Detail publikačního výsledku

Adaptive Neuro-Fuzzy Inference System for Power Oil Transformer Diagnostics

JANDA, O.; HAMMER, M.; MINISTR, M.; ERTL, J.

Originální název

Adaptive Neuro-Fuzzy Inference System for Power Oil Transformer Diagnostics

Anglický název

Adaptive Neuro-Fuzzy Inference System for Power Oil Transformer Diagnostics

Druh

Článek recenzovaný mimo WoS a Scopus

Originální abstrakt

The adaptive neuro-fuzzy inference system for the diagnostics of power oil trans-formers is described in this paper. The system is based on the DGA methods where the ANFIS system is applied on the results. A new fuzzy system can be created or the original model of the diagnostic method according to the IEC 60599 standard can be adapted. The real data of gas concentrations was used for the learning process. The resulting accuracy of the optimized fuzzy model achieved up to 78%

Anglický abstrakt

The adaptive neuro-fuzzy inference system for the diagnostics of power oil trans-formers is described in this paper. The system is based on the DGA methods where the ANFIS system is applied on the results. A new fuzzy system can be created or the original model of the diagnostic method according to the IEC 60599 standard can be adapted. The real data of gas concentrations was used for the learning process. The resulting accuracy of the optimized fuzzy model achieved up to 78%

Klíčová slova

DGA, ANFIS, Fuzzy, IEC ratio, Transformer

Klíčová slova v angličtině

DGA, ANFIS, Fuzzy, IEC ratio, Transformer

Autoři

JANDA, O.; HAMMER, M.; MINISTR, M.; ERTL, J.

Rok RIV

2012

Vydáno

05.10.2011

Nakladatel

MM publishing

Místo

Praha

ISSN

1803-1269

Periodikum

MM Science Journal

Svazek

4

Číslo

3

Stát

Česká republika

Strany od

325

Strany do

331

Strany počet

7

BibTex

@article{BUT74336,
  author="Ondřej {Janda} and Miloš {Hammer} and Martin {Ministr} and Jakub {Ertl}",
  title="Adaptive Neuro-Fuzzy Inference System for Power Oil Transformer Diagnostics",
  journal="MM Science Journal",
  year="2011",
  volume="4",
  number="3",
  pages="325--331",
  issn="1803-1269"
}