Translation Invariance-Based Deep Learning for Rotating Machinery Diagnosis
Discriminative feature extraction is a challenge for data-driven fault diagnosis. Although deep learning algorithms can automatically learn a good set of features without manual intervention, the lack of domain knowledge greatly limits the performance improvement, especially for nonstationary and no...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2020-01-01
|
| Series: | Shock and Vibration |
| Online Access: | http://dx.doi.org/10.1155/2020/1635621 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832566323852345344 |
|---|---|
| author | Wenliao Du Shuangyuan Wang Xiaoyun Gong Hongchao Wang Xingyan Yao Michael Pecht |
| author_facet | Wenliao Du Shuangyuan Wang Xiaoyun Gong Hongchao Wang Xingyan Yao Michael Pecht |
| author_sort | Wenliao Du |
| collection | DOAJ |
| description | Discriminative feature extraction is a challenge for data-driven fault diagnosis. Although deep learning algorithms can automatically learn a good set of features without manual intervention, the lack of domain knowledge greatly limits the performance improvement, especially for nonstationary and nonlinear signals. This paper develops a multiscale information fusion-based stacked sparse autoencoder fault diagnosis method. The autoencoder takes advantage of the multiscale normalized frequency spectrum information obtained by dual-tree complex wavelet transform as input. Accordingly, the multiscale normalized features guarantee the translational invariance for signal characteristics, and the stacked sparse autoencoder benefits the unsupervised feature learning and ensures accurate and stable diagnosis performance. The developed method is performed on motor bearing vibration signals and worm gearbox vibration signals, respectively. The results confirm that the developed method can accommodate changing working conditions, be free of manual feature extraction, and perform better than the existing intelligent diagnosis methods. |
| format | Article |
| id | doaj-art-1e3142ac9f4f45ce9da7d92848f6ab78 |
| institution | Kabale University |
| issn | 1070-9622 1875-9203 |
| language | English |
| publishDate | 2020-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Shock and Vibration |
| spelling | doaj-art-1e3142ac9f4f45ce9da7d92848f6ab782025-02-03T01:04:22ZengWileyShock and Vibration1070-96221875-92032020-01-01202010.1155/2020/16356211635621Translation Invariance-Based Deep Learning for Rotating Machinery DiagnosisWenliao Du0Shuangyuan Wang1Xiaoyun Gong2Hongchao Wang3Xingyan Yao4Michael Pecht5Henan Provincial Key Laboratory of Intelligent Manufacturing of Mechanical Equipment, Zhengzhou University of Light Industry, 5 Dongfeng Road, Zhengzhou 450002, ChinaSchool of Mechanical Engineering, University of Shanghai for Science and Technology, Shanghai 200093, ChinaHenan Provincial Key Laboratory of Intelligent Manufacturing of Mechanical Equipment, Zhengzhou University of Light Industry, 5 Dongfeng Road, Zhengzhou 450002, ChinaHenan Provincial Key Laboratory of Intelligent Manufacturing of Mechanical Equipment, Zhengzhou University of Light Industry, 5 Dongfeng Road, Zhengzhou 450002, ChinaSchool of Computer Science and Information Engineering, Chongqing Technology and Business University, Chongqing 400067, ChinaCenter for Advanced Life Cycle Engineering (CALCE), University of Maryland, College Park, MD 20742, USADiscriminative feature extraction is a challenge for data-driven fault diagnosis. Although deep learning algorithms can automatically learn a good set of features without manual intervention, the lack of domain knowledge greatly limits the performance improvement, especially for nonstationary and nonlinear signals. This paper develops a multiscale information fusion-based stacked sparse autoencoder fault diagnosis method. The autoencoder takes advantage of the multiscale normalized frequency spectrum information obtained by dual-tree complex wavelet transform as input. Accordingly, the multiscale normalized features guarantee the translational invariance for signal characteristics, and the stacked sparse autoencoder benefits the unsupervised feature learning and ensures accurate and stable diagnosis performance. The developed method is performed on motor bearing vibration signals and worm gearbox vibration signals, respectively. The results confirm that the developed method can accommodate changing working conditions, be free of manual feature extraction, and perform better than the existing intelligent diagnosis methods.http://dx.doi.org/10.1155/2020/1635621 |
| spellingShingle | Wenliao Du Shuangyuan Wang Xiaoyun Gong Hongchao Wang Xingyan Yao Michael Pecht Translation Invariance-Based Deep Learning for Rotating Machinery Diagnosis Shock and Vibration |
| title | Translation Invariance-Based Deep Learning for Rotating Machinery Diagnosis |
| title_full | Translation Invariance-Based Deep Learning for Rotating Machinery Diagnosis |
| title_fullStr | Translation Invariance-Based Deep Learning for Rotating Machinery Diagnosis |
| title_full_unstemmed | Translation Invariance-Based Deep Learning for Rotating Machinery Diagnosis |
| title_short | Translation Invariance-Based Deep Learning for Rotating Machinery Diagnosis |
| title_sort | translation invariance based deep learning for rotating machinery diagnosis |
| url | http://dx.doi.org/10.1155/2020/1635621 |
| work_keys_str_mv | AT wenliaodu translationinvariancebaseddeeplearningforrotatingmachinerydiagnosis AT shuangyuanwang translationinvariancebaseddeeplearningforrotatingmachinerydiagnosis AT xiaoyungong translationinvariancebaseddeeplearningforrotatingmachinerydiagnosis AT hongchaowang translationinvariancebaseddeeplearningforrotatingmachinerydiagnosis AT xingyanyao translationinvariancebaseddeeplearningforrotatingmachinerydiagnosis AT michaelpecht translationinvariancebaseddeeplearningforrotatingmachinerydiagnosis |