Fuzzy Decision on Transformer Fault Diagnosis using Dissolved Gas Analysis and IEC Ratio Codes

This paper emphasizes on the use of fuzzy approach in dealing with the incipient fault conditions of power transformer. DGA (Dissolved Gas in Oil Analysis) cannot provide better fault diagnosis results when multiple faults are involved. The boundary values specified by the ratio methods are of limited concern .Considering the limitations of convergence of the conventional neural networks in the local minima, fuzzy logic may take care of the uncertainties in fault conditions. Fuzzy decision is used to deal with various incipient fault conditions including the normal condition with major faults as per the IEC Ratio codes. The comparative of the fuzzy system, conventional IEC method of fault diagnosis and Roger’s ratio diagnosis of some oil samples of transformer and data from research paper are presented. Rule based fuzzy decision on various faults is made using IEC Ratio codes.