Journal of Drive and Control (드라이브 ㆍ 컨트롤)

The Korean Society for Fluid Power and Construction Equipment (유공압건설기계학회)
권호 표지
DOI QR코드

DOI QR Code

Reliable Fault Diagnosis Method Based on An Optimized Deep Belief Network for Gearbox

  • Oybek Eraliev (Department of Control & Mechanical Engineering, the Graduate School, Hankook University) ;
  • Ozodbek Xakimov (Department of Mechanical Engineering, Jeju National University) ;
  • Chul-Hee Lee (Department of Control & Mechanical Engineering, the Graduate School, Hankook University)
  • Received : 2023.10.18
  • Accepted : 2023.11.16
  • Published : 2023.12.01
Download PDF
⟨ Previous Next ⟩

Abstract

High and intermittent loading cycles induce fatigue damage to transmission components, resulting in premature gearbox failure. To identify gearbox defects, numerous vibration-based diagnostics techniques, using several artificial intelligence (AI) algorithms, have recently been presented. In this paper, an optimized deep belief network (DBN) model for gearbox problem diagnosis was designed based on time-frequency visual pattern identification. To optimize the hyperparameters of the model, a particle swarm optimization (PSO) approach was integrated into the DBN. The proposed model was tested on two gearbox datasets: a wind turbine gearbox and an experimental gearbox. The optimized DBN model demonstrated strong and robust performance in classification accuracy. In addition, the accuracy of the generated datasets was compared using traditional ML and DL algorithms. Furthermore, the proposed model was evaluated on different partitions of the dataset. The results showed that, even with a small amount of sample data, the optimized DBN model achieved high accuracy in diagnosis.

Keywords

Acknowledgement

This research was supported by a grant(2023-MOIS35-005) of Policy-linked Technology Development Program on Natural Disaster Prevention and Mitigation funded by Ministry of Interior and Safety (MOIS, Korea).

