Multi-objective optimization of concrete mixture proportions using machine learning and metaheuristic algorithms

Zhang, Junfei; Huang, Yimiao; Wang, Yuhang; Ma, Guowei

doi:10.1016/j.conbuildmat.2020.119208

Cited by 170 publications

(52 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…The combined algorithms RF, gradient boosting tree (GBDT), and extreme gradient boosting tree (XGBoost) all combine many DTs. In terms of model performance, XGBoost > GBDT > RF, RF prediction accuracy is comparable to ANN (Chopra et al 2018), better than RVM (Shaqadan and Alrawashdeh 2018), and has a superior fitting impact on the discretized distribution sample space when compared to ANN (Zhang et al 2020b). GBRT helps to address RF model overfitting and other difficulties by continuously adding new trees and altering the residual region, resulting in higher prediction performance (Zhang et al 2020a).…”

Section: Compressive Strengthmentioning

confidence: 98%

Progress in Artificial Intelligence-based Prediction of Concrete Performance

et al. 2021

ACT

View full text Add to dashboard Cite

Artificial intelligence technology has super high-dimensional nonlinear computing capabilities, intelligent comprehensive analysis and judgment functions, and self-learning knowledge reserve expression functions. It can unlock the potential of high-dimensional nonlinearity relation between tangible components and performance indicators when compared to the empirical formula generated from classic statistical approaches. This article summarizes the types of artificial intelligence algorithms used to predict concrete performance, comprehensively sorts out the research progress of artificial intelligence technology in predicting the mechanical properties, work performance, and durability of concrete, and compares and analyzes the effects of algorithm selection, sample data, and model construction on the concrete compressive prediction system. The analysis shows that artificial intelligence technology has obvious advantages in measurement accuracy in predicting concrete performance compared to conventional statistical methods. Multiple algorithms should be used to cross-validate the model prediction findings. For tiny data sets, support vector machines are utilized. Decision tree evolution techniques should be used in algorithm models that require feature optimization or dispersed index prediction. Artificial neural networks can be used to solve different challenges. To improve the prediction model and boost its prediction accuracy, measures such as optimized features, integrated algorithms, hyperparameter optimization, enlarged sample data set, richer data sources, and data pretreatment are proposed.

show abstract

Section: Compressive Strengthmentioning

confidence: 98%

Progress in Artificial Intelligence-based Prediction of Concrete Performance

et al. 2021

ACT

View full text Add to dashboard Cite

show abstract

“…The separations (d + i and d − i ) of each solution from the positive ideal solution and negative ideal solution are given as Equations (16) and (17). The relative closeness coefficient (R i ) can be calculated by Equation (18).…”

Section: Choice Criteria For Moo Problemsmentioning

confidence: 99%

“…to achieve a designed performance that meets requirements [16,17]. When there is only one objective to be optimized, the traditional experimental-based method is generally feasible [18]. However, 2 of 26 in most cases, proportion optimization of asphalt mixture is a complex problem, which needs to optimize multiple objectives simultaneously, such as performances, cost and environmental pollution, etc.…”

Section: Introductionmentioning

confidence: 99%

Machine Learning Approach to Develop a Novel Multi-Objective Optimization Method for Pavement Material Proportion

Liang

Chen

et al. 2021

Applied Sciences

View full text Add to dashboard Cite

Asphalt mixture proportion design is one of the most important steps in asphalt pavement design and application. This study proposes a novel multi-objective particle swarm optimization (MOPSO) algorithm employing the Gaussian process regression (GPR)-based machine learning (ML) method for multi-variable, multi-level optimization problems with multiple constraints. First, the GPR-based ML method is proposed to model the objective and constraint functions without the explicit relationships between variables and objectives. In the optimization step, the metaheuristic algorithm based on adaptive weight multi-objective particle swarm optimization (AWMOPSO) is used to achieve the global optimal solution, which is very efficient for the objectives and constraints without mathematical relationships. The results showed that the optimal GPR model could describe the relationship between variables and objectives well in terms of root-mean-square error (RMSE) and R2. After the optimization by the proposed GPR-AWMOPSO algorithm, the comprehensive pavement performances were enhanced in terms of the permanent deformation resistance at high temperature, crack resistance at low temperature as well as moisture stability. Therefore, the proposed GPR-AWMOPSO algorithm is the best option and efficient for maximizing the performances of composite modified asphalt mixture. The GPR-AWMOPSO algorithm has advantages of less computational time and fewer samples, higher accuracy, etc. over traditional laboratory-based experimental methods, which can serve as guidance for the proportion optimization design of asphalt pavement.

show abstract

“…Therefore, it is increasingly common to optimize the mixture using computational optimization methods. There are many different optimization methods such as Taguchi [15][16][17] and machine learning [18][19][20][21][22] methods. The Taguchi optimization method is one of the most preferred optimization methods to optimize experimental parameters because it is used both in analysis and optimization [23,24].…”

Section: Introductionmentioning

confidence: 99%