Abstract

VERNA, Damián et al. Frequency domain Partial Discharge (PD) pulse parameters for the construction of classifications maps that allow the separation of PD sources in online measurements. Tecnol. cienc. [online]. 2022, n.45, pp.46-66.  Epub Dec 21, 2022. ISSN 1666-6933.  http://dx.doi.org/10.33414/rtyc.45.46-66.2022.

In this work, the frequency domain characteristics of partial discharge (PD) pulses are studied to construct classification maps that allow different sources of pulses to be separated. These maps are applied in online PD measurements, where multiple sources are present and through them, it is possible to separate the different sources and graph their corresponding subpattern to identify the source.

The characteristics that were analyzed in the frequency domain were: energy (Vef), maximum modulus (Xmax), maximum modulus frequency (fmax) and bandwidth (BW_3db). These parameters were calculated on the decomposition of each pulse using the Discrete Wavelet Transform (DWT). The algorithm was tested both with signals obtained from test specimens and on real equipment. The results showed that it is feasible to separate PD sources using the maximum modulus parameter (Xmax) for the frequency bands of detail 5 (7.8 to 15.6 MHz) and detail 4 (16.5 to 31.2 MHz).

Keywords : Partial Discharge; Wavelet transform; Classification map; frequency domain.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )