DOI QR코드

DOI QR Code

A Study on the Calculation of Ternary Concrete Mixing using Bidirectional DNN Analysis

양방향 DNN 해석을 이용한 삼성분계 콘크리트의 배합 산정에 관한 연구

  • Choi, Ju-Hee (Department of Smart-City Engineering, Hanyang University) ;
  • Ko, Min-Sam (Department of ICT, Hanyang University) ;
  • Lee, Han-Seung (Department of Architectural Engineering, Hanyang University)
  • Received : 2022.10.20
  • Accepted : 2022.12.04
  • Published : 2022.12.20

Abstract

The concrete mix design and compressive strength evaluation are used as basic data for the durability of sustainable structures. However, the recent diversification of mixing factors has created difficulties in calculating the correct mixing factor or setting the reference value concrete mixing design. The purpose of this study is to design a predictive model of bidirectional analysis that calculates the mixing elements of ternary concrete using deep learning, one of the artificial intelligence techniques. For the DNN-based predictive model for calculating the concrete mixing factor, performance evaluation and comparison were performed using a total of 8 models with the number of layers and the number of hidden neurons as variables. The combination calculation result was output. As a result of the model's performance evaluation, an average error rate of about 1.423% for the concrete compressive strength factor was achieved. and an average MAPE error of 8.22% for the prediction of the ternary concrete mixing factor was satisfied. Through comparing the performance evaluation for each structure of the DNN model, the DNN5L-2048 model showed the highest performance for all compounding factors. Using the learned DNN model, the prediction of the ternary concrete formulation table with the required compressive strength of 30 and 50 MPa was carried out. The verification process through the expansion of the data set for learning and a comparison between the actual concrete mix table and the DNN model output concrete mix table is necessary.

콘크리트의 배합설계와 압축강도 평가는 지속가능한 구조물의 내구성을 위한 기초적인 자료로서 활용되고 있다. 하지만, 콘크리트 배합설계는 최근 배합요소의 다변화 등의 이유로 인하여 정확한 배합요소 산정이나 기준값 설정에 어려움을 겪고 있다. 본 연구에서는 인공지능 기법 중 하나인 딥러닝 기법을 사용하여 삼성분계 콘크리트의 배합요소를 산정하는 양방향 해석의 예측모델을 설계하는 것을 목적으로 한다. 콘크리트 배합요소 산정을 위한 DNN 기반 예측모 델은 층 수, 은닉 뉴런 수를 변수로 한 총 8개의 모델을 사용하여 성능평가 및 비교를 실시하였으며, 이후 학습된 DNN 모델을 사용하여 소요압축강도에 따른 콘크리트 배합 산정 결과를 출력하였다. 모델의 성능평가 결과, 콘크리트 압축 강도 인자에 대하여 평균 약 1.423%의 오류율을 나타내었으며, 삼성분계 콘크리트 배합인자 예측에 대하여 평균 8.22%의 MAPE 오차를 만족하였다. DNN 모델의 구조별 성능평가 비교 결과, 모든 배합인자에 대하여 DNN5L-2048 모델이 가장 높은 성능을 보였다. 학습된 DNN 모델을 사용하여 30, 50MPa의 소요압축강도를 가지는 삼성분계 콘크 리트 배합표 예측을 진행하였으며, 추후 학습을 위한 데이터 세트 확장과 실제 콘크리트 배합표와 DNN 모델 출력 콘 크리트 배합표 간의 비교를 통한 검증 과정이 필요할 것으로 판단된다.

Keywords

Acknowledgement

This work was supported by the National Research Foundation of Korea(NRF) grant funded by the Korea government(MSIT)(No. NRF-2018R1A5A1025137).

References

  1. Kim IS, Lee JH, Yang DS, Park SK. Prediction on mix proportion factor and strength of concrete using neural network. Journal of the Korea Concrete Institute. 2002 Aug;14(4):457-66. https://doi.org/10.4334/JKCI.2002.14.4.457
  2. Bengio Y, Courville A, Vincent P. Representation learning: A review and new perspectives. IEEE transactions on pattern analysis and machine intelligence. 2013 Mar;35(8):1798-828. https://doi.org/10.1109/TPAMI.2013.50
  3. Bello SA, Oyedele L, Olaitan OK, Olonade KA, Olajumoke AM, Ajayi A, Akanbi L, Akinade O, Sanni ML, Bello AL. A deep learning approach to concrete water-cement ratio prediction. Results in Materials. 2022 Sep;15:100300. https://doi.org/10.1016/j.rinma.2022.100300
  4. Werbos PJ. Generalization of backpropagation with application to a recurrent gas market model. Neural networks. 1988;1(4):339-56. https://doi.org/10.1016/0893-6080(88)90007-X
  5. Squartini S, Hussain A, Piazza F. Preprocessing based solution for the vanishing gradient problem in recurrent neural networks. Proceedings of the 2003 International Symposium on Circuits and Systems; 2003 May 25-28. Bangkok (Thailand): NY: Institute of Electrical and Electronics Engineers; 2003. p. 7762895. https://doi.org/10.1109/ISCAS.2003.1206412
  6. Chicco D, Warrens M J, Jurman G. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation. PeerJ Computer Science. 2021 Jul;7:e623. https://doi.org/10.7717/peerj-cs.623
  7. Im CH, Jee NY, Cho HB. The estimation of compressive strength of concrete used admixture on the basis of mix design. Spring Annual Conference of Architectural Institute of Korea; 2003 Apr 26; Yongin (Korea): Seoul (Korea): Architectural Institute of Korea; 2003. p. 251-4.