Mutual Evaluation of Different Machine Learning Algorithms for Effective Self-Compacting Lightweight Concrete Strength Prediction

Öz

Machine Learning (ML) methods are frequently used to model the mix design and to predict some properties of concrete. Traditional mix design calculations require a trial-and-error process to produce concrete with the desired properties. In this process, time, labor, and material losses are experienced. By using ML methods, the number of trial-and-error mixtures can be reduced, making it easier to produce concrete with the required properties. As a special type of concrete, Self-Compacting Lightweight Concrete (SCLWC), it becomes more important to speed up the production process by making the mix design practical. SCLWC, which contains both self-compactibility and lightweight aggregates, is more difficult to produce with the desired properties than conventional concrete. In this study, four different ML algorithms were comparatively evaluated to predict the compressive strength of SCLWC. The aim of the study is to predict the compressive strength from the composition of SCLWCC. In addition, to prove the success of ML in predicting concrete performance and thus increase the use of the method primarily for concrete mix calculation. For this purpose, 4 different prediction models were developed using Linear Regression, Gaussian Process Regression, Quadratik and Kernel Support Vector Machine methods with a data set containing some experimental properties and composition of SCLWC mix. The performance of the developed models was evaluated in terms of the linear correlation ratio between the experimental results and the model output and the amount of error. The Gaussian Process Regression model was the most successful ML method in this study.

Anahtar Kelimeler

• Self-compacting Lightweight Concrete
• Linear Regression
• Gaussian Process Regression
• Support Vector Machines

Efektif Kendiliğinden Yerleşen Hafif Beton Dayanımı Tahmini için Farklı Makine Öğrenmesi Algoritmalarının Karşılıklı Değerlendirilmesi

Öz

İnşaat mühendisliği literatüründe, özel beton türlerinin karışım tasarımını modellemek ve bu amaçla betonun bazı özelliklerini tahmin etmek için makine öğrenmesi yöntemleri sıklıkla kullanılmaktadır. Geleneksel karışım tasarımı hesapları, istenilen özellikte betonun üretilebilmesi için bir deneme yanılma süreci gerektirmektedir. Bu süreçte zaman, işgücü ve malzeme kayıpları yaşanır. Makine öğrenmesi yöntemleri kullanılarak deneme yanılma karışımlarının sayısı azaltılabileceğinden, gerek duyulan özelliklere sahip betonun üretilebilmesi kolaylaşabilir. Burada söz konusu malzeme Kendiliğinden Yerleşen Hafif Beton (KYHB) gibi özel bir tür beton ise karışım tasarımının pratik hale getirilmesi ile üretim sürecinin hızlandırılması daha büyük önem kazanır. Hem kendiliğinden yerleşen özellikte hem de hafif agrega içeren bu beton türünün istenilen özelliklerde üretilebilmesi süreci geleneksel betona göre daha zordur. Bu çalışmada, KYHB’nin basınç dayanımını tahmin etmek için dört farklı makine öğrenmesi algoritması karşılaştırılmalı değerlendirilmiştir. Çalışmanın amacı, KYHB bileşiminden basınç dayanımının tahmin edilmesidir. Ek olarak makine öğrenmesi algoritmalarının beton performansını tahmin etmedeki başarısını kanıtlamak ve böylece yöntemin öncelikle beton karışım hesabı için kullanımını artırmaktır. Bu amaçla, KYHB karışımının bazı deneysel özelliklerini ve bileşimini içeren bir veri seti ile Lineer Regresyon, Gauss Süreç Regresyonu, Quadratik ve Kernel Destek Vektör Makinası yöntemleri kullanılarak 4 farklı tahmin modeli geliştirilmiştir. Geliştirilen modellerin performansı, deneysel sonuçlar ile model çıktısı arasındaki doğrusal korelasyon oranı ve hata miktarları açısından değerlendirilmiştir. Gauss Süreç Regresyonu modeli ise bu çalışmada en başarılı makine öğrenmesi metodu olmuştur.

