Controlled formulation study of ψ – p vs ψ – ω Physics-Informed Neural Networks (PINNs) for high-Re ( Re = 1000) incompressible Navier-Stokes. Investigates operator conditioning, false convergence, and exact hard BCs.
deep-learning pytorch scientific-computing differential-equations navier-stokes fluid-dynamics computational-fluid-dynamics pinn fluid-mechanics lid-driven-cavity scientific-machine-learning sciml physics-informed-neural-networks streamfunction-vorticity high-reynolds-number operator-conditioning
-
Updated
Sep 3, 2026 - Python