Mixture optimization for environmental, economical and mechanical objectives in silica fume concrete: A novel frame-work based on machine learning and a new meta-heuristic algorithm

Partial replacement of cement by silica fume in concrete provides advantages such as mitigation the impact on environment carbon dioxide emitted during production, recycling industrial by-products and improvement strength durability. The optimization mixture (SFC) requires trade-off among multiple objectives (strength, cost embodied CO2) consideration a large number variables under highly nonlinear constraints. Obtaining Pareto front this multi-objective (MOO) problem is computationally expensive. To address issue, present study develops MOO model using machine learning (ML) techniques new meta-heuristic algorithm. Firstly, relationships between components SFC properties are modelled dataset back propagation neural network (BPNN) model. Then an individual-intelligence-based beetle antennae search algorithm (MOBAS) developed to for optimal mixtures that maximize UCS, minimize CO2 defined Results indicate proposed MOBAS more efficient with satisfactory accuracy comparison algorithms based swarm intelligence. achieves reliable predictions UCS very high correlation coefficient (0.9663) test set. proportions successfully obtained frame-work improves efficiency can facilitate appropriate decision making before construction.