Korean Journal of Construction Engineering and Management (한국건설관리학회논문집)

Korea Institute of Construction Engineering and Management (한국건설관리학회)
권호 표지
DOI QR코드

DOI QR Code

Condition Estimation of Facility Elements Using XGBoost

XGBoost를 활용한 시설물의 부재 상태 예측

  • Chang, Taeyeon (Department of Civil and Environmental Engineering, Seoul National University) ;
  • Yoon, Sihoo (Department of Civil and Environmental Engineering, Seoul National University) ;
  • Chi, Seokho (Department of Civil and Environmental Engineering, Seoul National University 4) ;
  • Im, Seokbeen (Research Institute for Safety Performance, Korea Authority of Land & Infrastructure Safety (KALIS))
  • 장태연 (서울대학교 건설환경공학부) ;
  • 윤시후 (서울대학교 건설환경공학부) ;
  • 지석호 (서울대학교 건설환경공학부) ;
  • 임석빈 (국토안전관리원 안전성능연구소 )
  • Received : 2022.09.30
  • Accepted : 2022.12.22
  • Published : 2023.01.31
Download PDF
⟨ Previous Next ⟩

Abstract

To reduce facility management costs and safety concerns due to aging of facilities, it is important to estimate the future facilities' condition based on facility management data and utilize predictive information for management decision making. To this end, this study proposed a methodology to estimate facility elements' condition using XGBoost. To validate the proposed methodology, this study constructed sample data for road bridges and developed a model to estimate condition grades of major elements expected in the next inspection. As a result, the developed model showed satisfactory performance in estimating the condition grades of deck, girder, and abutment/pier (average F1 score 0.869). In addition, a testbed was established that provides data management function and element condition estimation function to demonstrate the practical applicability of the proposed methodology. It was confirmed that the facility management data and predictive information in this study could help managers in making facility management decisions.

시설물의 고령화로 인한 유지관리 비용을 줄이고 안전성을 확보하기 위해서는 시설물 유지관리 데이터를 활용하여 향후 시설물의 상태를 예측하고 이를 유지관리 의사결정에 활용하는 것이 중요하다. 이를 위해 본 연구는 XGBoost를 활용하여 다양한 유지관리 정보로부터 향후 시설물의 부재 상태를 추정하는 방법론을 제안함을 목표로 한다. 방법론의 유효성을 검증하기 위해 교량시설물을 대상으로 샘플 데이터를 구축하고, 차기 정밀안전점검 및 정밀안전진단 시 부재 상태등급 예측모델을 개발 및 평가했다. 예측모델의 성능 평가 결과, 주요 부재(바닥판, 주형, 교대/교각) 상태등급을 예측하는 데 준수한 성능을 보였다(평균 F1 score 0.869). 또한 개발된 예측모델의 실무적 활용 가능성을 실증하기 위해 FMS 유지관리 데이터 관리 기능과 주요부재 상태등급 예측 기능을 제공하는 테스트베드를 구축했다. 이를 통해 본 연구에서 구축한 샘플 데이터와 예측모델을 활용하여 시설물 관리자에게 유지관리 의사결정에 필요한 시설물 정보 및 시설물 상태 예측정보를 제공할 수 있음을 확인할 수 있었다. 향후에는 추가적으로 데이터를 수집하고 다량의 데이터가 축적된 경우 좋은 성능을 보인다고 알려진 딥러닝 알고리즘을 활용함으로써 예측 성능을 높일 수 있다. 또한 제안된 방법론을 터널, 항만 등 다양한 시설물에 적용하여 상태등급 예측모델을 개발할 수 있다.

Keywords

Acknowledgement

본 연구는 국토안전관리원에서 수행하는 기본연구사업 (건설시설 안전분야 정보관리체계 개선 및 활용방안에 관한 연구)의 지원으로 수행되었습니다.

