Normwise scaling of second order polynomial matrices

被引:48
作者
Fan, HY [1 ]
Lin, WW
Van Dooren, P
机构
[1] Natl Tsing Hua Univ, Dept Math, Hsinchu 300, Taiwan
[2] Univ Catholique Louvain, Dept Engn Math, B-1348 Louvain, Belgium
关键词
generalized eigenvalues; QZ algorithm; balancing;
D O I
10.1137/S0895479803434914
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
We propose a minimax scaling procedure for second order polynomial matrices that aims to minimize the backward errors incurred in solving a particular linearized generalized eigenvalue problem. We give numerical examples to illustrate that it can significantly improve the backward errors of the computed eigenvalue-eigenvector pairs.
引用
收藏
页码:252 / 256
页数:5
相关论文
共 4 条
[1]  
Lemonnier D., 2003, P SIAM APPL LIN ALG
[2]   Backward error and condition of polynomial eigenvalue problems [J].
Tisseur, F .
LINEAR ALGEBRA AND ITS APPLICATIONS, 2000, 309 (1-3) :339-361
[3]   The quadratic eigenvalue problem [J].
Tisseur, F ;
Meerbergen, K .
SIAM REVIEW, 2001, 43 (02) :235-286
[4]   THE EIGENSTRUCTURE OF AN ARBITRARY POLYNOMIAL MATRIX - COMPUTATIONAL ASPECTS [J].
VANDOOREN, P ;
DEWILDE, P .
LINEAR ALGEBRA AND ITS APPLICATIONS, 1983, 50 (APR) :545-579