Abstract

In this paper, an expert system for transformer fault diagnosis was developed. This expert system utilizes Dissolve Gas in oil Analysis (DGA) techniques to diagnose fault conditions of transformers. Several methods were implemented in this expert system, namely the fuzzy Roger’s four ratio method; NTT Flagpoint method and Generation Rate ratio method. This is to increase the accuracy of both methods, as fuzzy logic helps to manage the uncertainties exhibited by transformer DGA data. All three methods then were integrated into this expert system, to yield greater accuracy in analyzing a transformer fault condition.