Use of Time-Frequency map combined with DBSCAN algorithm for separation of partial discharge pulses under DC voltage

Abstract

The Phase-Resolved-Partial-Discharge pattern (PRPD) is a conventional technique used for the evaluation of partial discharges (PD) phenomena in High-Voltage-Alternating-Current (HVAC) systems. This map is constructed by plotting the peak of each detected pulses as a function of the phase angle of the supply voltage. Therefore it is obvious that this technique cannot be used for the analysis of data from PD mesaurement under different supply voltage condition (DC). The aim of this paper is to evaluate the application of the Time-Frequency map (TF map) for the analysis of a dataset obtained from PD measurement under DC voltage. A density-based clustering algorithm was also used to gain more insight from the collected data. The results show that, with this approach, it's possible to perform a noise rejection and identify PD pulses.