Structure-preserving Kernel-based methods for solving dissipative PDEs on surfaces

Zhengjie Sun,Leevan Ling,Meng Chen
DOI: https://doi.org/10.48550/arXiv.2312.17478
2023-12-29
Abstract:In this paper, we propose a general meshless structure-preserving Galerkin method for solving dissipative PDEs on surfaces. By posing the PDE in the variational formulation and simulating the solution in the finite-dimensional approximation space spanned by (local) Lagrange functions generated with positive definite kernels, we obtain a semi-discrete Galerkin equation that inherits the energy dissipation property. The fully-discrete structure-preserving scheme is derived with the average vector field method. We provide a convergence analysis of the proposed method for the Allen-Cahn equation. The numerical experiments also verify the theoretical analysis including the convergence order and structure-preserving properties.
Numerical Analysis
What problem does this paper attempt to address?
### Problems the paper attempts to solve This paper aims to solve the problem of how to preserve the intrinsic structural properties, especially the energy dissipation property, when solving dissipative partial differential equations (PDEs) on surfaces. Specifically: 1. **Problem background**: - Partial differential equations (PDEs) usually have intrinsic properties related to geometric structures (such as symplectic structures, multisymplectic structures, Lie group structures, symmetries, etc.) or physical principles (such as energy conservation, energy dissipation, momentum conservation, positivity, etc.). - Preserving these intrinsic features is crucial for designing more accurate and stable numerical methods. - For dissipative PDEs (such as the gradient flow model), the energy dissipation property is particularly important. 2. **Research motivation**: - PDEs on surfaces describe a large number of phenomena in many applied science fields, but due to their complexity, analytical solutions are often not feasible. - Therefore, it is necessary to develop suitable numerical methods to approximately solve these problems. - Existing numerical methods are mainly divided into three types: intrinsic methods, embedding methods, and extrinsic methods, but these methods may overlook the inherent properties of PDEs. 3. **Paper goals**: - Propose a mesh - free structure - preserving Galerkin method based on kernel functions for solving general dissipative PDEs on surfaces. - This method can not only handle PDEs on surfaces flexibly and efficiently, but also preserve the energy dissipation properties of PDEs. - Verify the convergence and structure - preserving properties of the proposed method through numerical experiments. ### Main contributions - **Theoretical framework**: Proposed a semi - discrete Galerkin equation that inherits the energy dissipation property, and derived a fully - discrete structure - preserving scheme using the average vector field method. - **Numerical verification**: Provided a convergence analysis of the Allen - Cahn equation, and verified the theoretical analysis results through numerical experiments, including the convergence order and structure - preserving properties. - **Practical applications**: Demonstrated the potential application value of this method in multiple fields such as phase change, multiphase fluid flow, crystal growth, and image analysis. Through these works, the paper provides a new and effective numerical method for solving dissipative PDEs on surfaces, especially suitable for application scenarios where the energy dissipation property needs to be preserved.