
Number of Journals | 31 |
Number of Issues | 805 |
Number of Articles | 7,756 |
Article View | 35,545,660 |
PDF Download | 7,908,157 |
Evaluation of construction components with eco-friendly recycled concrete with artificial neural network modeling approach and adaptive neural fuzzy inference system optimized with meta-heuristic algorithm | ||
تحقیقات بتن | ||
Articles in Press, Accepted Manuscript, Available Online from 05 May 2025 | ||
Document Type: Research Paper | ||
DOI: 10.22124/jcr.2025.29506.1689 | ||
Authors | ||
Mahyar azizkhani1; Ali Asghar AmirKardoost* 2; davood sedaghat shayegan2 | ||
1PhD candidate, Islamic azad university, roudehen branch | ||
2Department of civil engineering, Roudehen Branch, Islamic Azad University, Roudehen, Iran | ||
Abstract | ||
]n this study, modeling with environmentally friendly recycled concrete construction components has been investigated. To model the compressive strength of recycled concrete, two methods of artificial neural network and adaptive neural fuzzy inference system were used. Then, in order to increase the performance of smart methods and prevent the trial and error approach to achieve the most optimal parameter values of each model, the particle swarm optimization algorithm was used. The data of the mixing plan from the research background to create the proposed models were 201 laboratory data that were collected to estimate the compressive strength of recycled concrete. Then, to determine the best input parameters to the model, four scenarios were defined, from which the best scenario was selected. Also, for this index, the ANFIS-FCM method is more accurate than the rest of the neurophasic based methods. In the test phase, the artificial neural network method based on the particle swarm algorithm had a better performance than the ANN method. However, the ANFIS-FCM-PSO method performed slightly weaker than the ANFIS-FCM method in the training phase, but it performed better than the rest of the proposed methods in the test phase based on the three presented indicators. The results of the uncertainty analysis indicated that the proposed method had a very small margin of error in the term of average prediction. Also, among the proposed intelligent methods, the ANFIS-FCM-PSO method has better performance than other models with a slight difference. | ||
Keywords | ||
sustainable development; recycled concrete; compressive strength prediction; artificial neural network; adaptive neural fuzzy inference system; particle swarm optimization algorithm | ||
Statistics Article View: 117 |