Search papers, labs, and topics across Lattice.
The INI-VPINN framework introduces a novel weak-form Physics-Informed Neural Network that effectively integrates Neumann boundary and interface conditions without requiring additional loss terms or multiple subdomain networks. By utilizing compact support weighting functions and integration by parts, it ensures physical consistency across material boundaries while addressing Poisson and Laplace problems with complex geometries. Experimental results demonstrate that INI-VPINN outperforms existing Physics-Informed Neural Network formulations in terms of accuracy, convergence speed, and smoothness.
Achieving higher accuracy and faster convergence in multi-material problems without the complexity of additional loss terms could revolutionize how we approach neural network training for physics-based simulations.
We propose a new weak-form Physics-Informed Neural Network approach (named INI-VPINN). INI-VPINN naturally incorporates Neumann boundary and interface conditions into the variational formulation. It removes the need for additional loss terms or multiple subdomain networks. This framework employs compact support weighting functions and integration by parts to implicitly impose flux and continuity constraints. In this way, it implicitly ensures physical consistency across material boundaries. The proposed method is tested on Poisson and Laplace problems with sharp interfaces and complex geometries. Results show that, compared with several other Physics Informed Neural Networks-based formulations, the INI-VPINN consistently achieves higher accuracy, smoother and faster convergence. The proposed framework provides a general approach for solving multimaterial problems with complex geometries and mixed Neumann-Dirichlet boundary conditions using neural networks. The implementation is publicly available in a GitHub repository.
