Journal of the Korea Academia-Industrial cooperation Society (한국산학기술학회논문지)

The Korea Academia-Industrial cooperation Society (한국산학기술학회)
권호 표지
DOI QR코드

DOI QR Code

Modelling on the Carbonation Rate Prediction of Non-Transport Underground Infrastructures Using Deep Neural Network

심층신경망을 이용한 비운송 지중구조물의 탄산화속도 예측 모델링

  • Youn, Byong-Don (Division of R & D, PLANALL Engineering & Construction Inc.)
  • 윤병돈 ((주)플랜올이엔씨 연구개발부문)
  • Received : 2021.01.04
  • Accepted : 2021.04.02
  • Published : 2021.04.30
Download PDF
⟨ Previous Next ⟩

Abstract

PCT (Power Cable Tunnel) and UT (Utility Tunnel), which are non-transport underground infrastructures, are mostly RC (Reinforced Concrete) structures, and their durability decreases due to the deterioration caused by carbonation over time. In particular, since the rate of carbonation varies by use and region, a predictive model based on actual carbonation data is required for individual maintenance. In this study, a carbonation prediction model was developed for non-transport underground infrastructures, such as PCT and UT. A carbonation prediction model was developed using multiple regression analysis and deep neural network techniques based on the actual data obtained from a safety inspection. The structures, region, measurement location, construction method, measurement member, and concrete strength were selected as independent variables to determine the dependent variable carbonation rate coefficient in multiple regression analysis. The adjusted coefficient of determination (Ra2) of the multiple regression model was found to be 0.67. The coefficient of determination (R2) of the model for predicting the carbonation of non-transport underground infrastructures using a deep neural network was 0.82, which was superior to the comparative prediction model. These results are expected to help determine the optimal timing for repair on carbonation and preventive maintenance methodology for PCT and UT.

비운송 지중구조물인 전력구와 공동구는 대부분 철근 콘크리트 구조물로서 공용기간이 경과함에 따라 탄산화에 의한 열화로 내구성이 저하된다. 특히, 전력구 및 공동구는 용도별, 지역별로 탄산화 속도가 상이하므로 개별적인 유지관리를 위해서는 탄산화 실측 데이터에 기반한 예측 모델이 요구된다. 본 연구에서는 노후화 된 전력구 및 공동구와 같이 기존 비운송 지중구조물에 대한 탄산화 예측 모델을 개발하였다. 탄산화 예측 모델 개발을 위해 안전점검에서 확보한 실측 데이터를 기반으로 다중회귀분석 및 심층신경망 기법을 활용하였다. 다중회귀분석에서 종속 변수인 탄산화 속도계수 결정을 위해 독립 변수로서 구조물, 지역, 측정 위치, 시공 유형, 측정 부재, 콘크리트 강도를 선정하였으며, 다중회귀 예측 모델의 수정결정계수(Ra2)는 0.67로 분석되었다. 심층신경망을 이용한 비운송 지중구조물의 탄산화 예측 모델결정계수(R2)는 0.82로 나타났으며, 비교대상 모델보다 우수한 예측 성능을 보였다. 심층신경망을 이용한 비운송 지중구조물의 탄산화 예측 모델은 콘크리트 강도에 기초한 것으로, 본 연구의 결과가 노후화 된 전력구 및 공동구에 대한 탄산화 유지보수 최적 시기 결정 및 예방적 유지관리 방법론에 기여되길 기대한다.

