DOI QR코드

DOI QR Code

ANN 및 SVR 알고리즘을 활용한 최적 터널지보패턴 선정에 관한 연구

A study on the optimization of tunnel support patterns using ANN and SVR algorithms

  • 이제겸 (한양대학교 자원환경공학과) ;
  • 김양균 (한양대학교 자원환경공학과) ;
  • 이승원 (한양대학교 자원환경공학과)
  • Lee, Je-Kyum (Dept. of Earch Resources and Environmental Engineering, Hanyang University) ;
  • Kim, YangKyun (Dept. of Earch Resources and Environmental Engineering, Hanyang University) ;
  • Lee, Sean Seungwon (Dept. of Earch Resources and Environmental Engineering, Hanyang University)
  • 투고 : 2022.10.13
  • 심사 : 2022.10.27
  • 발행 : 2022.11.30

초록

터널 건설 시 암반 등급에 따라 다양한 지보재를 적절히 병용하여 지보패턴을 결정하고 시공이 이루어진다. 이 과정에서 시공 경험이 풍부한 전문가의 기술적 판단이 필요한데, 터널 설계의 초기 단계인 타당성 조사 및 기본설계 단계에는 상대적으로 짧은 수행기간과 부족한 자료 및 예산으로 인해 설계에 많은 어려움이 존재한다. 터널 건설의 급증과 함께 축적된 설계 데이터와 머신러닝을 활용한다면, 지보패턴 설계를 보다 신속하고 신뢰도 있게 수행할 수 있다. 따라서 본 연구에서는 암반등급 판정 및 해당 암반등급에 적합한 지보패턴 설계를 자동화하고자 국내 48개 도로터널의 설계자료 및 지반조사 자료를 수집하였으며, 8개의 입력항목(암종, 전기비저항, 심도, 터널연장, 터널연장에 따른 방재등급, 위험도지수에 따른 방재등급, 터널 종류, 터널 단면적)과 11개의 출력항목(암반등급, 숏크리트 제원 2개 항목, 록볼트 제원 3개 항목, 강지보 제원 3개 항목, 콘크리트 라이닝 2개 항목)에 대한 데이터를 정리하였다. 이와 같이 정리된 데이터를 활용하여 2가지 머신러닝 알고리즘(SVM, ANN)을 활용하여 3가지 머신러닝 모델(S1, A1, A2)을 개발하였으며, 세 가지 모델의 성능을 비교해본 결과 출력값의 데이터 형식에 따라 서로 다른 손실함수를 적용한 ANN 기반의 A2 모델이 가장 뛰어난 성능을 보였다. 본 연구를 통해 머신러닝을 활용한 지보패턴 설계의 가능성을 확인할 수 있었으며, 향후 지속적으로 실제 설계에 사용함으로써 단점을 보완하고 적용성을 개선해 나간다면 설계에 보다 큰 도움을 줄 수 있는 지보패턴 설계 모델을 개발할 수 있을 것으로 기대된다.

A ground support pattern should be designed by properly integrating various support materials in accordance with the rock mass grade when constructing a tunnel, and a technical decision must be made in this process by professionals with vast construction experiences. However, designing supports at the early stage of tunnel design, such as feasibility study or basic design, may be very challenging due to the short timeline, insufficient budget, and deficiency of field data. Meanwhile, the design of the support pattern can be performed more quickly and reliably by utilizing the machine learning technique and the accumulated design data with the rapid increase in tunnel construction in South Korea. Therefore, in this study, the design data and ground exploration data of 48 road tunnels in South Korea were inspected, and data about 19 items, including eight input items (rock type, resistivity, depth, tunnel length, safety index by tunnel length, safety index by rick index, tunnel type, tunnel area) and 11 output items (rock mass grade, two items for shotcrete, three items for rock bolt, three items for steel support, two items for concrete lining), were collected to automatically determine the rock mass class and the support pattern. Three machine learning models (S1, A1, A2) were developed using two machine learning algorithms (SVR, ANN) and organized data. As a result, the A2 model, which applied different loss functions according to the output data format, showed the best performance. This study confirms the potential of support pattern design using machine learning, and it is expected that it will be able to improve the design model by continuously using the model in the actual design, compensating for its shortcomings, and improving its usability.

