A novel fuzzy logic approach to transformer fault diagnosis

Dissolved gas in oil analysis is an well established in-service technique for incipient fault detection in oil-insulated power transformers. A great deal of experience and data in dissolved gas in oil analysis (DGA) is now available within the utilities. Actually, diagnostic interpretations were solely done by human experts using past knowledge and standard techniques such as the ratio method. In this paper, a novel fuzzy logic approach is adopted to develop a computer based intelligent interpretation of transformer faults using Visual basic and C/sup ++/ programming. The proposed fuzzy logic based software as been tested and tuned using over 800 dissolved gas in oil analysis (DGA) case histories. This highly reliable tool has then been utilized in detection and verification of 20 transformer faults. The proposed diagnostic tool is very useful to both expert and novice engineers in DGA result interpretation.