Anahtar Kelimeler

• Kendiliğinden Yerleşen Hafif Beton
• Lineer Regresyon
• Gauss Süreç Regresyonu
• Destek Vektör Makineleri

Kaynakça

1. Bibm C, Ermco, EFNARC E. European Guidelines for Self-Compacting Concrete: Specification, Production and Use. SCC European Project Group; 2005.
2. Sağlam RN, Açıkgenç Ulaş M, Alyamaç KE. Hafif Beton Üretimi İçin Gerekli Olan Hafif Agrega Miktarının Yapay Sinir Ağı ile Tahmin Edilmesi. Fırat Üniversitesi Mühendislik Bilim Derg. 2022;34(2):889-898.
3. Vakhshouri B, Nejadi S. Mix design of light-weight self-compacting concrete. Case Stud Constr Mater. 2016;4:1-14.
4. TS EN 206 Beton- Özellik, Performans, Imalat ve Uygunluk. Published online 2013.
5. TS 802 Beton Karışım Tasarımı Hesap Esasları. Published online 2016.
6. Agrawal Y, Gupta T, Sharma R, Panwar NL, Siddique S. A Comprehensive Review on the Performance of Structural Lightweight Aggregate Concrete for Sustainable Construction. Constr Mater. 2021;1(1):39-62.
7. Hu X, Li B, Mo Y, Alselwi O. Progress in Artificial Intelligence-based Prediction of Concrete Performance. J Adv Concr Technol. 2021;19(8):924-936.
8. Ziolkowski P, Niedostatkiewicz M, Kang SB. Model-Based Adaptive Machine Learning Approach in Concrete Mix Design. Materials (Basel). 2021;14(7).

