Journal of Korean Tunnelling and Underground Space Association (한국터널지하공간학회 논문집)

Korean Tunnelling and Underground Space Association (한국터널지하공간학회)
권호 표지
DOI QR코드

DOI QR Code

A study on the rock mass classification in boreholes for a tunnel design using machine learning algorithms

머신러닝 기법을 활용한 터널 설계 시 시추공 내 암반분류에 관한 연구

  • Lee, Je-Kyum (Dept. of Earth Resources and Environmental Engineering, Hanyang University) ;
  • Choi, Won-Hyuk (Dept. of Earth Resources and Environmental Engineering, Hanyang University) ;
  • Kim, Yangkyun (Dept. of Earth Resources and Environmental Engineering, Hanyang University) ;
  • Lee, Sean Seungwon (Dept. of Earth Resources and Environmental Engineering, Hanyang University)
  • 이제겸 (한양대학교 자원환경공학과) ;
  • 최원혁 (한양대학교 자원환경공학과) ;
  • 김양균 (한양대학교 자원환경공학과) ;
  • 이승원 (한양대학교 자원환경공학과)
  • Received : 2021.10.19
  • Accepted : 2021.11.01
  • Published : 2021.11.30
Download PDF
⟨ Previous Next ⟩

Abstract

Rock mass classification results have a great influence on construction schedule and budget as well as tunnel stability in tunnel design. A total of 3,526 tunnels have been constructed in Korea and the associated techniques in tunnel design and construction have been continuously developed, however, not many studies have been performed on how to assess rock mass quality and grade more accurately. Thus, numerous cases show big differences in the results according to inspectors' experience and judgement. Hence, this study aims to suggest a more reliable rock mass classification (RMR) model using machine learning algorithms, which is surging in availability, through the analyses based on various rock and rock mass information collected from boring investigations. For this, 11 learning parameters (depth, rock type, RQD, electrical resistivity, UCS, Vp, Vs, Young's modulus, unit weight, Poisson's ratio, RMR) from 13 local tunnel cases were selected, 337 learning data sets as well as 60 test data sets were prepared, and 6 machine learning algorithms (DT, SVM, ANN, PCA & ANN, RF, XGBoost) were tested for various hyperparameters for each algorithm. The results show that the mean absolute errors in RMR value from five algorithms except Decision Tree were less than 8 and a Support Vector Machine model is the best model. The applicability of the model, established through this study, was confirmed and this prediction model can be applied for more reliable rock mass classification when additional various data is continuously cumulated.

터널 설계 시 지반조사를 통한 암반분류 결과는 공사기간 및 공사비 산출, 그리고 터널안정성 평가에 지대한 영향을 미친다. 국내에서 지금까지 완공된 3,526개소의 터널들의 설계 및 시공을 통해 관련 기술들은 지속적으로 발전되어 왔지만, 터널 설계 시 암질 및 암반등급을 보다 정확하게 평가하기 위한 방법에 대한 연구는 미미하여 평가자의 경험 및 주관에 따라 결과의 차이가 큰 경우가 적지 않다. 따라서 본 연구에서는 암석샘플에 대한 주관적 평가를 통한 기존의 인력에 의한 암반분류 대신, 최근 지반분야에서도 그 활용도가 급증하고 있는 머신러닝 알고리즘을 이용하여 시추조사에서 획득한 다양한 암석 및 암반정보를 분석하여 보다 신뢰성있는 RMR에 의한 암반분류 모델을 제시하고자 하였다. 국내 13개 터널을 대상으로 11개의 학습 인자(심도, 암종, RQD, 전기비저항, 일축압축강도, 탄성파 P파속도 및 S파 속도, 영률, 단위중량, 포아송비, RMR)를 선정하여 337개의 학습 데이터셋과 60개의 시험 데이터셋을 확보하였으며, 모델의 예측성능을 향상시키기 위해 6개의 머신러닝 알고리즘(DT, SVM, ANN, PCA & ANN, RF, XGBoost)과 각 알고리즘별 다양한 초매개변수(hyperparameter)를 적용하였다. 학습된 모델의 예측성능을 비교한 결과, DT 모델을 제외한 5개의 머신러닝 모델에서 시험데이터에 대한 RMR 평균절대오차 값이 8 미만으로 수렴되었으며, SVM 모델에서 가장 우수한 예측성능을 나타내었다. 본 연구를 통해 암반분류 예측에 대한 머신러닝 기법의 적용 가능성을 확인하였으며, 향후 다양한 데이터를 지속적으로 확보하여 예측모델의 성능을 향상시킨다면 보다 신뢰성 있는 암반 분류에 활용될 수 있을 것으로 기대된다.

