ABSTRACT Power transformers are the most valuable assets in the power system. An appropriate asset management is therefore essential for ensuring a reliable power supply to the utility managers. This is accomplished through continuous transformer condition monitoring. The transformers’ asset management involves a severe investigation into its insulation health index. It is determined by utilising several diagnostic parameters of transformer insulation. In this context, a novel fuzzy logic-based health index model has been proposed in the present paper to find the overall health condition of transformers. This model utilises the results extracted from several important diagnostic tests of transformer insulation, such as furan analysis and dissolved gas analysis. The validity of the proposed fuzzy logic model has been tested using real data from 200 transformer samples. These samples are collected from different-rated transformers of the Himachal Pradesh State Electricity Board (HPSEB), India. Further, the efficacy of the proposed fuzzy logic model has been validated against the expert model presented in literature. It is revealed from the comparison that the accuracy of the present proposed model is 92%. Also, it is observed that the proposed fuzzy logic model is very effective in interpreting dissolved gases and the other diagnostic test results of the transformer.
Read full abstract