키워드

과제정보

본 연구는 국토교통부/국토교통과학기술진흥원의 지원으로 수행되었습니다(과제번호: RS-2021-KA163775, 빅데이터와 인공지능 기반의 발파굴착터널 자동설계기술 개발을 위한 기초연구). 이에 감사드립니다.

참고문헌

  1. Carbonell, A., Gonzalez-Vidosa, F., Victor, Y. (2011), "Design of reinforced concrete road vaults by heuristic optimization", Advances in Engineering Software, Vol. 42, No. 4, pp. 151-159. https://doi.org/10.1016/j.advengsoft.2011.01.002
  2. Galende-Hernandez, M., Menendez, M., Fuente, M.J., Sainz-Palmero, G.I. (2018), "Monitor-while-drillingbased estimation of rock mass rating with computational intelligence: the case of tunnel excavation front", Automation in Construction, Vol. 93, pp. 325-338. https://doi.org/10.1016/j.autcon.2018.05.019
  3. Kang, K.N., An, J.S., Kim, B.C., Song, K.I. (2018), "Optimization of tunnel support patterns using DEA", Journal of Korean Tunnelling and Underground Space Association, Vol. 20, No. 1, pp. 211-224. https://doi.org/10.9711/KTAJ.2018.20.1.211
  4. Kim, B.C., Moon, H.K. (2016), "A study on the optimization technique for the plan of slope reinforcement arrangement of soil-nailing in tunnel portal area", Journal of Korean Tunnelling and Underground Space Association, Vol. 18, No. 6, pp. 569-579. https://doi.org/10.9711/KTAJ.2016.18.6.569
  5. Kim, J.Y., Seo, J.W., Kang, K.I. (2003). "A study on the selection model of retaining wall bearing methods using neural network", Journal of the Architectural of Korea Structure & Construction, Vol. 19, No. 5, pp. 121-128.
  6. Kim, M., Jung, J., Lee, J., Park, M., Bak, J., Lee, S.S. (2021), "Development and application of largediameter cut-hole exploration system for assessment of the geological condition beyond NATM tunnel face", Tunnel and Underground Space, Vol. 31, No. 1, pp. 1-9. https://doi.org/10.7474/TUS.2021.31.1.001
  7. Kim, M.S., Lee, S.S. (2022), "Investigation of geological conditions beyond the excavation face using a MSP boring data monitoring system", International Journal of Rock Mechanics and Mining Sciences, Vol. 157, 105161.
  8. 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
  9. Lee, J.K., Choi, W.H., Kim, Y., Lee, S.S. (2021), "A study on the rock mass classification in boreholes for a tunnel design using machine learning algorithms", Journal of Korean Tunnelling and Underground Space Association, Vol. 23, No. 6, pp. 469-484. https://doi.org/10.9711/KTAJ.2021.23.6.469
  10. MOLIT (2011), Road design guideline (the chapter of tunnel), Ministry of Land, Infrastructure and Transport, Korea, pp. 606-3 - 606-4.
  11. MOLIT (2022), Yearbook of road bridge and tunnel statistics, Ministry of Land, Infrastructure and Transport, Korea, pp. 13.
  12. Rai, P., Schunesson, H., Lindqvist, P.A., Kumar, U. (2015), "An overview on measurement-while-drilling technique and its scope in excavation industry", Journal of The Institution of Engineers (India), Vol. 96, No. 1, pp. 57-66.
  13. Shin, H.S., Kwon, Y.C. (2009), "Development of a window-shifting ANN training method for a quantitative rock classification in unsampled rock zone", Journal of Korean Tunnelling and Underground Space Association, Vol. 11, No. 2, pp. 151-162. https://doi.org/10.9711/KTAJ.2009.11.2.151
  14. Soranzo, E., Guardiani, C., Wu, W. (2022), "The application of reinforcement learning to NATM tunnel design", Underground Space (In Press), Vol. 7, No. 6, pp. 990-1002. https://doi.org/10.1016/j.undsp.2022.01.005