Roller bearing fault diagnosis using stacked denoising autoencoder in deep learning and Gath–Geva clustering algorithm without principal component analysis and data label

Most deep learning models such as stacked autoencoder (SAE) and stacked denoising autoencoder (SDAE) are used for fault diagnosis with a data label. These models are applied to extract the useful features with several hidden layers, then a classifier is used to complete the fault diagnosis. However,...

Celý popis

Uloženo v:

Podrobná bibliografie
Vydáno v:	Applied soft computing Ročník 73; s. 898 - 913
Hlavní autoři:	Xu, Fan, Tse, Wai tai Peter, Tse, Yiu Lun
Médium:	Journal Article
Jazyk:	angličtina
Vydáno:	Elsevier B.V 01.12.2018
Témata:	Deep learning Gath–Geva clustering algorithm Roller bearing fault diagnosis Stacked denoising autoencoder Deep learning Roller bearing fault diagnosis Gath–Geva clustering algorithm Stacked denoising autoencoder
ISSN:	1568-4946, 1872-9681
On-line přístup:	Získat plný text
Tagy:	Přidat tag Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!

Popis
Shrnutí:	Most deep learning models such as stacked autoencoder (SAE) and stacked denoising autoencoder (SDAE) are used for fault diagnosis with a data label. These models are applied to extract the useful features with several hidden layers, then a classifier is used to complete the fault diagnosis. However, these fault diagnosis classification methods are only suitable for tagged datasets. Actually, many datasets are untagged in practical engineering. The clustering method can classify data without a label. Therefore, a method based on the SDAE and Gath–Geva (GG) clustering algorithm for roller bearing fault diagnosis without a data label is proposed in this study. First, SDAE is selected to extract the useful feature and reduce the dimension of the vibration signal to two or three dimensions direct without principal component analysis (PCA) of the final hidden layer. Then GG is deployed to identify the different faults. To demonstrate that the feature extraction performance of the SDAE is better than that of the SAE and EEMD with the FE model, the PCA is selected to reduce the dimension of eigenvectors obtained from several previously hidden layers, except for the final hidden layer. Compared with SAE and ensemble empirical mode decomposition (EEMD)-fuzzy entropy (FE) models, the results show that as the number of the hidden layers increases, all the fault samples under different conditions are separated better by using SDAE rather than those feature extraction models mentioned. In addition, three evaluation indicators such as PC, CE, and classification accuracy are used to assess the performance of the method presented. Finally, the results show that the clustering effect of the method presented, and its classification accuracy are superior to those of the other combination models, including the SAE-fuzzy C-means (FCM)/Gustafson–Kessel (GK)/GG and EEMD-fuzzy entropy FE-PCA-FCM/GK/GG. •Reduce the dimension of extracted feature using SDAE directly without PCA.•Fulfilling the bearing fault diagnosis by using SDAE and clustering model without data label.•Using the SDAE reduce the high dimension to 2 or 3 directly from frequency domain feature directly after FFT decomposition.
ISSN:	1568-4946 1872-9681
DOI:	10.1016/j.asoc.2018.09.037