← Papers
Operators

PI-DeepONet

Physics-informed DeepONets for parametric PDEs

Paper title Learning the solution operator of parametric partial differential equations with physics-informed DeepOnets

Wang et al. (University of Pennsylvania) · General PDEs · Public: code, data

Paper arXiv 2103.10974PDF

Trains a DeepONet with the residual of the governing equation, computed by automatic differentiation at collocation points, added to the loss as a soft penalty. With only initial and boundary conditions given, the network learns the map from an input function (a source term, an initial condition, a shape) to the solution without any solved examples.

Published
Science Advances

Share on LinkedIn, X or Bluesky, or or .

Know related work that belongs here?

Something wrong on this page?

More papers

All papers
  1. OperatorsTurtLES — A 3D turbulent-wake benchmark where operators fail over long rollouts
  2. OperatorsFunction-space certification — Domain-wide guarantees for neural operators, not only at grid points
  3. OperatorsL-SplineFNO — Learnable splines take the FNO to unstructured meshes