Title: Bio-inspired mix design optimisation of self-compacting concrete using machine learning algorithms

Authors: Sriman Pankaj Boindala; Vasan Arunachalam; Jabez J. Christopher

Addresses: Department of Civil and Environmental Engineering, Technion – Israel Institute of Technology, Haifa 32000, Israel ' Department of Civil Engineering, Birla Institute of Technology and Science-Pilani, Hyderabad Campus, Telangana, 500078, India ' Department of Computer Science and Information Systems, Birla Institute of Technology and Science-Pilani, Hyderabad Campus, Telangana, 500078, India

Abstract: This study focuses on optimising concrete mix design using a hybrid bio-inspired optimisation algorithm that combines differential evolution (DE) and cuckoo search (CS). The study also evaluates the performance of two strength prediction models, artificial neural networks (ANNs) and support vector machine regression (SVR), in determining optimal mix proportions. The hybrid algorithm is tested using 11 benchmark test functions and the best approach is chosen to solve a mix design optimisation problem with the objectives of maximising compressive strength, minimising carbon emissions, and minimising cost. Results show that ANN outperforms SVR in terms of compressive strength, with a 30% increase observed. Both prediction models produce optimal mix proportions with minimal variation for cost and embodied carbon minimisation scenarios. The study demonstrates the efficacy of the hybrid optimisation algorithm in conjunction with a prediction model in determining optimal concrete mix proportions.

Keywords: bio-inspired optimisation; swarm intelligence algorithms; machine learning; compressive strength prediction model; concrete mix design optimisation; cuckoo search; differential evolution; differential cuckoo search; DCS; artificial neural networks; ANNs; support vector machine regression; SVR.

DOI: 10.1504/IJMMNO.2024.143237

International Journal of Mathematical Modelling and Numerical Optimisation, 2024 Vol.14 No.1/2, pp.1 - 31

Received: 07 Dec 2021
Accepted: 14 Apr 2023
Published online: 11 Dec 2024 *