Loading...

Discrimination and Identification of Multiple Partial Discharge Sources in a Transformer Insulation

Javandel Ajirloo, Vahid | 2017

727 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 50173 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Vakilian, Mahdi
  7. Abstract:
  8. Partial discharges that occur in a transformer insulation, generate current pulses. If these pulses be recorded, they can be used for transformer insulation condition assessment. Through processing of these recorded partial discharge signals, the PRPD patterns are generated and used to identify the source type of partial discharge defect. If multiple partial discharge defects exist in a transformer insulation, the related PRPD pattern, doesn’t look like any PRPD patterns of single defects. In this case, we need in the first step to discriminate the partial discharge signals stemmed from all the existing multiple partial discharge sources. To simulate the occurrence of multiple partial discharge defects in a transformer insulation, the PD defect models developed in the laboratory are put in parallel and the high voltage is applied. The generated partial discharge signals are captured. In the first step, the time, frequency and wavelet features of the signals are extracted, where the total number of extracted features from the signals added up to 35. To employ the most efficient part of this processed data, and to minimize the memory size needed for the following steps of the process, the six best features, among them, are selected through application of Laplacian score and correlation coefficient algorithms. In this thesis to discriminate the multiple partial discharge signals sources, the density based spatial clustering (DBSCAN), and the Gaussian mixture model (GMM) are used. The discriminated signals, as independent PD sources, are found more efficiently through DBSCAN, rather than GMM which is not able to discriminate the independent PD sources, accurately. However, if a specific method is able to determine the number of clusters accurately, then GMM will more accurately discriminate the multiple PD sources
  9. Keywords:
  10. Partial Discharge ; Feature Extraction ; Gaussian Mixture Modeling ; Phase Resolved Partial Discharge (PRPD)Pattern ; Density Based Method ; Laplacian Matrix ; Density-based Spatial Clustering Applicoction with Noise (DBSCAN)

 Digital Object List

 Bookmark

No TOC