Skip to main content
SpringerLink
Log in

Bearing condition monitoring via an unsupervised and enhanced stacked auto-encoder

  • Technical Paper
  • Published:
Journal of the Brazilian Society of Mechanical Sciences and Engineering Aims and scope Submit manuscript

Abstract

Supervised deep learning models have been widely used in the construction of bearing health indicators (HIs) for performance degradation. Such models require a series of model improvement combinations, which are dependent on manual experience, to extract HIs. Additionally, noise leads to a large, extracted health indicator value that easily exceeds the preset threshold, causing an incorrect evaluation of health status. Therefore, this paper proposes an improved and unsupervised stacked auto-encoder without an output label layer and with local weighted regression and smoothing scatterplot (Lowess). That is, a stacked auto-encoder adds a Lowess filter after each hidden layer to perform multiple de-noising operations. We used encoders and decoders at several hidden layers within a stacked noise auto-encoder (though only with an output layer) to extract HIs and directly eliminate the noise from the original bearing vibration. Hence, our proposed method reduces the modeling complexity. The results show that our proposed method outperforms others of its type, such as stacked auto-encoders, stacked de-noising auto-encoders, deep belief neural networks, root-mean-square kurtosis, empirical mode decomposition, singular-value decomposition-K-medoids, and self-organizing map networks. Moreover, by using the monotonicity evaluating indicator, the extracted HI curve was found to be smoother than in other models and the probability of erroneous judgments for bearing performance degradation assessment was reduced.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Log in via an institution

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6
Fig. 7
Fig. 8
Fig. 9

Similar content being viewed by others

References

  1. Heng A, Zhang S, Tan AC (2009) Rotating machinery prognostics: state of the art, challenges and opportunities. Mech Syst Signal Process 23(3):724–739

    Article  Google Scholar 

  2. Huang Y, Wu BX, Wang JQ (2012) Test for active control of boom vibration of a concrete pump truck. J Vib Shock. 31(2):91–94

    Google Scholar 

  3. Resta F, Ripamonti F (2011) Cazzluani G (2011) Independent modal control for nonlinear flexible structures: an experimental test rig. J Sound Vib 329(8):961–972

    Article  Google Scholar 

  4. Bagordo G, Cazzluani G, Resta F (2011) A modal disturbance estimator for vibration suppression in nonlinear flexible structures. J Sound Vib 330(25):6061–6069

    Article  Google Scholar 

  5. Theodoros HL, Dimitrios R, George G (2013) Remaining useful life estimation in rolling bearings utilizing data-driven probabilistic e-support vectors regression. IEEE Trans Reliab 62(4):821–832

    Article  Google Scholar 

  6. Williams T, Ribadeneira X, Billington S, Kurfess T (2001) Rolling element bearing diagnostics in run-to-failure lifetime testing. Mech Syst Signal Process 15(5):979–993

    Article  Google Scholar 

  7. Antoni J (2006) The spectral kurtosis: a useful tool for characterising nonstationary signals. Mech Syst Signal Process 20:282–307

    Article  Google Scholar 

  8. Antoni J, Randall RB (2006) The spectral kurtosis: Application to the vibratory surveillance and diagnostics of rotating machines. Mech Syst Signal Process 20:308–331

    Article  Google Scholar 

  9. Rai A, Upadhyay SH (2017) Bearing performance degradation assessment based on a combination of empirical mode decomposition and K-medoids clustering. Mech Syst Signal Process 93:16–29

    Article  Google Scholar 

  10. Qiu H, Lee J, Lin J, Yu G (2003) Robust performance degradation assessment methods for enhanced rolling element bearing prognostics. Adv Eng Inform 17:127–140

    Article  Google Scholar 

  11. Tse PW, Wang D (2017) Enhancing the abilities in assessing slurry pumps’ performance degradation and estimating their remaining useful lives by using captured vibration signals. J Vib Control 23(12):1925–1937

    Article  Google Scholar 

  12. Pan Y, Chen J, Li XL (2010) Bearing performance degradation assessment based on lifting wavelet packet decomposition and fuzzy c-means. Mech Syst Signal Process 24(5):559–566

    Article  Google Scholar 

  13. Rai A, Upadhyay SH (2018) Intelligent bearing performance degradation assessment and remaining useful life prediction based on self-organising map and support vector regression. Proc Inst Mech Eng Part C J Mech Eng Sci 232(6):1118–1132

    Article  Google Scholar 

  14. Zhang Y, Tang BP, Han Y (2017) Bearing performance degradation assessment based on time-frequency code features and SOM network. Meas Sci Technol 28(4):045601

    Article  Google Scholar 

  15. Sun S, Zhang BB, Xie L, Zhang Y (2017) An unsupervised deep domainadaptation approach for robust speech recognition. Neurocomputing 257:79–87

    Article  Google Scholar 

  16. Affonso C, Rossi ALD, Vieira FHA, Carvalho ACPLF (2017) Deep learning for biological image classification, Expert Syst. Appl 85:114–122

    Google Scholar 

  17. Lv FY, Wen CL, Liu MQ (2017) Weighted time series fault diagnosis based on a stacked sparse autoencoder. J Chemometrics 31(9):1–16

    Article  Google Scholar 

  18. Qi YM, Shen CQ, Wang D (2017) Stacked sparse autoencoder-based deep networkfor fault diagnosis of rotating machinery. IEEE Access 5:15066–15079

    Article  Google Scholar 

  19. Wang LK, Zhao XY, Pei JG (2016) Transformer fault diagnosis using continuous sparse autoencoder. Springerplus 5:1–13

    Google Scholar 

  20. Feng J, Lei YG, Jing L (2016) Deep neural networks: A promising tool for fault characteristic mining and intelligent diagnosis of rotating machinery with massive data. Mech Syst Signal Process 72–73:303–315

    Google Scholar 

  21. Vincent P, Larochelle H, Bengio Y (2008) Extracting and composing robust features with denoising autoencoders, Proceedings of the 25th international conference on machine learning international conference. 1096–1103.

