Published Paper Details:

Reliability of power system could be very essential to generate and transmit energy. Power transformer is one of the most important electric apparatus and therefore it need to be saved in good state. The incipient fault identification and classification is a major research area. The dissolved gas analysis (DGA) is a technique being extensively used to find out incipient faults but diverse strategies had been developed to analyze DGA effects, they will once in a while fail to diagnose exactly. The accurate identification of incipient fault using numerous artificial intelligence (AI) is varied with variation of input parameters. Principle Component Analysis using Rapidminer Software is applied to IEC TC10 databases and associated datasets to find out most influencing input parameters for incipient fault classification. Further, one of the Machine learning algorithm namely Extreme Learning Machine (ELM) is implemented to categorize the power transformer incipient faults accurately.