Coupling-Robust Accuracy in Multiphysics Physics Informed Neural Networks via Kronecker-Preconditioned Optimization
Researchers have developed a new optimization technique called SOAP+GN to improve the accuracy of physics-informed neural networks (PINNs) when dealing with complex, coupled multiphysics systems. This method addresses a known issue where PINN accuracy degrades as the inter-equation coupling strengthens. By employing Kronecker-preconditioned optimization and inverse-gradient-norm loss balancing, SOAP+GN demonstrates robust accuracy across numerous experiments, even in challenging 2D systems that previously overwhelmed standard optimization methods like Adam+GN. AI
IMPACT Introduces a novel optimization method that significantly enhances the performance and applicability of physics-informed neural networks in complex multiphysics simulations.