  22. Zhao X, Wu J, Zhang Shi YY, Wang L (2020) Fault diagnosis of motor in frequency domain signal by stacked de-noising auto-encoder. CMC Comp Mater Contin 57(2):223–242

    Google Scholar 

  23. Chen LR (2020) A novel method of combining nonlinear frequency spectrum and deep learning for complex system fault diagnosis. Measurement 151:107190

    Article  Google Scholar 

  24. Xuejiao P, Shaojiang D, Xuewu P et al (2023) A method for rolling bearing life state recognition by combining health indicator and anti-noise deep residual shrinkage network[J]. J Braz Soc Mech Sci Eng 45(1):37

    Article  Google Scholar 

  25. Mao W, Chen J, Chen Y et al (2021) Construction of health indicators for rotating machinery using deep transfer learning with multiscale feature representation[J]. IEEE Trans Instrum Meas 70:1–13

    Google Scholar 

  26. Wang Z, Guo J, Wang J, Yang Y, Dai L, Huang CG, Wan JL (2023) A deep learning based health indicator construction and fault prognosis with uncertainty quantification for rolling bearings. Meas Sci Technol 34(10):105105

    Article  Google Scholar 

  27. Shi M, Ding C, Que H et al (2023) Multilayer-graph-embedded extreme learning machine for performance degradation prognosis of bearing[J]. Measurement 207:112299

    Article  Google Scholar 

  28. Chen X, Wang Y, Meng L et al (2023) An interpretable health indicator for bearing condition monitoring based on semi-supervised autoencoder latent space variance maximization[J]. Meas Sci Technol 34(12):125135

    Article  Google Scholar 

  29. Chen Z, Zhu H, Wu J et al (2022) Health indicator construction for degradation assessment by embedded LSTM–CNN​ autoencoder and growing self-organized map[J]. Knowl-Based Syst 252:109399

    Article  Google Scholar 

  30. Xu F, Tse WP, Tse YL (2018) Roller bearing fault diagnosis using stacked denoising autoencoder in deep learning and Gath-Geva clustering algorithm without principal component analysis and data label. Appl Soft Comput 73:898–913

    Article  Google Scholar 

  31. Xu F, Huang ZL, Yang FF, Wang D, Tsui KL (2020) Constructing a health indicator for roller bearings by using a stacked auto-encoder with an exponential function to eliminate concussion. Appl Soft Comput 89:106119

    Article  Google Scholar 

  32. Xu XuF, Yang FF, Fan XM, Huang ZL, Tsui KL (2020) Extracting degradation trends for roller bearings by using a moving-average stacked auto-encoder and a novel exponential function. Measurement 152:107371

    Article  Google Scholar 

  33. Zhu KH (2018) Performance degradation assessment of rolling element bearingsbased on hierarchical entropy and general distance. J Vib Control 24(14):3194–3205

    Article  Google Scholar 

  34. Qian Y, Yan RQ, Hu SJ (2014) Bearing degradation evaluation using recurrence quantification analysis and kalman filter. IEEE Trans Instrum Meas 63(11):2599–2610

    Article  Google Scholar 

  35. Javed K, Gouriveau R, Zerhouni N, Nectoux P (2015) Enabling health monitoring approach based on vibration data for accurate prognostics. IEEE Trans Industr Electron 62(1):647–656

    Article  Google Scholar 

  36. Nectoux P, Gouriveau R, Medjaher K (2012) PRONOSTIA: an experimental platform for bearings accelerated life test, IEEE international conference on prognostics and health management, Denver, CO, USA.

  37. Lee J, Qiu H, Yu G, Lin J (2007) Bearing data set, IMS, University of Cincinnati. NASA ames prognostics data repository. Rexnord technical services.

Download references

Acknowledgements

This paper supported by The National Natural Science Foundation of China (NO.52205168).

Funding

National Natural Science Foundation of China, 52205168, Fan Xu

Author information

Authors and Affiliations

  1. School of Mechanical and Electronic Engineering, Wuhan University of Technology, Wuhan, 430070, Hubei, China

    Fan Xu & Zhenyu Hao

  2. Institute of Laser and Photontic Engineering, CASIC, Hubei, 430075, Wuhan, China

    Chao Zhou

  3. School of Management, Wuhan University of Technology, Wuhan, 430070, Hubei, China

    Yaling Deng

  4. Research Institute of Digital Governance and Management Decision Innovation, Wuhan University of Technology, Wuhan, 430070, Hubei, China

    Yaling Deng

Authors
  1. Fan Xu
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Zhenyu Hao
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Chao Zhou
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Yaling Deng
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Yaling Deng.

Ethics declarations

Conflict of interest

We declare that we have no financial and personal relationships with other people or organizations that can inappropriately influence our work. There is no professional or other personal interest of any nature or kind in any product, service, and/or company that could be construed as influencing the position presented in, or the review of, the manuscript entitled.

Additional information

Technical Editor: Jarir Mahfoud.

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Xu, F., Hao, Z., Zhou, C. et al. Bearing condition monitoring via an unsupervised and enhanced stacked auto-encoder. J Braz. Soc. Mech. Sci. Eng. 46, 367 (2024). https://doi.org/10.1007/s40430-024-04866-2

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1007/s40430-024-04866-2

Keywords

Search

Navigation