The ML-based framework for the MOO design of RHEAs. (IMAGE)
Caption
(a) The workflow contains three parts: ML model selection to calculate the expected improvement (EI) values of the target properties for a given alloy; the non-dominated sorting genetic algorithm (NSGA)-II to evolutionarily search for candidate alloy compositions based on EI values; and experimental feedback including alloy selection by cluster analysis and experimental verification. The EI utility indicator is used as the objective in the genetic search rather than the ML-predicted values. (b) Sketch map of a comparison of materials selection using the EI indicator with uncertainty considerations and direct ML prediction. (c) For the two objective RHEA designs, a clustering-based selector is used to obtain potential alloys on the PF, which leads to possible improvement in the HT strength in comparison with TaNbHfZrTi and greater RT ductility than that of NbMoTaW(V).
Credit
Cheng Wen et al.
Usage Restrictions
Credit must be given to the creator. Only noncommercial uses of the work are permitted. No derivatives or adaptations of the work are permitted.
License
CC BY-NC-ND