Reproduce “Hard Constraint Projection in a Physics Informed Neural Network (Burgers)” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Miranda J S Horne, Peter K Jimack, Amirul Khan, He Wang
The HCP-PINN adds a fixed local projection that adjusts predictions to satisfy a discretized Burgers’ equation before loss evaluation. In one 1D test with solution labels only on initial and spatial boundaries, its final test data error was more than an order of magnitude below a matched PINN’s. This preliminary result does not establish broader performance or exact continuous-PDE satisfaction.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.