期刊名称：Facta universitatis - series: Electronics and Energetics
印刷版ISSN：0353-3670
电子版ISSN：2217-5997
出版年度：2021
卷号：34
期号：1
页码：71-88
语种：English
出版社：University of Niš
摘要：The usage of wavelets is widespread in many fields nowadays, especially insignal processing. Their nature provides some advantages in comparison to the Fouriertransform, and therefore many applications rely on wavelets rather than on othermethods. The decomposition of wavelets into detail and approximation coefficients isone of the methods to extract representative audio features. They can be used in signalanalysis and further classification. This paper investigates the usage of various waveletfamilies in the wavelet decomposition to extract audio features of direct current (DC)motor sounds recorded in the production environment. The purpose of featurerepresentation and analysis is the detection of DC motor failures in motor production.The effects of applying different wavelet families and parameters in the decompositionprocess are studied using sounds of more than 60 motors. Time and frequency analysisis also done for the tested DC motor sounds.
关键词：wavelets; detail coefficients; approximation coefficients; audio features; DC motors
