Sustainable mixture design of rice husk ash cement based concrete: performance optimization through data driven modeling and multi objective analysis

Abstract

Rice husk ash (RHA), a silica-rich by-product from rice husk combustion, serves as an eco-friendly partial cement replacement, contributing to reduced greenhouse gas emissions and sustainable construction practices. This study develops a practical predictive optimization framework for Rice Husk Ash Concrete (RHAC) targeting compressive strength (CS) alongside CO2 and SO2 emission reductions. Two ensemble learning strategies, voting and stacking, integrate Histogram Gradient Boosting (HGB) and Light Gradient Boosting (LGB) models, complemented by a Dempster–Shafer-based ensemble for enhanced prediction accuracy. Multi-objective optimization is performed using the Artificial Protozoa Optimizer (APO) and Electric Eel Foraging Optimization (EEFO) to explore trade-offs between mechanical and environmental performance. The Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) supports final decision-making, identifying balanced RHAC mixture designs. The proposed approach delivers high predictive accuracy and effective optimization, offering a robust pathway toward greener and more sustainable concrete production.