Data mining on partial discharge signals of power transformer´s defect models

Partial discharge (PD) is a common phenomenon which occurs in insulation of high voltage equipments, such as; transformers and has a damaging effect on the insulation. If data mining techniques be used to find specifications and features of different types of partial discharges in power transformers, one can monitor the insulation condition of such equipment online and continuously. Those results can be employed to develop preventive measures more exactly and consequently the maintenance would require less time and cost for electric utility and improve the life time expectancy of the transformers. In this paper experiments are set up to create models for some types of PD that occurs in Power transformers, and features that can differentiate those PD types are extracted.