arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2602.14449math.NAcs.NA
收藏

关于两阶段Householder正交化

On Two-Stage Householder Orthogonalization

School of Data Science, Fudan University(复旦大学数据科学学院);MOE Key Laboratory for Computational Physical Sciences, Fudan University(复旦大学计算物理科学教育部重点实验室)

Zhuang-Ao He, Meiyue Shao

AI总结:

本文提出一种基于广义Householder变换的两阶段Householder正交化算法,解决了传统方法对矩阵条件数依赖的问题,通过仅正交化V的平方子矩阵提升了稳定性。

摘要:

两阶段正交化在数值算法如Krylov子空间方法中至关重要。我们需要将矩阵A与具有正交列的矩阵V正交化。常用方法是使用块Gram-Schmidt算法。然而,其稳定性很大程度上依赖于[V,A]的条件数。在对[V,A]执行Householder正交化时,稳定性无条件成立,但并未利用V具有正交列的知识。为了解决这些问题,我们提出了一种基于广义Householder变换的两阶段Householder正交化算法。与传统方法不同,我们的算法不需要显式地正交化整个V,只需正交化V的一个平方子矩阵。理论分析和数值实验表明,我们的方法也是无条件稳定的。

英文摘要:

Two-stage orthogonalization is essential in numerical algorithms such as Krylov subspace methods. For this task we need to orthogonalize a matrix $A$ against another matrix $V$ with orthonormal columns. A common approach is to employ the block Gram--Schmidt algorithm. However, its stability largely depends on the condition number of $[V,A]$. While performing a Householder orthogonalization on $[V,A]$ is unconditionally stable, it does not utilize the knowledge that $V$ has orthonormal columns. To address these issues, we propose a two-stage Householder orthogonalization algorithm based on the generalized Householder transformation. Instead of explicitly orthogonalizing the entire $V$, our algorithm only needs to orthogonalizes a square submatrix of $V$. Theoretical analysis and numerical experiments demonstrate that our method is also unconditionally stable.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