NEUP CINR-25-34261: Integrating Graph Neural Operator with MOOSE Framework for Enhanced Transient Modeling and Real-Time Data Assimilation

We introduce the GNO-MOOSE hybrid solver, which combines Graph Neural Operators with MOOSE's automatic differentiation features to enhance computational efficiency without sacrificing accuracy. This integration reduces computational time by allowing larger timestep sizes and coarser meshes, and it enables real-time data assimilation through coupling with an Ensemble Kalman Filter. Our approach aims to make multi-dimensional transient simulations feasible in real time.

Additional Info

Field	Value
Awarded Institution	Texas A&M University
PI	Yang Liu