Extended Interface Physics-Informed Neural Networks Method for Moving Interface Problems
Ran Bi, Weibing Deng, Yameng Zhu
arXiv:2508.01463·math.NA·Published 2025-08-02
Physics-Informed Neural Networks (PINNs) have emerged as a powerful class of mesh-free numerical methods for solving partial differential equations (PDEs), particularly those involving complex geometries. In this work, we present an innovative Extended Interface Physics-Informed Neural Network (XI-PINN) framework specifically designed to solve parabolic moving interface problems. The proposed approach incorporates a level set function to characterize the interface, which can be obtained either directly or through a neural network solution. We conduct a rigorous a priori error analysis for the XI-PINN method, providing error bounds for the approximation. Leveraging the Neural Tangent Kernel (NTK) theory, we further demonstrate that XI-PINN achieves a faster training convergence rate compared to conventional PINN approaches. The method's versatility is further demonstrated by its application to the Oseen equations. We perform comprehensive numerical experiments to validate the efficacy, accuracy, and robustness of the proposed framework.
TopicsScientific Machine Learning & PINNs
Tagspartial-differential-equations physics-informed-neural-networks pinns
arXiv categoriesmath.NA
arXiv abstract pagePDF