9. Golafshani EM, Kim T, Behnood A, Ngo T, Kashani A. Sustainable mix design of recycled aggregate concrete using artificial intelligence. J Clean Prod. 2024;442:140994.
10. Marani A, Jamali A, Nehdi ML. Predicting Ultra-High-Performance Concrete Compressive Strength Using Tabular Generative Adversarial Networks. Mater (Basel, Switzerland). 2020;13(21).
11. Farooq F, Ahmed W, Akbar A, Aslam F, Alyousef R. Predictive modeling for sustainable high-performance concrete from industrial wastes: A comparison and optimization of models using ensemble learners. J Clean Prod. 2021;292:126032.
12. Dao D Van, Adeli H, Ly HB, et al. A Sensitivity and Robustness Analysis of GPR and ANN for High-Performance Concrete Compressive Strength Prediction Using a Monte Carlo Simulation. Sustainability. 2020;12(3).
13. Fan D, Yu R, Fu S, et al. Precise design and characteristics prediction of Ultra-High Performance Concrete (UHPC) based on artificial intelligence techniques. Cem Concr Compos. 2021;122:104171.
14. Moein MM, Saradar A, Rahmati K, et al. Predictive models for concrete properties using machine learning and deep learning approaches: A review. J Build Eng. 2023;63:105444.
15. Ahmad A, Farooq F, Niewiadomski P, et al. Prediction of Compressive Strength of Fly Ash Based Concrete Using Individual and Ensemble Algorithm. Materials (Basel). 2021;14(4).
16. Abuodeh OR, Abdalla JA, Hawileh RA. Assessment of compressive strength of Ultra-high Performance Concrete using deep machine learning techniques. Appl Soft Comput. 2020;95:106552.
17. Kaloop MR, Kumar D, Samui P, Hu JW, Kim D. Compressive strength prediction of high-performance concrete using gradient tree boosting machine. Constr Build Mater. 2020;264:120198.
18. Zheng W, Shui Z, Xu Z, Gao X, Zhang S. Multi-objective optimization of concrete mix design based on machine learning. J Build Eng. 2023;76:107396.
19. Andiç-Çakır Ö, Hızal S. Influence of elevated temperatures on the mechanical properties and microstructure of self consolidating lightweight aggregate concrete. Constr Build Mater. 2012;34:575-583.
20. Kaffetzakis M, Papanicolaou C. Mix Proportioning Method for Lightweight Aggregate SCC (LWASCC) Based on the Optimum Packing Point Concept BT - Innovative Materials and Techniques in Concrete Construction. In: Fardis MN, ed. Springer Netherlands; 2012:131-151.
21. Ofuyatan OM, Olutoge F, Omole D, Babafemi A. Influence of palm ash on properties of light weight self-compacting concrete. Clean Eng Technol. 2021;4:100233.
22. Aslani F, Ma G, Yim Wan DL, Muselin G. Development of high-performance self-compacting concrete using waste recycled concrete aggregates and rubber granules. J Clean Prod. 2018;182:553-566.
23. Mazloom M, Mahboubi F. Evaluating the settlement of lightweight coarse aggregate in self-compacting lightweight concrete. Comput Concr. 2017;19(2):203–210.
24. Ghanbari M, Kohnehpooshi O, Tohidi M. Experimental Study of the Combined Use of Fiber and Nano Silica Particles on the Properties of Lightweight Self Compacting Concrete. Int J Eng. 2020;33(8):1499-1511.
25. Andiç-Çakır Ö, Yoğurtçu E, Yazici S, Ramyar K. Self-compacting lightweight aggregate concrete: design and experimental study. Mag Concr Res. 2009;61(7):519-527.
26. Mazaheripour H, Ghanbarpour S, Mirmoradi SH, Hosseinpour I. The effect of polypropylene fibers on the properties of fresh and hardened lightweight self-compacting concrete. Constr Build Mater. 2011;25(1):351-358.
27. Maghsoudi AA, Mohamadpour S, Maghsoudi M. Mix design and mechanical properties of self compacting light weight concrete. Int J Civ Eng. 2011;9(3):230-236. https://www.scopus.com/inward/record.uri?eid=2-s2.0-80052764281&partnerID=40&md5=fbdde60f5121d637865440a4b9e9a0fa
28. Güneyisi E, Gesoğlu M, Booya E. Fresh properties of self-compacting cold bonded fly ash lightweight aggregate concrete with different mineral admixtures. Mater Struct. 2012;45(12):1849-1859.
29. Papanicolaou CG, Kaffetzakis MI. Lightweight Aggregate Self-Compacting Concrete: State-of-the-Art & Pumice Application. J Adv Concr Technol. 2011;9(1):15-29.
30. Zarnaghi VN, Fouroghi-Asl A, Nourani V, Ma H. On the pore structures of lightweight self-compacting concrete containing silica fume. Constr Build Mater. 2018;193:557-564.
31. Hilal NN, Sahab MF, Mohammed Ali TK. Fresh and hardened properties of lightweight self-compacting concrete containing walnut shells as coarse aggregate. J King Saud Univ - Eng Sci. 2021;33(5):364-372.
32. Li J, Chen Y, Wan C. A mix-design method for lightweight aggregate self-compacting concrete based on packing and mortar film thickness theories. Constr Build Mater. 2017;157:621-634.
33. Priya KS. Linear regression algorithm in machine learning through MATLAB. Int J Res Appl Sci Eng Technol. 2021;9(12):989-995.
34. Liemohn MW, Shane AD, Azari AR, Petersen AK, Swiger BM, Mukhopadhyay A. RMSE is not enough: Guidelines to robust data-model comparisons for magnetospheric physics. J Atmos Solar-Terrestrial Phys. 2021;218:105624.
35. Chicco D, Warrens MJ, Jurman G. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation. PeerJ Comput Sci. 2021;7:e623.
36. Kumar A, Verma A, Shinde G, Sukhdeve Y, Lal N. Crime Prediction Using K-Nearest Neighboring Algorithm. In: 2020 International Conference on Emerging Trends in Information Technology and Engineering (Ic-ETITE). ; 2020:1-4.
37. Purwanto A. Partial Least Squares Structural Squation Modeling (PLS-SEM) Analysis for Social and Management Research: A Literature Review. ournal Ind Eng Manag Res. 2021;2(4):114-123.
38. Açıkgenç Ulaş M. Gauss Süreç Regresyonu ve Destek Vektör Makineleri Kullanılarak Değerlendirilen Kendiliğinden Yerleşen Beton Davranışının Deneysel Veri İle Doğrulanması. Fırat Üniversitesi Mühendislik Bilim Derg. 2023;35(1):379-388.
39. Mahdaviara M, Rostami A, Keivanimehr F, Shahbazi K. Accurate determination of permeability in carbonate reservoirs using Gaussian Process Regression. J Pet Sci Eng. 2021;196:107807.
40. Dagher I. Quadratic kernel-free non-linear support vector machine. J Glob Optim. 2008;41(1):15-30.
41. Ye J, Yang Z, Ma M, Wang Y, Yang X. ∊-Kernel-free soft quadratic surface support vector regression. Inf Sci (Ny). 2022;594:177-199.
42. Jun Z. The Development and Application of Support Vector Machine. J Phys Conf Ser. 2021;1748(5):52006.