References

  1. Alipour, M., Harris D.K., and Barnes L.E. (2017). "Pattern Recognition in the National Bridge Inventory for Automated Screening and the Assessment of Infrastructure." Proc., Structures Congress 2017, ASCE, Reston, pp. 279-291.
  2. Bansal, S. (2018). "Data Science Trends on Kaggle." Kaggle, San Francisco, (April 1, 2022).
  3. Bektas, B.A., Carriquiry, A., and Smadi, O. (2013). "Using Classification Trees for Predicting National Bridge Inventory Condition Ratings." Journal of Infrastructure Systems, 19(4), pp. 425-433. https://doi.org/10.1061/(ASCE)IS.1943-555X.0000143
  4. Bolukbasi, M., Mohammadi, J., and Arditi, D. (2004). "Estimating the Future Condition of Highway Bridge Components Using National Bridge Inventory Data." Practice Periodical on Structural Design and Construction, 9(1), pp. 16-25. https://doi.org/10.1061/(ASCE)1084-0680(2004)9:1(16)
  5. Cesare, B.M.A., Santamarina, C., Members, A., Turkstra, C., and Vanmarcke, E.H. (1993). "Modeling Bridge Deterioration with Markov Chains." Journal of Transportaion Engineering, 118(6), pp. 820-833.
  6. Chawla, N.V., Bowyer, K.W., Hall, L.O., and Kegelmeyer, W.P. (2002). "SMOTE: Synthetic Minority Over-sampling Technique." Journal of Artificial Intelligence Research, 16, pp. 321-357. https://doi.org/10.1613/jair.953
  7. Chen, T., and Guestrin, C. (2016). "XGBoost: A Scalable Tree Boosting System." Proc., the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Association for Computing Machinery, New York, pp. 785-794.
  8. Friedman, J.H. (2001). "Greedy function approximation: a gradient boosting machine." Annals of statistics, pp. 1189-1232.
  9. Han, J., Kamber, M., and Pei, J. (2011). Data mining: concepts and techniques. Morgan Kaufmann, Waltham.
  10. Huang, J., Huang, N., Zhang, L., and Xu, H. (2012). "A method for feature selection based on the correlation analysis." Proc., 2012 International Conference on Measurement, Information, and Control, IEEE, Piscataway, pp. 529-532.
  11. Huang, Y.H. (2010). "Artificial Neural Network Model of Bridge Deterioration." Journal of Performance of Constructed Facilities, 24(6), pp. 597-602. https://doi.org/10.1061/(ASCE)CF.1943-5509.0000124
  12. Kim, Y.J., and Yoon, D.K. (2010). "Identifying Critical Sources of Bridge Deterioration in Cold Regions through the Constructed Bridges in North Dakota." Journal of Bridge Engineering, 15(5), pp. 542-552. https://doi.org/10.1061/(ASCE)BE.1943-5592.0000087
  13. Korea Authority of Land & Infrastructure Safety (KALIS) (2021). 2021 Statistics on Facilities.
  14. Lim, S., and Chi, S. (2019). "Xgboost application on bridge management systems for proactive damage estimation." Advanced Engineering Informatics, 41, 100922.
  15. Martinez, P., Mohamed, E., Mohsen, O., and Mohamed, Y. (2020). "Comparative Study of Data Mining Models for Prediction of Bridge Future Conditions." Journal of Performance of Constructed Facilities, 34(1), 04019108.
  16. Ministry of Land, Infrastructure and Transport (MOLIT) (2021a). Facility Management System, (Oct. 1, 2021).
  17. Ministry of Land, Infrastructure and Transport (MOLIT) (2021b, Dec. 1). Guidelines for Maintenance and Performance Assessments.
  18. Ministry of Land, Infrastructure and Transport (MOLIT) (2022). Traffic Monitoring System, (Feb. 10, 2022).
  19. Morcous, G. (2006). "Performance Prediction of Bridge Deck Systems Using Markov Chains." Journal of Performance of Constructed Facilities, 20(2), pp. 146-155. https://doi.org/10.1061/(ASCE)0887-3828(2006)20:2(146)
  20. Reddy, M.C., Balasubramanyam, P., and Subbarayudu, M. (2013). "An Effective Approach to Resolve Multicollinearity in Agriculture Data." International Journal of Research in Electronics and Computer Engineering, 1(1), pp. 27-30.
  21. Shmueli, G., Patel, N.R., and Bruce, P.C. (2010). Data Mining for Business Intelligence: Concepts, Techniques, and Applications in Microsoft Office Excel with XLMiner, Wiley, New York.
  22. Wu, B., Zhang, L., and Zhao, Y. (2014). "Feature selection via Cramer's V-test discretization for remote-sensing image classification." IEEE Transactions on Geoscience and Remote Sensing, 52(5), pp. 2593-2606.  https://doi.org/10.1109/TGRS.2013.2263510