On-line diagnosis of incipient faults and cellulose degradation based on artificial intelligence methods

In this paper, a new artificial intelligence technique is proposed to detect incipient faults and cellulose degradation in power transformers using dissolved gas analysis. The proposed technique is based on a combination between neural networks and fuzzy logic theory. Incipient faults diagnosis is based on hydrocarbon gases as an input while cellulose degradation detection is based on carbon monoxide and carbon dioxide. The capabilities of the proposed diagnostic system have been verified through practical test data collected from the Egyptian electricity network. A comparison between the proposed technique and reported methods is carried out.