A Hybrid Tool for Detection of Incipient Faults in Transformers Based on the Dissolved Gas Analysis of Insulating Oil

Abstract

This paper describes the development and implementation of a tool for the diagnosis of faults in power transformers through the analysis of dissolved gases in oil. The computational system approach is based on the combined use of some traditional criteria of the dissolved gas analysis published in standards, an artificial neural network, and a fuzzy logic system. The objective of the tool is to provide the user with an answer obtained from analysis not only of the traditional methods already consolidated in the technical literature, but also via artificial-intelligence techniques, reaching a higher degree of reliability with respect to each technique individually. The results obtained with this tool are promising in the diagnosis of incipient faults in transformers, reaching success levels of more than 80%.