Bearing Fault Diagnosis Based on Improved Morlet Wavelet Transform and Shallow Residual Neural Network

Autor:	Lijin Guo, Bintao Han, Qilan Huang
Jazyk:	angličtina
Rok vydání:	2024
Předmět:	wavelet admissibility condition improved Morlet wavelet wavelet residual network Technology Engineering (General). Civil engineering (General) TA1-2040 Biology (General) QH301-705.5 Physics QC1-999 Chemistry QD1-999
Zdroj:	Applied Sciences, Vol 14, Iss 11, p 4542 (2024)
Druh dokumentu:	article
ISSN:	14114542 2076-3417
DOI:	10.3390/app14114542
Popis:	To solve the problem of the Morlet wavelet transform not satisfying wavelet admissibility conditions, which results in fuzzy and aliasing time–frequency image features, this paper proposes an improved Morlet wavelet. The scale parameter is the wavelet’s dominant frequency parameter. It largely satisfies the admissibility conditions and effectively improves the fuzzy aliasing effect of time–frequency image features. To further improve feature learning, this paper proposes a combined Morlet wavelet and residual network. The experimental results show that the accuracy of the proposed method is 10%, 4.5%, and 3% higher than that of LeNets, CNNs, and ResNets under varying load and noise disturbance levels, respectively.
Databáze:	Directory of Open Access Journals
Externí odkaz:	https://doaj.org/article/641aa5e2c9ae487fa696a4ef1966df58 Zobrazit plný text záznamu View record in DOAJ