AUTOMATIC SCALING PROCEDURE FOR A DYAKANOV-GUNN ITERATION SCHEME

被引：9

作者：

BANK, RE

机构：

[1] Department of Mathematics The University of Texas at Austin, Austin

来源：

LINEAR ALGEBRA AND ITS APPLICATIONS | 1979年 / 28卷 / DEC期

关键词：

D O I：

10.1016/0024-3795(79)90114-9

中图分类号：

O29 [应用数学];

学科分类号：

070104 ;

摘要：

An automatic computational procedure for scaling a D'Yakanov-Gunn iteration for nonseparable elliptic equations is described. This method is used in conjunction with a preconditioned conjugate gradient iteration employing a nearby separable approximation of the scaled nonseparable matrix. This results in an operation count of O(n2log( n k)log( 1 ε)) to reduce the initial error by a factor of ε. © 1979.

引用

页码：17 / 33

页数：17

共 29 条

[1]

AUMANN G, 1958, BAYER AKAD WISS M SB, P137

[2]

AUMANN G, 1959, BAYER AKAD WISS M SB, P125

[3]

AXELSSON O, 1974, CERN7410

[4]

AXELSSON O, 1977, LECTURE NOTES MATH