Aplicação da lógica fuzzy nos resultados das análises de óleo isolante dos transformadores da UHE de Tucuruí
This work aims to develop an intelligent predictive maintenance system to detect incipiente faults in power transformers from the cromatographic analysis of developed gases in the insulating oil of these devices using the IEC and Rogers methods.The proposed system is developed based on artificial in...
| Autor principal: | FERNANDES, Keyliane da Silva |
|---|---|
| Grau: | Trabalho de Conclusão de Curso - Graduação |
| Publicado em: |
2018
|
| Assuntos: | |
| Acesso em linha: |
http://bdm.ufpa.br/jspui/handle/prefix/490 |
| Resumo: |
|---|
|
This work aims to develop an intelligent predictive maintenance system to detect incipiente faults in power transformers from the cromatographic analysis of developed gases in the insulating oil of these devices using the IEC and Rogers methods.The proposed system is developed based on artificial intelligence techniques using the ratio of dissolved gases in oil.It will be presented the basic structure of the fuzzy system developed based on IEC and Rogers methods, since this is the artificial intelligence technique used. To validate the proposed methodology were used historical data from trials of Tucuruí’s Hydropower Plant transformers, conducted by the Laboratory Lacen. The results were very favorable, especially when the methods of Rogers and IEC were combined, thus supplying individual deficiencies and reaching levels of match with values greater than 80%. |