Journal of Korea Society of Industrial Information Systems (한국산업정보학회논문지)

Korea Society of Industrial Information Systems (한국산업정보학회)
권호 표지
DOI QR코드

DOI QR Code

Comparative Study of AI Models for Reliability Function Estimation in NPP Digital I&C System Failure Prediction

원전 디지털 I&C 계통 고장예측을 위한 신뢰도 함수 추정 인공지능 모델 비교연구

  • 이대영 ((주)미래와도전 디지털기획부) ;
  • 이정훈 ((주)미래와도전 사업개발그룹) ;
  • 양승혁 ((주)미래와도전 디지털기획부)
  • Received : 2023.10.30
  • Accepted : 2023.12.02
  • Published : 2023.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

The nuclear power plant(NPP)'s Instrumentation and Control(I&C) system periodically conducts integrity checks for the maintenance of self-diagnostic function during normal operation. Additionally, it performs functionality and performance checks during planned preventive maintenance periods. However, there is a need for technological development to diagnose failures and prevent accidents in advance. In this paper, we studied methods for estimating the reliability function by utilizing environmental data and self-diagnostic data of the I&C equipment. To obtain failure data, we assumed probability distributions for component features of the I&C equipment and generated virtual failure data. Using this failure data, we estimated the reliability function using representative artificial intelligence(AI) models used in survival analysis(DeepSurve, DeepHit). And we also estimated the reliability function through the Cox regression model of the traditional semi-parametric method. We confirmed the feasibility through the residual lifetime calculations based on environmental and diagnostic data.

원전 계측제어계통은 정상운전 시 자가 진단기능의 유지보수를 위해 일정 주기로 건전성을 확인하고 있으며, 계획예방정비 기간 동안 기능 및 성능점검을 실시하여 필요한 경우 유지보수를 하고 있다. 하지만 원전의 정보를 계측하고 제어하는 계측제어계통에서도 선제적으로 고장을 진단하고 대처하여 사고전파를 방지할 수 있는 기술개발이 필요하다. 이에 본 논문에서는 계측제어 장비의 환경조건과 자가 진단 데이터를 활용한 신뢰도 함수 추정 방안을 연구하였으며, 고장데이터의 획득을 위해 계측제어 장비의 부품에 대한 Feature 별 확률분포를 가정하여 가상 고장데이터를 생산하였다. 이러한 고장데이터를 바탕으로 생존분석에서 활용되는 대표적인 인공지능 모델(DeepSurve, DeepHit)을 이용하여 신뢰도 함수를 추정하였고, 그와 동시에 전통적인 준모수적 방법론인 Cox 회귀모델을 통해 신뢰도 함수를 추정하여 환경조건과 진단 데이터를 바탕으로 한 잔여 수명 계산을 통해 적용 가능성을 확인하였다.

Keywords

Acknowledgement

본 연구는 산업통상자원부와 한국에너지기술평가원의 지원을 받아 수행한 연구임.(No.20224B10100120)

References

  1. Choi, C. W., Kim, H. J., Shim, E. E., Im, A. Y., Lee, Y. J., Jeong, S. J. and Kim, K. B. (2020). Review of Lung Cancer Survival Analysis with Multimodal Data. Annual Conference of KIP S Vol.27 No.2.
  2. D'Agostino, R. B. and Nam, B. H. (2003). Evaluation of the performance of survival analysis models: discrimination and calibration measures. Handbook of Statistics 2003;23:1-25. DOI: 10.1016/S0169-7161(03)23001-7
  3. Faraggi, D. and Simon, R. (1995). A neural network model for survival data. Statistics in Medicine, 14(1):73-82. https://doi.org/10.1002/sim.4780140108
  4. Fotso, S. (2018). Deep neural networks for survival analysis based on a multi-task framework. arXiv preprints arXiv:1801.05512.
  5. Harrell, F. E., Califf, R. M., Pryor, D. B., Lee, K. L. and Rosati, R. A. (1982). Evaluating the yield of medical tests. Journal of the American Medical Association, 247 (18):2543-2546. https://doi.org/10.1001/jama.1982.03320430047030
  6. Heagerty, P. J. and Zheng, Y. (2005). Survival model predictive accuracy and ROC curves. Biometrics; 61(1):92-105. https://doi.org/10.1111/j.0006-341X.2005.030814.x
  7. Ishwaran, H., Kogalur, U. B., Blackstone, E. H. and Lauer, M. S. (2008) Random survival forests. The Annals of Applied Statistics, 2(3) 841-860. https://doi.org/10.1214/08-AOAS169
  8. Kamarudin, A. N., Cox, T., Kolamunnage-Dona, R. (2017). Time-dependent ROC curve analysis in medical research: current methods and applications. MC Medical Research Methodology, 17(1):53.
  9. Kapur, K. C. and Pecht, M. (2014). Reliability Engineering, New Jersey, John Wiley & Sons
  10. Katzman, J. L., Shaham, U., Cloninger, A., Bates, J., Jiang, T., and Kluger, Y. (2018). DeepSurv: personalized treatment recommender system using a Cox proportional hazards deep neural network. BMC Medical Research Methodology, 18(1).
  11. Lee, C., Zame, Z., Yoon, J., and Schaar, M. (2018). Deephit: A deep learning approach to survival analysis with competing risks. In Thirty-Second AAAI Conference on Artificial Intelligence, Vol.32 No.1.
  12. Luck, M., Sylvain, T., Cardinal, H., Lodi, A., and Bengio, Y. (2017). Deep learning for patient-specific kidney graft survival analysis. arXiv preprint, arXiv:1705.10245.
  13. Raykar, V. C., Steck, H., Krishnapuram, B., Dehing-Oberije, C. and Lambin, P. (2007). On ranking in survival analysis: bounds on the concordance index. In Proceedings of the 20th International Conference on Neural Information Processing Systems, pages 1209-1216.
  14. Sargent, D. J. (2001). Comparison of artificial neural networks with other statistical approaches. Cancer, 91(8):1636-42. https://doi.org/10.1002/1097-0142(20010415)91:8+<1636::AID-CNCR1176>3.0.CO;2-D
  15. Xiang, A., Lapuerta, P., Ryutov, A., Buckley, J., and Azen, S. (2000). Comparison of the performance of neural network methods and Cox regression for censored survival data. Computational Statistics & Data Analysis, 34:243-257. https://doi.org/10.1016/S0167-9473(99)00098-5
  16. Yang, J., Jeon, S., Lee, H. S. (2022) Predictive models and visualizations according to outcome variables using R-focusing on regression analyses. Journal of Health Informatics and Statistics, Vol.47 : 21-30. https://doi.org/10.21032/jhis.2022.47.S2.S21
  17. Yousefi, S., Amrollahi, F., Amgad, M., Dong, C., Lewis, J. E., Song, C., Gutman, D. A., Halani, S. H., Vega, J. E. V., Brat, D. J. and Lee, A. D. (2017). Predicting clinical outcomes from large scale cancer genomic profiles with deep survival models. Scientific Reports, 7(1): 11707.
  18. Zhu, X., Yao, J. and Huang, J. (2016). Deep convolutional neural network for survival analysis with pathological images. In 2016 IEEE International Conference on Bioinformatics and Biomedicine, pages 544-547.
  19. Zhu, X., Yao, J., Zhu, F., and Huang, J. (2017). WSISA: Making survival prediction from whole slide histopathological images. In 2017 IEEE Conference on Computer Vision and Pattern Recognition, pages 6855-6863.