References

  1. F. Elasha, M. Greaves, and D. Mba, "Planetary bearing defect detection in a commercial helicopter main gearbox with vibration and acoustic emission," Struct. Heal. Monit., vol. 17, no. 5, pp. 1192-1212, 2018, doi: 10.1177/1475921717738713.
  2. S. Butzmann and J. Melbert, "Sensorless Control of Electromagnetic Actuators for Variable Train," SAE Paper 2000-01-1225, pp. 325-423, 2000.
  3. Y. Wang, S. Yang, and R. V. Sanchez, "Gearbox fault diagnosis based on a novel hybrid feature reduction method," IEEE Access, vol. 6, pp. 75813-75823, 2018, doi: 10.1109/ACCESS.2018.2882801.
  4. Y. B. Lee, G. C. Lee, J. J. Lee, S. Y. Lim, W. J. Kim, and K. M. Kim, " A Study on the Acceleration Durability Test of In-Wheel Drive Gearbox for Military Special Vehicles," vol. 19, no. 3, pp. 32-38, 2022. 
  5. Z. Chen, C. Li, and R. Sanchez, "Gearbox Fault Identification and Classification with Convolutional Neural Networks," Shock Vib., vol. 2015, pp. 1-10, 2015, doi: 10.1155/2015/390134.
  6. C. Li and M. Liang, "Extraction of oil debris signature using integral enhanced empirical mode decomposition and correlated reconstruction," Meas. Sci. Technol., vol. 22, no. 8, 2011, doi: 10.1088/0957-0233/22/8/085701.
  7. A. M. D. Younus and B. S. Yang, "Intelligent fault diagnosis of rotating machinery using infrared thermal image," Expert Syst. Appl., vol. 39, no. 2, pp. 2082-2091, 2012, doi: 10.1016/j.eswa.2011.08.004.
  8. T. Toutountzakis, C. K. Tan, and D. Mba, "Application of acoustic emission to seeded gear fault detection," NDT E Int., vol. 38, no. 1, pp. 27-36, 2005, doi: 10.1016/j.ndteint.2004.06.008.
  9. J. R. Ottewill and M. Orkisz, "Condition monitoring of gearboxes using synchronously averaged electric motor signals," Mech. Syst. Signal Process., vol. 38, no. 2, pp. 482-498, 2013, doi: 10.1016/j.ymssp.2013.01.008.
  10. J. W. Park, S. Han, H. S. Lee, and S. Yun, "A Study on the Hydraulic Vibration Characteristics of the Prefill Check Valve-monitoring," vol. 18, no. 3, pp. 8-15, 2021.
  11. H. S. Baek, J. H. Shin, and S. J. Kim, "Development of AI-Based Condition Monitoring System for Failure Diagnosis of Excavator ' s Travel Device," vol. 18, no. 1, pp. 24-30, 2021.
  12. Y. K. Kang and J. S. Jang, "Feasibility Study on the Vibration Reduction for Hydraulic Breaker by the Dynamic Vibration Absorber," vol. 18, no. 4, pp. 65-71, 2021.
  13. S. Khan and T. Yairi, "A review on the application of deep learning in system health management," Mech. Syst. Signal Process., vol. 107, pp. 241-265, 2018, doi: 10.1016/j.ymssp.2017.11.024.
  14. T. Wang, Q. Han, F. Chu, and Z. Feng, "Vibration based condition monitoring and fault diagnosis of wind turbine planetary gearbox: A review," Mech. Syst. Signal Process., vol. 126, pp. 662-685, 2019, doi: 10.1016/j.ymssp.2019.02.051.
  15. S. R. Saufi, Z. A. Bin Ahmad, M. S. Leong, and M. H. Lim, "Low-Speed Bearing Fault Diagnosis Based on ArSSAE Model Using Acoustic Emission and Vibration Signals," IEEE Access, vol. 7, pp. 46885-46897, 2019, doi: 10.1109/ACCESS.2019.2909756.
  16. J. Xu, L. Xiang, R. Hang, and J. Wu, "Stacked Sparse Autoencoder (SSAE) based framework for nuclei patch classification on breast cancer histopathology," 2014 IEEE 11th Int. Symp. Biomed. Imaging, ISBI 2014, no. 61273259, pp. 999-1002, 2014, doi: 10.1109/isbi.2014.6868041.
  17. L. Deng, G. Hinton, and B. Kingsbury, "New types of deep neural network learning for speech recognition and related applications: An overview," ICASSP, IEEE Int. Conf. Acoust. Speech Signal Process. - Proc., pp. 8599-8603, 2013, doi: 10.1109/ICASSP.2013.6639344.
  18. R. Zhao, R. Yan, Z. Chen, K. Mao, P. Wang, and R. X. Gao, "Deep learning and its applications to machine health monitoring," Mech. Syst. Signal Process., vol. 115, pp. 213-237, 2019, doi: 10.1016/j.ymssp.2018.05.050.
