Fault detection and classification in transmission lines based on a PSD index. Issue 18 (4th September 2018)

• Record Type: Journal Article
• Authors: Guillen, Daniel; Paternina, Mario R. Arrieta; Ortiz‐Bejar, Jose; Tripathy, Rajesh Kumar; Zamora‐Mendez, Alejandro; Tapia‐Olvera, Ruben; Tellez, Eric S.
• Is Part Of: IET generation, transmission & distribution. Volume 12:Issue 18(2018)
• Page Start: 4070
• Page End: 4078
• Publication Date: 2018-09-04
• Subjects: power transmission lines -- power transmission faults -- fault diagnosis -- signal classification -- covariance matrices -- approximation theory -- discrete wavelet transforms -- power grids -- power system interconnection; fault detection -- fault classification -- AC transmission lines -- PSD index -- power spectral density index -- transient information analysis -- discrete wavelet transform -- current signal scaling -- decomposition levels -- approximation coefficient -- detail coefficient -- wavelet‐covariance matrix -- Hellinger distance -- radial power grid -- interconnected power grid; Electric power production -- Periodicals; Electric power transmission -- Periodicals; Electric power distribution -- Periodicals; 621.3105
• DOI: 10.1049/iet-gtd.2018.5062
• Languages: English
• ISSNs: 1751-8687

Abstract

Abstract : This study deals with the faults' detection and classification in AC transmission lines based on power spectral density (PSD), introducing PSD in time and frequency for analysing transient information under faulted conditions. The discrete wavelet transform is used for scaling current signals in time–frequency at different decomposition levels by approximation and detail coefficients. Then, a wavelet‐covariance matrix is shaped with the aim to obtain its PSD, being this the key to detect and classifying faults. Results show that the proposed method attains the detection in a short time and the classification is accomplished via the Hellinger distance, whose straightforward implementation is carried out in this study. The classification process is compared adopting different classifiers to cope with a set of signals in a time frame. Finally, the proposal is extensively assessed using real and simulated signals stemming from multiple fault cases of radial and interconnected power grids.