Domain Decomposition for Data-Driven Reduced Modeling of Large-Scale Systems

What is it about?

Constructing reduced models of complex real-world problems from data is challenging in practice, often requiring powerful computers with large memory. To address these challenges, we introduce a domain decomposition formulation into the construction of a data-driven reduced model to increase the accuracy of the domain-decomposed reduced model and decrease the overall computational cost by processing the large data set into more manageable subdomains. Our method demonstrates improved prediction accuracy and decreased computational demands in a large-scale, complex simulation involving rotating detonation rocket engines.

Featured Image