An efficient PD data mining method for power transformer defect models using SOM technique

• Laboratory PD measurement set-up is developed.
• HFCT is employed to pick up PD current pulses.
• Texture feature extraction method is applied on time-domain PD data.
• Capability in different PD sources discrimination (using texture features) is assessed by SOM.

Suggestion and application of a set of new features for on-line Partial Discharge (PD) monitoring, where there is no information about the type of PD is a challenging task for condition assessment of power equipments, such as a power transformer. This is looked for in this paper. So far, in past various techniques have been employed to develop a comprehensive PD monitoring system, however limited success has been achieved. One of the challenging issues in this field is the discovering of proper features capable of differentiating the involvement of possible types of PD sources. In order to examine the efficiency of the method established in this paper, which is based on application of a set of new feature spaces, texture feature analysis, followed by application of principal component analysis (PCA) and self-organizing map (SOM) is used to analyze and interpret the time-domain-captured PD data. The results of this work demonstrate the capabilities of the aforementioned features space to be used as a supplementary knowledge-base to help experts making their decisions confidently.