Abstract:We present a reduced basis method for cheaply constructing (possibly rough) approximations to the nodal basis functions of the virtual element space, and propose to use such approximations for the design of the stabilization term in the virtual element method and for the post-processing of the solution.
What problem does this paper attempt to address?
This paper attempts to solve some key problems in the Virtual Element Method (VEM), especially the limitations of VEM in dealing with anisotropic problems and the post - processing of discrete solutions. Specifically:
1. **Dealing with anisotropic problems**: The stabilization term in VEM is isotropic in nature, which may lead to inaccurate results or performance degradation when dealing with anisotropic problems. To solve this problem, the author proposes a method based on the Reduced Basis Method (RBM) to approximate local basis functions in order to design more effective stabilization terms.
2. **Post - processing of discrete solutions**: Although the numerical solution of VEM belongs to a continuous discrete space, it is usually not directly accessible unless the PDE is solved in each polygonal cell. Therefore, in practical applications, only a discontinuous polynomial space projection can be obtained, which is an obstacle for some tasks (such as the design of geometric multigrid preconditioners). To overcome this limitation, the author proposes to use RBM to locally reconstruct virtual basis functions, so that a continuous discrete solution can be obtained.
### Main contributions
- **Application of the Reduced Basis Method**: By taking the shape of the polygonal cell as a parameter, the non - polynomial part of the virtual element function is efficiently reconstructed using the Reduced Basis Method. This method can provide approximate values of local basis functions when needed and has a low computational cost.
- **Design of stabilization terms**: For anisotropic problems, traditional stabilization methods may perform poorly. The author proposes to use the virtual basis functions reconstructed by RBM to design better stabilization terms, which can significantly improve the performance of the method even in the case of low - precision reconstruction.
- **Post - processing techniques**: By reconstructing virtual basis functions with RBM, a continuous discrete solution can be obtained, so that tasks such as visualization, point - value evaluation and error estimation can be better carried out.
### Method overview
- **Parameterized equations**: The local Laplace equation defining the virtual basis functions is rewritten as a parameterized equation on the reference cell, where the parameter is the shape of the polygonal cell.
- **Reduced Basis Method**: The Reduced Basis Method is used to efficiently solve these parameterized equations and generate approximate basis functions for stabilization and post - processing.
- **Implementation and verification**: The effectiveness of the RBM - VEM method is verified by numerical experiments, especially its advantages in dealing with anisotropic problems and post - processing.
In conclusion, this paper aims to improve the stabilization and post - processing of VEM by introducing the Reduced Basis Method, so as to better deal with complex geometric shapes and anisotropic problems.