Abstract:In this paper, a new block preconditioner is proposed for the saddle point problem arising from the Neumann boundary control problem. In order to deal with the singularity of the stiffness matrix, the saddle point problem is first extended to a new one by a regularization of the pure Neumann problem. Then after row permutations of the extended saddle point problem, a new block triangular preconditioner is constructed based on an approximation of the Schur complement. We analyze the eigenvalue properties of the preconditioned matrix and provide eigenvalue bounds. Numerical results illustrate the efficiency of the proposed preconditioning method.
What problem does this paper attempt to address?
### What problem does this paper attempt to solve?
This paper aims to solve the pre - processing methods for saddle - point problems related to pure Neumann boundary control problems. Specifically, the paper has conducted research on the following issues:
1. **Singularity of saddle - point problems**: Due to the singular nature of the stiffness matrix \(K\), traditional pre - processing methods face challenges when dealing with saddle - point problems. To overcome this problem, the author introduced a new block pre - processing method to expand the saddle - point problem through regularization and row permutation.
2. **Construction of efficient pre - processing methods**: The paper proposed a new block triangular pre - processing method based on Schur complement approximation, analyzed the eigenvalue properties of the pre - processing matrix, and provided the eigenvalue range. In addition, numerical results were also presented to prove the effectiveness of the proposed method.
3. **Solving PDE - constrained optimization problems**: The paper paid special attention to how to solve optimization problems constrained by PDEs more effectively, especially control problems with Neumann boundary conditions. By improving the pre - processing technique, the efficiency of the Krylov subspace iteration method can be improved.
#### Specific problem description
The paper considers the boundary control problem constrained by the Poisson equation with pure Neumann boundary conditions, which has the form:
\[
\min_{y, u} \frac{1}{2}\|y - y_d\|^2_{L^2(\Omega)}+\frac{\beta}{2}\|u\|^2_{L^2(\partial \Omega)},
\]
subject to
\[
\begin{cases}
-\nabla^2 y = f & \text{in } \Omega, \\
\frac{\partial y}{\partial n}=u & \text{on } \partial \Omega,
\end{cases}
\]
where \( \Omega \subset \mathbb{R}^2\), \(y\) is the state variable, \(u\) is the control variable, \(f\) is the known source term, \(y_d\) is the desired state, and \( \beta\) is the regularization parameter.
#### Solutions
1. **Regularization and expansion of saddle - point problems**: The original problem is regularized by introducing the global condition \( \int_\Omega y_0 \, d\Omega = 0\), thus obtaining an extended saddle - point problem, where the extended stiffness matrix \(K_e\) is non - singular.
2. **Block triangular pre - processing method**: A new block triangular pre - processing matrix \( \hat{P}_2\) is constructed by performing row permutation on the extended saddle - point problem and based on Schur complement approximation.
3. **Eigenvalue analysis**: The spectral characteristics of the pre - processing matrix \( \hat{P}_2^{-1}A\) are analyzed, and the eigenvalue range is provided, indicating that the new method has good performance under different mesh sizes and regularization parameters.
4. **Numerical experiments**: The effectiveness of the proposed pre - processing method is verified through numerical experiments, demonstrating its superior performance in solving pure Neumann boundary control problems.
In summary, this paper solves the singularity problem of saddle - point problems in pure Neumann boundary control problems and improves the solution efficiency through the introduction of new pre - processing techniques and theoretical analysis.