  19. Y. Guo, Y. Liu, A. Oerlemans, S. Lao, S. Wu, and M. S. Lew, "Deep learning for visual understanding: A review," Neurocomputing, vol. 187, pp. 27-48, 2016, doi: 10.1016/j.neucom.2015.09.116.
  20. W. Zhang, C. Li, G. Peng, Y. Chen, and Z. Zhang, "A deep convolutional neural network with new training methods for bearing fault diagnosis under noisy environment and different working load," Mech. Syst. Signal Process., vol. 100, pp. 439-453, 2018, doi: 10.1016/j.ymssp.2017.06.022.
  21. M. Sadoughi and C. Hu, "Physics-Based Convolutional Neural Network for Fault Diagnosis of Rolling Element Bearings," IEEE Sens. J., vol. 19, no. 11, pp. 4181-4192, 2019, doi: 10.1109/JSEN.2019.2898634.
  22. B. Ibrokhimov, C. Hur, H. Kim, and S. Kang, "ADBNF: adaptive deep belief network framework for regression and classification tasks," Appl. Intell., vol. 51, no. 7, pp. 4199-4213, 2021, doi: 10.1007/s10489-020-02050-2.
  23. H. Chen, J. Wang, B. Tang, K. Xiao, and J. Li, "An integrated approach to planetary gearbox fault diagnosis using deep belief networks," Meas. Sci. Technol., vol. 28, no. 2, 2017, doi: 10.1088/1361-6501/aa50e7.
  24. H. Zheng and Y. Dai, "Fault Prediction of Fan Gearbox Based on Deep Belief Network," J. Phys. Conf. Ser., vol. 1449, no. 1, 2020, doi: 10.1088/1742-6596/1449/1/012050.
  25. J. Yu and G. Liu, "Knowledge extraction and insertion to deep belief network for gearbox fault diagnosis," Knowledge-Based Syst., vol. 197, p. 105883, 2020, doi: 10.1016/j.knosys.2020.105883.
  26. Y. Qin, X. Wang, and J. Zou, "The Optimized Deep Belief Networks with Improved Logistic Sigmoid Units and Their Application in Fault Diagnosis for Planetary Gearboxes of Wind Turbines," IEEE Trans. Ind. Electron., vol. 66, no. 5, pp. 3814-3824, 2019, doi: 10.1109/TIE.2018.2856205.
  27. D. Verstraete, A. Ferrada, E. L. Droguett, V. Meruane, and M. Modarres, "Deep Learning Enabled Fault Diagnosis Using Time-Frequency Image Analysis of Rolling Element Bearings," Shock Vib., vol. 2017, pp. 1-17, 2017, doi: 10.1155/2017/5067651.
  28. G. H.-N. computation and undefined 2002, "Training products of experts by minimizing contrastive divergence," ieeexplore.ieee.org, Accessed: Jun. 06, 2022. [Online]. Available: https://ieeexplore.ieee.org/abstract/document/6789337/
  29. M. N. Ab Wahab, S. Nefti-Meziani, and A. Atyabi, "A comprehensive review of swarm optimization algorithms," PLoS One, vol. 10, no. 5, pp. 1-36, 2015, doi: 10.1371/journal.pone.0122827.
  30. J. Y. Lee, "Variable short-time Fourier transform for vibration signals with transients," JVC/Journal Vib. Control, vol. 21, no. 7, pp. 1383-1397, 2015, doi: 10.1177/1077546313499389.
  31. O. Eraliev, K.-H. Lee, and C.-H. Lee, "VibrationBased Loosening Detection of a Multi-Bolt Structure Using Machine Learning Algorithms," Sensors, vol. 22, no. 3, p. 1210, 2022, doi: 10.3390/s22031210.
  32. J. Huang, B. Chen, B. Yao, and W. He, "ECG Arrhythmia Classification Using STFT-Based Spectrogram and Convolutional Neural Network," IEEE Access, vol. 7, pp. 92871-92880, 2019, doi: 10.1109/ACCESS.2019.2928017.
  33. "Encyclopedia of Vibration | ScienceDirect." https://www.sciencedirect.com/referencework/9780122270857/encyclopedia-of-vibration (accessed Jan. 20, 2022).
  34. S. Shao, S. McAleer, R. Yan, and P. Baldi, "Highly Accurate Machine Fault Diagnosis Using Deep Transfer Learning," IEEE Trans. Ind. Informatics, vol. 15, no. 4, pp. 2446-2455, Apr. 2019, doi: 10.1109/TII.2018.2864759.
  35. S. R. Saufi, Z. A. Bin Ahmad, M. S. Leong, and M. H. Lim, "Gearbox Fault Diagnosis Using a Deep Learning Model With Limited Data Sample," IEEE Trans. Ind. Informatics, vol. 16, no. 10, pp. 6263-6271, Oct. 2020, doi: 10.1109/TII.2020.2967822.