IET GENERATION TRANSMISSION & DISTRIBUTION, vol.14, no.21, pp.4859-4865, 2020 (Peer-Reviewed Journal)
Article / Article
IET GENERATION TRANSMISSION & DISTRIBUTION
Science Citation Index Expanded, Scopus, Academic Search Premier, PASCAL, Aerospace Database, Applied Science & Technology Source, Business Source Elite, Business Source Premier, Communication Abstracts, Compendex, Computer & Applied Sciences, Environment Index, INSPEC, Metadex, Civil Engineering Abstracts
feature extraction, principal component analysis, time-frequency analysis, partial discharge measurement, monopole antennas, polymer insulators, conical antennas, packaging, wavelet transforms, image colour analysis, insulator contamination, radiotelemetry, insulator surface, simultaneous PD signals, PD sources, polymeric insulators, RF PD measurement, power grid, grid reliability, housing-erosion, power network reliability, contaminated insulators, polymeric housing, contamination layer, partial discharge activities, RF-PRPD subpatterns, radiofrequency-PRPD patterns, conical monopole antenna, time-frequency matrix, greyscale image, wavelet transform, histogram of oriented gradients method, principal component analysis, Davies-Bouldin criterion, k-medoids clustering method, FLASHOVER VOLTAGE, CORONA
A better performance and consequently the widespread use of polymeric insulators in different parts of the power grid can increase their role in the grid reliability. The accumulation of contamination and housing-erosion are the two most effective factors in undermining the performance of this type of insulators. Therefore, electric utilities need to identify contaminated insulators for washing and cracks in polymeric housing to replace them with healthy specimens. This paper discusses the impact of contamination layer and housing-erosion of polymeric insulators on the partial discharges (PD) at the insulator surface, through RF-PRPD (phase resolved partial discharge) patterns. The existence of different sources of PDs in a real environment (transmission line or station) makes it difficult to use the PRPD patterns to distinguish them from each other. Therefore, using a conical monopole antenna, the simultaneous PD signals and the related RF-PRPD pattern of samples under test are captured. The grayscale image was obtained using the time-frequency matrix of the PD signals transform, by wavelet. Then, features are extracted and selected from grayscale image. By clustering of the PD signals, the resulted RF-PRPD sub-patterns are well separated and provided the necessary means to distinguish among the status of different samples under test.