Accelerated design for magnetic high entropy alloys using data-driven multi-objective optimization

Abstract

High entropy alloys (HEAs) allow for attractive combinations of excellent mechanical and magnetic properties. The composition space of HEAs is enormous as it comprises multiple principal elements. Therefore, it is...

Accelerated Design for High-Entropy Alloys Based on Machine Learning and Multiobjective Optimization

“…Debnath et al 49 developed an artificial neural network (ANN) to map the relationship between the composition and properties of HEAs and then employed a genetic algorithm (GA) for multiobjective optimization to identify HEAs with high strength and ductility. Recently, Li et al 27 combined ML models with NSGA-II to design magnetic HEAs with high hardness and selected three candidates for experimental validation. Therefore, multiobjective optimization algorithms may be helpful for the multitargeted property design of HEAs.…”

“…However, the time-consuming and costly limitations of this method have become more apparent as research on HEAs has progressed. Therefore, methods based on phase diagram calculations (CALPHAD), first-principles calculations, and machine learning (ML) , to assist in the design of HEAs with desired properties are gaining importance, where ML-based methods are simpler and more effective than the first two methods . Many ML classification models have been developed to predict the phases of HEAs, − and some ML regression models have also been applied to predict hardness, − strength, − and elastic properties − with satisfactory accuracy.…”

“…Therefore, methods based on phase diagram calculations (CALPHAD), 23 first-principles calculations, 24 and machine learning (ML) 25,26 to assist in the design of HEAs with desired properties are gaining importance, where ML-based methods are simpler and more effective than the first two methods. 27 Many ML classification models have been developed to predict the phases of HEAs, 28−34 and some ML regression models have also been applied to predict hardness, 35−38 strength, 39−42 and elastic properties 43−45 with satisfactory accuracy. These ML models have contributed significantly to the design of new HEAs with outstanding properties.…”

“…For hardness studies, candidates are usually selected from specific HEA systems. 27,35 It is worthwhile to investigate the application of ML in designing HEAs that possess multiple desired properties for practical purposes, but related research is still limited. Menou et al 46,47 used a nondominated sorting genetic algorithm version II 48 (NSGA-II) to simultaneously optimize the strength, solid solution stability, and density of single-phase HEAs.…”

Bayesian optimization of the composition of the lanthanide metal–organic framework MIL-103 for white-light emission

“…30 Bayesian optimization methods are useful tools for yield improvement and composition optimization. [31][32][33][34][35][36][37][38][39][40][41] The use of Bayesian optimization methods for composition optimization could help resolve the problems associated with the complex energy transfer and trace amounts of lanthanide impurities in Ln-MOFs. Bayesian optimization searches for conditions that help achieve white luminescence properties using the amount of rare earth salts in the raw material as a parameter, eliminating the need for the experimenter to directly consider…”