Cable insulation assessment by a fuzzy-based algorithm

In this paper, a fuzzy-based clustering algorithm with features extracted from partial discharge waveform for condition assessment of power cables has been presented. Since the nature of cable insulation assessment already has remained unsupervised, fuzzy clustering algorithms seem to be more applicable in this context. The decision making process has been designed based on static features which is capable of considering trend of results. Also, the measuring and operation conditions have been considered to present final judgment. The results have shown successful clustering of the insulation condition.