Stationary splitting iterative methods for the matrix equation A X B = C

Zhongyun Liu,Zhen Li,Carla Ferreira,Yulin Zhang

DOI: https://doi.org/10.1016/j.amc.2020.125195

IF: 4.397

2020-08-01

Applied Mathematics and Computation

Abstract:<p>Stationary splitting iterative methods for solving <span class="math"><math>AXB=C</math></span> are considered in this paper. The main tool to derive our new method is the induced splitting of a given nonsingular matrix <span class="math"><math>A=M−N</math></span> by a matrix <span class="math"><math>H</math></span> such that <span class="math"><math>(I−H)−1</math></span> exists. Convergence properties of the proposed method are discussed and numerical experiments are presented to illustrate its computational efficiency and the effectiveness of some preconditioned variants. In particular, for certain surface fitting applications our method is much more efficient than the <strong>progressive iterative approximation</strong> (PIA), a conventional iterative method often used in computer aided geometric design (CAGD).</p>

mathematics, applied

What problem does this paper attempt to address?

The problem that this paper attempts to solve is the solution of the matrix equation \( \textbf{AXB} = \textbf{C} \). Specifically, the paper focuses on how to efficiently solve such matrix equations by introducing the splitting - iteration method. This type of matrix equations plays an important role in many practical applications such as Computer - Aided Geometric Design (CAGD), signal and image processing, and photogrammetry. Given the existence of matrix equations and their wide application in many fields, researchers have studied the theory and algorithms of matrix equations for decades.

The paper proposes a new iteration method. This method is based on the induced splitting of a given non - singular matrix \( \textbf{A} = \textbf{M} - \textbf{N} \), and makes \( (\textbf{I} - \textbf{H})^{-1} \) exist through the matrix \( \textbf{H} \). The paper discusses the convergence properties of the proposed method and shows its computational efficiency and the effectiveness of some pre - processing variants through numerical experiments. In particular, in some surface - fitting applications, the proposed method is more efficient than the traditional iteration method - Progressive Iterative Approximation (PIA).

The main contributions of the paper are as follows:
1. Propose a new iteration method based on induced splitting to directly solve from the original matrix equation \( \textbf{AXB} = \textbf{C} \).
2. Discuss the convergence properties of this method and give sufficient conditions for convergence.
3. Verify the effectiveness and computational efficiency of the method through numerical experiments, especially its performance when the matrix scale is large.
4. Explore the application of pre - processing techniques to further improve the performance of the method.

These works not only enrich the theoretical basis for solving matrix equations but also provide efficient computational tools for practical applications.