Keywords

Acknowledgement

본 연구는 국토교통부/국토교통과학기술진흥원의 지원(빅데이터와 인공지능 기반의 발파굴착터널 자동설계기술 개발을 위한 기초연구, 과제번호: 21CTAP-C163775-01)으로 수행되었습니다. 이에 감사드립니다.

References

  1. Choi, J.H., Jo, C.H., Ryu, D.W., Kim, H., Oh, B.S., Kang, M.G., Suh, B.S. (2003), "A study on the correlation between the result of electrical resistivity survey and the rock mass classification values determined by the tunnel face mapping", Tunnel and Underground Space, Vol. 13, No. 4, pp. 279-286.
  2. Han, M.S., Yang, I.J., Kim, K.M. (2002), "A prediction of the rock mass rating of tunnelling area using artificial neural networks", Journal of Korean Tunnelling and Underground Space Association, Vol. 4, No. 4, pp. 277-286.
  3. Kim, H.J., Kim, W.S., Kim, Y.G., Lee, D.W., Park, J.W., Kim, K.B. (2004), "Reliability estimation of site investigation by geotechnical risk analysis in subway tunnel", Proceedings of the Korean Society for Rock Mechanics Conference, Seoul, pp. 13-22.
  4. Kim, J.H., Kwon, O.I., Baek, Y., Yoo, W.K. (2015), "Study on improvement of rock mass classification for tunnel", Crisisonomy, Vol. 11, No. 7, pp. 147-155.
  5. Kim, Y. (2021), "An analysis of artificial intelligence algorithms applied to rock engineering", Tunnel and Underground Space, Vol. 31, No. 1, pp. 25-40. https://doi.org/10.7474/TUS.2021.31.1.025
  6. Kwon, H.S., Hwang, S.H., Baek, H.J., Kim, K.S. (2008), "A study on the correlation between electrical resistivity and rock classification", Geophysics and Geophysical Exploration, Vol. 11, No. 4, pp. 350-360.
  7. Lee, K.H., Seo, H.J., Park, J.H., Ahn, H.Y., Kim, K.S., Lee, I.M. (2012), "A study on correlation between electrical resistivity obtained from electrical resistivity logging and rock mass rating in-situ tunnelling site", Journal of Korean Tunnelling and Underground Space Association, Vol. 14, No. 5, pp. 503-516. https://doi.org/10.9711/KTAJ.2012.14.5.503
  8. Lee, K.J., Ha, H.S., Ko, K.B., Kim, J.S. (2009), "Investigation of indicator kriging for evaluating proper rock mass classification based on electrical resistivity and RMR correlation analysis", Tunnel and Underground Space, Vol. 19, No. 5, pp. 407-420.
  9. Lee, K.N., Park, Y.J., Kim, K.S. (2018), "Comparison of seismic velocity and rock mass rating from in situ measurement", Tunnel and Underground Space, Vol. 28, No. 3, pp. 232-246. https://doi.org/10.7474/TUS.2018.28.3.232
  10. Loo, B.P.Y. (2003), "Tunnel traffic and toll elasticities in Hong Kong: some recent evidence for international comparisons", Environment and Planning A: Economy and Space, Vol. 35, No. 2, pp. 249-276. https://doi.org/10.1068/a3590
  11. MOLIT Statistics System (Website) Road Bridge and Tunnel Statistics, http://stat.molit.go.kr/portal/cate/statView.do?hRsId=65&hFormId=1040&hDivEng=&month_yn= (September 24, 2021a).
  12. MOLIT Statistics System (Website) Train Tunnel Statistics, http://stat.molit.go.kr/portal/cate/statView.do?hRsId=302&hFormId=4746&hDivEng=&month_yn= (September 24, 2021b).
  13. Ryu, H.H., Joo, G.W., Cho, G.C., Kim, K.Y., Lim, Y.D. (2013), "Probabilistic rock mass classification using electrical resistivity - Theoretical approach of relationship between RMR and electrical resistivity-", Journal of Korean Tunnelling and Underground Space Association, Vol. 15, No. 2, pp. 97-111. https://doi.org/10.9711/KTAJ.2013.15.2.097
  14. Srisa-An, C. (2021), "Guideline of collinearity - avoidable regression models on time-series analysis", Proceedings of the 2nd International Conference on Big Data Analytics and Practices, Tetouan, Morocco, pp. 28-32.
  15. Sunwoo, C., Ryu, D.W., Kim, H.M., Kim, K.S. (2011), "Study on the geotechnical characteristics of granite in Korea and their correlation with rock classification method", Tunnel and Underground Space, Vol. 21, No. 3, pp. 205-215. https://doi.org/10.7474/TUS.2011.21.3.205
  16. You, J.O., Baek, S.K. (2003), "A study of rock mass rating using elastic wave of TSP survey", Proceedings of the KSCE 2003 Convention Conference, Daegu, pp. 4833-4838.