit is difficult to control. The principles based on inversion control result in high control action. The semi-inversion controller was designed to tackle these problems, to exploit the high adaptability of neural networks and make use of inversion control principles.

Anglický abstrakt

When the basic approaches to control using neural networks had been studied, a new control algorithm - Semi-inversion controller was designed. Due to the high degree of freedom of neural networks in the form of circuits with a universal neural model, the adaptation behavior of the control circuit is difficult to control. The principles based on inversion control result in high control action. The semi-inversion controller was designed to tackle these problems, to exploit the high adaptability of neural networks and make use of inversion control principles.

Dokumenty

BibTex 


@inproceedings{BUT15249,
  author="Petr {Pivoňka} and Petr {Krupanský} and Václav {Veleba} and Jiří {Dohnal}",
  title="The Semi-Inversion Controller",
  annote="When the basic approaches to control using neural networks had been studied, a new control algorithm - Semi-inversion controller was designed. Due to the high degree of freedom of neural networks in the form of circuits with a universal neural model, the adaptation behavior of the control circuit is difficult to control. The principles based on inversion control result in high control action. The semi-inversion controller was designed to tackle these problems, to exploit the high adaptability of neural networks and make use of inversion control principles.",
  address="FME Zenica, Bosna a Hercegovina",
  booktitle="TMT 2005",
  chapter="15249",
  institution="FME Zenica, Bosna a Hercegovina",
  year="2005",
  month="september",
  pages="631",
  publisher="FME Zenica, Bosna a Hercegovina",
  type="conference paper"
}