Data-driven optimisation of crushing system configuration in large-scale copper mines to reduce CO 2 emissions

This study develops a novel data-driven framework to minimise energy consumption and CO2 emissions in copper mine crushing systems. Unlike previous works, it compares single-unit and multi-unit configurations from an environmental perspective. Using a Random Forest model, key design parameters are identified and ranked. K-means clustering then groups operational data into four representative clusters, and Particle Swarm Optimization determines the optimal variable combination, reducing the objective function. Results show stable convergence and demonstrate that the proposed hybrid approach significantly enhances energy efficiency and reduces the carbon footprint in crushing operations.