|About this Abstract
||7th World Congress on Integrated Computational Materials Engineering (ICME 2023)
||Machine Learning-enhanced Robust Co-design Exploration for Many Objective, Multilevel Materials Design Problems
||Anand Balu Nellippallil, Mathew Baby, Rashmi Rama Sushil, Palaniappan Ramu, Janet K. Allen, Farrokh Mistree
|On-Site Speaker (Planned)
||Anand Balu Nellippallil
ICME requires seamless integration and exploration of material, product, and manufacturing process design spaces across multilevel. This demands the capability to co-design, that is, share ranged sets of robust design specifications among distributed material, product, and manufacturing process stakeholders through the visualization and exploration of high-dimensional material design space under uncertainty.
In this paper, we present a machine learning-enhanced robust co-design exploration framework by integrating robust compromise decision support problem construct (r-cDSP) with interpretable self-organizing maps (iSOM). Using the framework, we facilitate systematic and efficient visualization, interpretation, and exploration of high-dimensional materials design space under uncertainty. The generic nature of the framework for multidisciplinary designers to a) understand the interactions and capture the dominant process mechanisms that affect materials responses and b) provide decision support for problems involving many conflicting goals under uncertainty is demonstrated using an industry-inspired steel manufacturing process chain problem.