Nuclear Engineering and Technology

  • 제55권2호
  • /
  • Pages.506-515
  • /
  • 2023
  • /
  • 1738-5733(pISSN)
  • /
  • 2234-358X(eISSN)
한국원자력학회 (Korean Nuclear Society)
권호 표지
DOI QR코드

DOI QR Code

Proposal of a new method for learning of diesel generator sounds and detecting abnormal sounds using an unsupervised deep learning algorithm

  • 투고 : 2022.06.12
  • 심사 : 2022.10.15
  • 발행 : 2023.02.25
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

This study is to find a method to learn engine sound after the start-up of a diesel generator installed in nuclear power plant with an unsupervised deep learning algorithm (CNN autoencoder) and a new method to predict the failure of a diesel generator using it. In order to learn the sound of a diesel generator with a deep learning algorithm, sound data recorded before and after the start-up of two diesel generators was used. The sound data of 20 min and 2 h were cut into 7 s, and the split sound was converted into a spectrogram image. 1200 and 7200 spectrogram images were created from sound data of 20 min and 2 h, respectively. Using two different deep learning algorithms (CNN autoencoder and binary classification), it was investigated whether the diesel generator post-start sounds were learned as normal. It was possible to accurately determine the post-start sounds as normal and the pre-start sounds as abnormal. It was also confirmed that the deep learning algorithm could detect the virtual abnormal sounds created by mixing the unusual sounds with the post-start sounds. This study showed that the unsupervised anomaly detection algorithm has a good accuracy increased about 3% with comparing to the binary classification algorithm.

키워드

과제정보

This work was supported by the Korea Institute of Energy Technology Evaluation and Planning [KETEP Project No. 20204010600130] and the Ministry of Trade, Industry & Energy (MOTIE) of the Republic of Korea.

참고문헌

  1. K.H. Choi, S.G. Lee, A study on the power plant application of engine condition diagnosis technology for diesel generator, J. Korean Soc. Power Syst. Eng. 17 (2013) 17-22. https://doi.org/10.9726/kspse.2013.17.4.017
  2. S.G. Lee, Development status of emergency diesel generator status diagnosis technology for nuclear power plants, J. Electr. World (2014) 82-88. Special And+.
  3. S.G. Lee, D.W. Kim, Development and study of engine condition diagnosis program for emergency diesel generator for power generation, J. Korean Soc. Power Syst. Eng. 19 (2015) 67-72. https://doi.org/10.9726/kspse.2015.19.5.067
  4. K.W. Kang, K.M. Lee, CNN-Based automatic machine Fault Diagnosis method using spectrogram images, J. Instit. Converg. Signal Process 21 (2020) 121-126.
  5. S.H. Lee, J.S. Kim, B.B. Shin, CNN based actuator fault cause classification system using noise, in: Proceedings of the Winter Conference of the Korean Society of Computer and Information Engineers 26, 2018, pp. 7-8.
  6. T. Tran, J. Lundgren, Drill fault diagnosis based on the scalogram and mel spectrogram of sound signals using artificial intelligence, IEEE Access 8 (2020) 203655-203666, https://doi.org/10.1109/ACCESS.2020.3036769.
  7. H. Liu, L. Li, J. Ma, Rolling Bearing Fault Diagnosis Based on STFT-Deep Learning and Sound Signals, Hindawi Publishing Corporation Shock and Vibration, 2016, 6127479, https://doi.org/10.1155/2016/6127479, 2016.
  8. K.M. Lee, A Machine Fault Diagnosis System Based on Convolutional Neural Network Using Acoustic Power Spectral Density, Pukyong National University, 2019. https://repository.pknu.ac.kr:8443/handle/2021.oak/23346.
  9. H. Nakamura, K. Asano, S. Usuda, Y. Mizuno, A diagnosis method of bearing and stator fault in motor using rotating sound based on deep learning, Energies 14 (2021) 1319, https://doi.org/10.3390/en14051319.
  10. C. Vununu, A Sound Diagnosis System with Deep Learning for Machine Fault Detection, Pukyong National University, 2017. https://repository.pknu.ac.kr:8443/handle/2021.oak/14281.
  11. D. Li, B. Li, C. Wang, P. Cheng, B. Jiao, A fault diagnosis equipment of motor bearing based on sound signal and CNN, J. Phys.: Conf. Ser. (2010), 012159, https://doi.org/10.1088/1742-6596/2010/1/012159.
  12. S.H. Lee, B.B. Sin, Y.J. Kim, J.S. Kim, Deep-learning based PHM embedded system using noise.vibration, in: Proceedings of the Summer Conference of the Korean Society of Computer and Information Engineers 25, 2017, pp. 9-10.
  13. S. Zhao, E. Wang, J. Hao, Fault diagnosis method for energy storage mechanism of high voltage circuit breaker based on CNN characteristic matrix constructed by sound-vibration signal, J. Vibroeng. 21 (2019) 1665-1678, https://doi.org/10.21595/jve.2019.20781.
  14. Y. Yao, H. Wang, S. Li, Z. Liu, G. Gui, Y. Dan, J. Hu, End-to-end convolutional neural network model for gear fault diagnosis based on sound signals, Appl. Sci. 8 (9) (2018) 1584, https://doi.org/10.3390/app8091584.
  15. L. Weng, From Autoencoder to Beta-VAE, GITHUB, 2018. https://lilianweng.github.io/lil-log/2018/08/12/from-autoencoder-to-beta-vae.html.
  16. S. Jin, Introduction to the spectrum spectrogram, J. Kor. Soc. Laryngol. Phoniatr. Logop. 19 (2008) 101-106. http://jkslp.org/journal/view.php?number=1878.
  17. S.A. Broughton, K. Bryan, Discrete Fourier Analysis and Wavelets: Applications to Signal and Image Processing, second ed., Wiley, 2018, ISBN 978-1-119-25824-7.
  18. F. Chollet, Deep Learning with Python, Manning, 2017, ISBN 9781617294433.