Machine learning-assisted dual-objective synergistic optimization for mechanical and electrical properties of CNTs/Cu composites

Abstract

Overcoming the tensile strength-electrical conductivity trade-off in CNTs/Cu composites remains challenging due to complex multi-parameter coupling. We present a novel machine learning (ML)-assisted synergistic design framework enabling simultaneous optimization of both properties. Systematically evaluating seven ML algorithms under diverse preprocessing strategies, we established an effective modeling approach for limited datasets, achieving sub-10 % prediction errors for ultimate tensile strength and conductivity. Crucially, integrating ML predictions with Pareto optimization via the weighted Tchebycheff methodology generated experimentally validated, manufacturable process parameter sets, resolving conventional trial-and-error limitations. Experimental fabrication and comprehensive characterization confirmed the framework's accuracy, practicality, scalability, and engineering significance. This research establishes an intelligent pathway for the synergistic design of advanced metal matrix composites.