Abstract:Efficient and stable solution of partial differential equations (PDEs) is central to scientific and engineering applications, yet existing numerical solvers rely heavily on matrix based discretizations, while learning based methods require costly training and often suffer from limited generalization. In this work, we proposes a PDE energy driven framework that solves PDEs through physically constrained diffusion iterations, without relying on classical matrix based finite element assembly or data driven neural network training. The proposed method evolves arbitrary random initial fields through PDE energy driven implicit iterations combined with Gaussian smoothing, while strictly enforcing boundary conditions at each iteration. The proposed formulation is applied to representative one dimensional Poisson, Heat, and viscous Burgers equations, covering both steady state and transient problems. Numerical results demonstrate stable convergence to the unique physical solution from random initializations, with accurate resolution of sharp gradients and controlled Mean Squared Error (MSE) across a wide range of discretization parameters. Detailed comparisons with analytical solutions indicate that the framework achieves competitive accuracy and stability. Overall, the proposed framework provides a fast, flexible, and physically consistent alternative to traditional numerical solvers, offering a potential pathway for scalable PDE solutions in both research and engineering applications.
| Subjects: | Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computational Physics (physics.comp-ph) |
| Cite as: | arXiv:2604.25943 [cs.LG] |
| (or arXiv:2604.25943v1 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2604.25943 arXiv-issued DOI via DataCite |
Submission history
From: Bing Yi [view email]
[v1]
Fri, 17 Apr 2026 06:53:03 UTC (3,180 KB)