Estimation of Charge, Energy and Polarity of Noisy Partial Discharge Pulses

Research output: Contribution to journalArticleScientificpeer-review

21 Citations (Scopus)

Abstract

The algorithms for the computation of charge, energy and polarity of partial discharge (PD) pulses are affected by noise, which can lead to over and underestimation of the PD quantities. These quantities can be computed in time domain, frequency domain and according to the impulse response method (standard IEC270). In this paper, a theoretical study is performed in which each computation method is applied to simulated PD pulses having different waveforms and noise level to evaluate the extent of affectation on the results due to the noise. The results suggest that the error in the estimation of the PD charge is higher for oscillatory pulses regardless of the method. In contrast, the estimation of energy is more resilient to the PD waveform and the estimation in frequency domain gives rise to low error. In time domain, the charge an energy estimation method can be improved by filtering the PD pulse and integrating
the output pulse to certain limits. A new method for the estimation of PD polarity is proposed based on the derivative of the filtered pulse, showing accurate estimation of the polarity even for the pulses with low signal to noise ratio.
Original languageEnglish
Pages (from-to)2511-2521
Number of pages11
JournalIEEE Transactions on Dielectrics and Electrical Insulation
Volume24
Issue number4
DOIs
Publication statusPublished - 2017

Keywords

  • Estimation
  • Frequency-domain analysis
  • Noise measurement
  • Partial discharges
  • Pulse measurements
  • Sensors
  • Transmission line measurements
  • Partial discharge
  • charge
  • energy
  • pulse filtering
  • quasi-integration

Fingerprint

Dive into the research topics of 'Estimation of Charge, Energy and Polarity of Noisy Partial Discharge Pulses'. Together they form a unique fingerprint.

Cite this