Artificial Intelligence Designer of Materials and Processes for Advanced Power Generation

Abstract

Motivation for this research is the need to accelerate design of high-performance materials and processes to be used in advanced fossil energy power plants and, by doing so, bridge the gap between the shortening technological cycles and the long qualification testing of new alloys for energy applications. The artificial intelligence can exploit causal graph neural networks and other advanced network architectures to represent domain knowledge and engineering constraints. In this presentation, ‘deep-freeze’ graphs, ‘convoluted filtering’ networks, ‘mirror-image’ graphs, and adversarial ensemble methods are utilized to support inversion modeling for optimization of the complex compositions and complex processes in design of high-performing alloys, with their properties tailored to the energy application specifications.