A Riccati-type algorithm for solving generalized Hermitian eigenvalue problems

A Riccati-type algorithm for solving generalized Hermitian eigenvalue problems
复制标题

求解广义埃尔米特特征值问题的 Riccati 型算法

DOI:
10.1007/s11227-020-03331-w
复制
发表时间:
2021
期刊:
The Journal of Supercomputing
影响因子:
--
通讯作者:
Takafumi Miyata
Takafumi Miyata
中科院分区:
--
文献类型:
--
作者:
K. Fukazawa;Y. Katoh;T. Nanri;Y. Miyake;Takafumi Miyata

文献摘要

参考文献

相似文献

本文提出了一种快速求解广义厄米特特征值问题的启发式算法。该算法在子空间中搜索问题的近似解。如果近似解不可接受,则将子空间扩展到更大的子空间,然后在扩展的子空间中计算可能更好的近似解。该算法交替迭代这两个步骤。因此,算法的收敛速度取决于如何生成子空间。本文导出了一个Riccati方程,它的解可以将广义Hermitian特征值问题的近似解校正为精确解。换句话说,如果一个子空间被Riccati方程的解展开,则可以找到特征值问题的解。这是现有算法如MATLAB中实现的Krylov子空间算法和Jacobi-Davidson算法所不具有的特征。然而,类似于求解特征值问题,求解Riccati方程是耗时的。我们考虑求解低精度的Riccati方程,并利用其近似解展开一个子空间。讨论了该启发式算法的实现,以节省算法的计算量。实验结果表明,与现有算法相比,该启发式算法在较少的迭代次数内收敛,因而所需的计算时间较少。
The paper describes a heuristic algorithm for solving a generalized Hermitian eigenvalue problem fast. The algorithm searches a subspace for an approximate solution of the problem. If the approximate solution is unacceptable, the subspace is expanded to a larger one, and then, in the expanded subspace a possibly better approximated solution is computed. The algorithm iterates these two steps alternately. Thus, the speed of the convergence of the algorithm depends on how to generate a subspace. In this paper, we derive a Riccati equation whose solution can correct the approximate solution of a generalized Hermitian eigenvalue problem to the exact one. In other words, the solution of the eigenvalue problem can be found if a subspace is expanded by the solution of the Riccati equation. This is a feature the existing algorithms such as the Krylov subspace algorithm implemented in the MATLAB and the Jacobi–Davidson algorithm do not have. However, similar to solving the eigenvalue problem, solving the Riccati equation is time-consuming. We consider solving the Riccati equation with low accuracy and use its approximate solution to expand a subspace. The implementation of this heuristic algorithm is discussed so that the computational cost of the algorithm can be saved. Some experimental results show that the heuristic algorithm converges within fewer iterations and thus requires lesser computational time comparing with the existing algorithms.
特征值问题的基于修正的迭代方法
DOI: 10.1587/transfun.e101.a.1668
发表时间: 2018
期刊: IEICE Transactions on Fundamentals of Electronics, Communications and Computer Sciences
影响因子: --
作者:
永野 浩大;鳥生 大祐;牛島 省;Takafumi Miyata
通讯作者: Takafumi Miyata
解决特征问题:从阿诺尔迪到雅可比-戴维森到里卡蒂方法
DOI: 10.1007/3-540-36487-0_18
发表时间: 2002
期刊: Comput. J.
影响因子: --
作者:
J. Brandts
通讯作者: J. Brandts
DOI: 10.1093/comjnl/4.4.332
发表时间: 1962-01-01
期刊: COMPUTER JOURNAL
影响因子: 1.4
作者:
FRANCIS, JGF
通讯作者: FRANCIS, JGF
计算波导横截面中选定本征模的改进雅可比-戴维森方法
DOI: 10.1109/tmag.2010.2046315
发表时间: 2010
影响因子: 2.1
作者:
B. Bandlow;D. Sievers;R. Schuhmann
通讯作者: R. Schuhmann
DOI: 10.1016/0041-5553(63)90168-x
发表时间: 1962
期刊: Ussr Computational Mathematics and Mathematical Physics
影响因子: --
作者:
V. Kublanovskaya
通讯作者: V. Kublanovskaya