Open AccessPower quality detection and classification using wavelet and support vector machine
Author(s) -
Victor Garrido-Arévalo,
Walter Gil-González,
Mauricio Holguín
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1448/1/012002
Subject(s) - harmonics , voltage sag , flicker , power quality , pattern recognition (psychology) , computer science , support vector machine , wavelet , wavelet transform , coding (social sciences) , artificial intelligence , energy (signal processing) , speech recognition , voltage , mathematics , engineering , statistics , electrical engineering , operating system
This work presents the identification and classification of various disturbances that affect the quality of energy, seen as the quality of the voltage wave (harmonics, sag, swell and flicker). For this, the wavelet transform is used, which allows to have characteristic patterns as input signals of the support vector machine, these are evaluated in their different configurations, bi-class, minimum output coding, error correcting output and one versus all. For all of them, in the first instance they were trained with 200 samples, then the results were validated with 100 samples and finally the evaluation was made with 500 different samples, obtaining that the best result is presented with the minimum output coding configuration.