A.S. Leonov
Abstract:We consider the solution of systems of linear algebraic equations (SLAEs) with an ill-conditioned or degenerate exact matrix and an approximate right-hand side. An approach to solving such a problem is proposed and justified, which makes it possible to improve the conditionality of the SLAE matrix and, as a result, obtain an approximate solution that is stable to perturbations of the right hand side with higher accuracy than using other methods. The approach is implemented by an algorithm that uses so-called minimal pseudoinverse matrices. The results of numerical experiments are presented that confirm the theoretical provisions of the article.
What problem does this paper attempt to address?
The problem that this paper attempts to solve is the problem of solving ill - conditioned systems of linear algebraic equations (SLAEs), especially when the exact matrices of these equations are ill - conditioned or degenerate and the right - hand side terms are approximate. Specifically, the paper proposes a method to improve the condition number of the SLAEs matrix, thereby obtaining an approximate solution with higher accuracy that is stable to perturbations of the right - hand side terms.
### Background and Problem Description of the Paper
In scientific research, it is often necessary to solve very large - scale systems of linear algebraic equations (SLAEs). However, these equations are often ill - conditioned or degenerate, causing classical linear algebraic solution methods to become unreliable due to the accumulation of rounding errors. Although the problem can be partially solved by using higher - precision calculations, this is not applicable to all computing systems, especially those on personal computers.
Therefore, the actual solution of ill - conditioned and degenerate SLAEs requires the use of special methods to combat numerical instability. There are already many methods for dealing with such problems, such as the Tikhonov regularization method, the truncated singular value decomposition (TSVD) method, etc. But the main goal of these methods is not to directly improve the condition number of the matrix, but to improve the stability of the solution by other means.
### Solution Proposed in the Paper
This paper proposes a new method aimed at significantly improving the condition number of the SLAEs matrix. The core idea of this method is to improve the condition number of the original matrix by constructing a special matrix, thereby making the solution process more stable and enabling the acquisition of a higher - accuracy approximate solution.
### Method Overview
The paper proposes an algorithm based on the minimum pseudo - inverse matrix (MPM), which can not only stably calculate the pseudo - inverse matrix but also has several optimal properties. The specific steps include:
1. **Singular Value Decomposition (SVD)**: Perform singular value decomposition on the given matrix \(A_h\).
2. **Select Function \(x_k(h)\)**: Select the function \(x_k(h)\) according to specific conditions, and calculate the corresponding \(\rho_k(\lambda)\) and \(\beta(\lambda)\).
3. **Solve the Equation**: Find the solution \(\lambda(h)\) that satisfies \(\beta(\lambda)=h^2\).
4. **Construct Regularized Singular Value Matrix**: Construct the regularized singular value matrix \(\tilde{R}_h\) and its pseudo - inverse \(\tilde{R}_h^+\) according to \(\lambda(h)\).
5. **Calculate the Final Result**: Use the above matrices to calculate the minimum pseudo - inverse matrix \(\tilde{A}_h^+\), and obtain the approximate solution \(z_\delta = \tilde{A}_h^+\ u_\delta\).
### Numerical Experiments and Conclusions
Through numerical experiments, the paper verifies the effectiveness of the proposed method. The experimental results show that, compared with other methods (such as TSVD and Tikhonov regularization), the MPMI method can provide higher solution accuracy at different error levels and significantly reduce the condition number of the matrix. For example, in some cases, the MPMI method can reduce the condition number to less than half of the condition number of the original matrix.
### Summary
The paper proposes a new method for solving ill - conditioned systems of linear algebraic equations by improving the matrix condition number. This method performs well in terms of numerical stability and solution accuracy. By introducing the concept of the minimum pseudo - inverse matrix, the paper not only provides a theoretical basis but also verifies the effectiveness of the method through numerical experiments.