Inferring partial discharge identification through fuzzy tools

This paper presents a new identification technique, based on fuzzy logic, which enables the recognition of PD generated from different sources, such as internal, surface and corona discharges. It is shown that the different sources typologies can be identified on the basis of fuzzy rules applied to a selection of parameters derived from PD-pulse phase and amplitude distributions. The proposed procedure is finally applied to rotating machines and cables, with real insulation defects, showing promising on-field applications.