Keywords

References

  1. Ministry of Culture, Sports and Tourism. Republic of Korea, Comprehensive measures to strengthen the safety of sustainable infrastructure [Internet]. Ministry of Culture, Sports and Tourism. Republic of Korea, c2019 [cited 2019 Jun. 18], Available From: http://www.korea.kr/news/policyBriefingView.do?newsId=156336845 (accessed Dec. 16, 2020)
  2. KEPRI, Development of Integrated Life Management System for Underground Structures based on LCC (Life Cycle Cost), Korea Electric Power Corporation, Korea, pp.312-327.
  3. Y. Lee, J. R. Lee, S. K. Woo, J. E. Nam, "Service life estimation of concrete box culvert for power transmission", Proceedings of the Korea Institute for Structural Maintenance and Inspection, Jeju, Korea, pp.474-475, Oct. 2014.
  4. J. H. Seong, Y. S. Lee, E. S. Hong, Y. S. Byun, "Development of performance assessment criterion for structures of shield TBM tunnel", Journal of Korean Tunnelling and Underground Space Association, Vol.17, No.5, pp.553-561, 2015. DOI: http://dx.doi.org/10.9711/KTAJ.2015.17.5.553
  5. L. Czarnecki, P. Woyciechowski, "Modelling of concrete carbonation; is it a process unlimited in time and restricted in space?", Bulletin of the Polish Academy of Science Technical Sciences, Vol.63, No.1, pp.43-54, 2015. DOI: http://dx.doi.org/10.1515/bpasts-2015-0006
  6. Alberto A. Sagues, Carbonation in Concrete and Effect on Steel Corrosion, WPI 0510685, Final Report, State Job 99700-3530-119, University of South Florida, College of Engineering, USA, pp.174-190.
  7. S. O. Ekolu, "Towards practical carbonation prediction and modelling for service life design of reinforced concrete structures", 2015 IOP Conf. Ser.: Mater. Sci. Eng., Riga, Latvia, Vol.96, 30th Sep. to 2nd Oct., 2015. DOI: http://dx.doi.org/10.1088/1757-899X/96/1/012065
  8. H. K. Kim, S. B. Kim, "Service life prediction and carbonation of bridge structures according to environmental conditions", Journal of the Korea Institute for Structural Maintenance and Inspection, Vol.14, No.4, pp.126-132, 2010. DOI: https://doi.org/10.11112/jksmi.2010.14.4.126
  9. S. W. Cho, C. S. Lee, "A proposal of durability prediction models and development of effective tunnel maintenance method through field application", Journal of the Korea Institute for Structural Maintenance and Inspection, Vol.16, No.5, pp.148-160, 2012. DOI: https://doi.org/10.11112/jksmi.2012.16.5.148
  10. T. K. Noh, Y. S. Shin, M. H. Go, H. S. Ryu, "Correlation analysis of compressive strength and carbonation depth in urban road tunnel", Proceedings of the Korea Institute for Structural Maintenance and Inspection, Busan, Korea, pp.57-85, Apr. 2015.
  11. C. S. Lee, Y. O. Kim, Y. H. Kim, "Probabilistic approach of carbonation speed of subway concrete structures", Proceedings of the Korean Society of Civil Engineers, Daejeon, Korea, pp.1468-1471, Oct. 2008.
  12. B. D. Youn, H. Hamada, "Probabilistic approach on the Carbonation Rate of Non-Transport Underground Infrastructures", Advances in Construction Materials Proceedings of the Conmat'20, Sixth International Conference on Construction Materials, Fukuoka, Japan, pp.290-299, Aug. 2020.
  13. JSCE, Standard Specification for Concrete Structures - 2007; Maintenance, p.278, Japan Society of Civil Engineering, 2007, pp.110-112.
  14. W. Z. Taffese, F. Al-Neshawy, E. Sistonen, M. Ferreira, "Optimized neural network based carbonation prediction model", International Symposium Non-Destructive Testing in Civil Engineering (NDT-CE), Berlin, Germany, pp.1074-1083, Sep. 2015.
  15. H. M. Lee, H. S. Lee, "Prediction of Carbonation Progress of Concrete Using Deep Learning", Proceedings of the Korea Concrete Institute, Jeju, Korea, pp.171-172, May 2017.
  16. D. H. Jung, H. S. Lee, "A Fundamental Study on the Prediction of Carbonation Progress Using Deep Learning Algorithm Considering Mixing Factors", Proceedings of the Korea Institute of Building Construction, Yeosu, Korea, pp.30-31, May 2019.
  17. D. H. Jung, H. S. Lee, "A Fundamental Study on the Effect of Activation Function in Predicting Carbonation Progress Using Deep Learning Algorithm", Proceedings of the Korea Institute of Building Construction, Chungju, Korea, pp.60-61, Nov. 2019.
  18. N. D. Lewise, Deep Learning: Made Easy with R, p.283, Acorn Publishing Co., 2017, pp.34-79.
  19. B. Lantz, Machine Learning with R: 2nd ed., p.566, Acorn Publishing Co., 2017, pp.122-310.
  20. G. Hinton, N. Srivastava, A. Krizhevsky, I. Sutskever, R. R. Salakhutdinov, Improving neural networks by preventing co-adaptation of feature detectors [Internet]. Department of Computer Science, University of Toronto, c2012 [cited 2020 Oct. 5], vailable From: http://arxiv.org/abs/1207.0580 (accessed Dec. 16, 2020)
  21. A. Nazari, F. P. Torgal, "Predicting compressive strength of different geopolymers by artificial neural networks", Ceramics International, Vol.39, No.3, pp.2247-2257, Apr. 2019. DOI: http://dx.doi.org/10.1016/j.ceramint.2012.08.070
  22. KCI, Standard Specification for Concrete; Maintenance Description, p.189, Korea Concrete Institute, 2005, pp.77-106.
  23. AIJ, Japanese Architectural Standard Specification JASS 5; Reinforced concrete construction, p.754, Architectural Institute of Japan, 1997.