Patterning of UHF partial discharge signals of various defects in GIS

摘要

In UHF PD detection technology, identification of PD signal is key point Exact classification of PD signals needs data base of PD pattern measured in UHF band. Various parameters such as amplitude, phase angle, and time interval of the UHF PD pulse train were used to visualize the acquired data. Several unique patterns were extracted for the defect types and these patterns will be used as a useful reference to analyze and identify various UHF PD patterns found in GIS substations. We will also discuss the possibilities of automated identification of various defect types using the visualization methods and available artificial neural network system.