Classification of lightning stroke on transmission line using multi-resolution analysis and machine learning

J. A. Morales, E. Orduña, C. Rehtanz

Research output: Contribution to journalArticlepeer-review

21 Scopus citations

Abstract

One of most important elements of Electric Power Systems (EPS) is the transmission line (TL), which is permanently under adverse conditions especially lightning strokes. At the moment, those phenomena have been the root cause of short circuits and the most important cause of mal-operation of transmission line protection relays. Thus, this paper develops the classification of lightning transient signals with and without fault. Multi-resolution analysis (MRA) is used to analyze those signals considering five mother wavelets and different decomposition levels of three phase voltages. In this manner, Spectral Energy and Machine Learning as Artificial Neural Network, K-Nearest Neighbors and Support Vector Machine are employed to classify those signals. On the other hand, the developed work in this paper analyzes most important parameters of lightning strokes, which are essentials in producing conditions with and without fault. Results show that the methodology presents an acceptable performance. © 2013 Elsevier Ltd. All rights reserved.
Original languageEnglish
Pages (from-to)19-31
Number of pages13
JournalInternational Journal of Electrical Power and Energy Systems
DOIs
StatePublished - 1 Jun 2014
Externally publishedYes

Fingerprint

Dive into the research topics of 'Classification of lightning stroke on transmission line using multi-resolution analysis and machine learning'. Together they form a unique fingerprint.

Cite this