The modeling of partial discharge waveforms in power systems equipment

This paper looks at the modeling of partial discharge (PD) waveforms to gain an understanding of the characteristics of PD signals. To accurately model the PD waveforms, a high sampling rate of 500 MHz was used in collecting the PD data with the use of a high-frequency current transformer. While characteristics such as rise time, fall time and pulse width have been used in the past to analyze PD waveforms, this paper uses Prony's method to fit a digital transfer function to the PD waveform, such that the impulse response of the digital filter approximately matches the waveform. Results show that this technique can be applied to internal, corona and surface partial discharge signals, and a model like this can be used for simulating and interpreting PD waveforms.