Utilization of Phase Resolved Partial Discharge (PRPD) Characteristics using K-NN Classifier for Defect Identification in GIS
Partial discharge (PD) is a major cause of insulation failure in high voltage electrical equipments. The main cause of such partial discharge in GIS is free/fixed particle contamination, insulator defect, surface roughness, and/or triple junction, etc. These partial discharges may lead to complete breakdown of the electrical equipment. The adequate knowledge of this activity i.e. occurrence, type and mode of partial discharge, etc. due to any type of defect helps in its prevention and effective preventive management of GIS equipment. The aim of this paper is to carry out a systematic study for the defect identification in GIS equipments, PRPD characteristics due to the above mentioned specific type of defect(s), Phase Resolved Partial Discharge (PRPD) were obtained for each type of defect, using Transient Earth Voltage (TEV) sensor. The obtained PRPD patterns were used to identify the type of defect using K-Nearest Neighborhood (K-NN) algorithm. The results will help in development of proper maintenance procedures before the failure of electrical equipment occurs and will therefore improve the reliability of electric power system. Keywords- High voltage, Partial discharge, PRPD, TEV, K-NN.