James H. Adler,Xiaozhe Hu,Xue Wang,Zhongqin Xue
Abstract:When developing robust preconditioners for multiphysics problems, fractional functions of the Laplace operator often arise and need to be inverted. Rational approximation in the uniform norm can be used to convert inverting those fractional operators into inverting a series of shifted Laplace operators. Care must be taken in the approximation so that the shifted Laplace operators remain symmetric positive definite, making them better conditioned. In this work, we study two greedy algorithms for finding rational approximations to such fractional operators. The first algorithm improves the orthogonal greedy algorithm discussed in [Li et al., SISC, 2024] by adding one minimization step in the uniform norm to the procedure. The second approach employs the weak Chebyshev greedy algorithm in the uniform norm. Both methods yield non-increasing error. Numerical results confirm the effectiveness of our proposed algorithms, which are also flexible and applicable to other approximation problems. Moreover, with effective rational approximations to the fractional operator, the resulting algorithms show good performance in preconditioning a Darcy-Stokes coupled problem.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is: in multi - physical - field problems, especially when dealing with fractional - order operators, how to perform rational approximation efficiently and accurately. Specifically, when developing robust preconditioners for multi - physical - field problems, it is often necessary to solve the inverse of the fractional - order Laplace operator. By using greedy algorithms for rational approximation under the uniform norm, the problem of solving the inverse of the fractional - order operator can be transformed into solving the inverses of a series of shifted Laplace operators. This not only improves the computational efficiency but also ensures that these shifted Laplace operators remain symmetric positive definite (SPD), thereby improving the condition number of the matrix.
### Main Research Contents
1. **Problem Background**:
- In multi - physical - field problems, the coupling interface conditions are usually imposed in a weak form through Lagrange multipliers.
- Effective preconditioners sometimes involve solving fractional - order or weighted Laplace operators, for example, a diagonal block in the Darcy - Stokes coupling problem:
\[
S=\mu^{-1}(-\Delta + I)^{-1/2}+K\mu^{-1}(-\Delta + I)^{1/2}
\]
- Directly solving this diagonal block will lead to low computational efficiency and increased storage requirements, so a rational approximation method is required.
2. **Importance of Rational Approximation**:
- Rational approximation can represent complex functions in a compact form, especially when the function has singularities.
- Through rational approximation, the inversion of the fractional - order operator can be transformed into the inversions of a series of shifted Laplace operators, thus maintaining symmetric positive definiteness and improving computational efficiency.
3. **Improved Greedy Algorithms**:
- Two greedy algorithms are proposed to find the rational approximation of the fractional - order operator:
1. **Improved Orthogonal Greedy Algorithm (OGA)**: An additional step of minimizing the uniform norm error is added on the basis of the original OGA.
2. **Weak Chebyshev Greedy Algorithm (WCGA)**: WCGA is applied under the uniform norm.
- Both of these algorithms can ensure that the uniform norm error is monotonically non - increasing.
4. **Numerical Experiments**:
- The effectiveness and flexibility of the proposed greedy algorithms are verified through several basic tests.
- These algorithms are applied to construct preconditioners for the Darcy - Stokes coupling problem, showing good performance.
### Summary
By proposing and improving two greedy algorithms and performing rational approximation under the uniform norm, this paper solves the difficult problem of inverting fractional - order operators in multi - physical - field problems, significantly improving the computational efficiency and the performance of preconditioners.