Recycling Krylov subspaces for sequences of linear systems

被引:244
作者
Parks, Michael L.
De Sturler, Eric
Mackey, Greg
Johnson, Duane D.
Maiti, Spandan
机构
[1] Univ Illinois, Dept Comp Sci, Urbana, IL 61801 USA
[2] Sandia Natl Labs, Albuquerque, NM 87185 USA
[3] Virginia Tech, Dept Math, Blacksburg, VA 24061 USA
[4] Univ Illinois, Dept Mat Sci, Urbana, IL 61801 USA
[5] Univ Illinois, Dept Aerosp Engn, Urbana, IL 61801 USA
关键词
sequence of linear systems; linear solvers; Krylov methods; truncation; restarting; Krylov subspace recycling; iterative methods;
D O I
10.1137/040607277
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
Many problems in science and engineering require the solution of a long sequence of slowly changing linear systems. We propose and analyze two methods that significantly reduce the total number of matrix-vector products required to solve all systems. We consider the general case where both the matrix and right-hand side change, and we make no assumptions regarding the change in the right-hand sides. Furthermore, we consider general nonsingular matrices, and we do not assume that all matrices are pairwise close or that the sequence of matrices converges to a particular matrix. Our methods work well under these general assumptions, and hence form a significant advancement with respect to related work in this area. We can reduce the cost of solving subsequent systems in the sequence by recycling selected subspaces generated for previous systems. We consider two approaches that allow for the continuous improvement of the recycled subspace at low cost. We consider both Hermitian and non-Hermitian problems, and we analyze our algorithms both theoretically and numerically to illustrate the effects of subspace recycling. We also demonstrate the effectiveness of our algorithms for a range of applications from computational mechanics, materials science, and computational physics.
引用
收藏
页码:1651 / 1674
页数:24
相关论文
共 39 条
[1]   Convergence of restarted Krylov subspaces to invariant subspaces [J].
Beattie, C ;
Embree, M ;
Rossi, J .
SIAM JOURNAL ON MATRIX ANALYSIS AND APPLICATIONS, 2003, 25 (04) :1074-1109
[2]  
Boisvert RF, 1997, QUALITY OF NUMERICAL SOFTWARE - ASSESSMENT AND ENHANCEMENT, P125
[3]   Galerkin projection methods for solving multiple linear systems [J].
Chan, TF ;
Ng, MK .
SIAM JOURNAL ON SCIENTIFIC COMPUTING, 1999, 21 (03) :836-850
[4]  
Creutz M., 1986, QUARKS GLUONS LATTIC
[5]   Truncation strategies for optimal Krylov subspace methods [J].
De Sturler, E .
SIAM JOURNAL ON NUMERICAL ANALYSIS, 1999, 36 (03) :864-889
[6]  
de Sturler E., 1996, HOUS S 13 P HOUS S N, P193
[7]   Nested Krylov methods based on GCR [J].
deSturler, E .
JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS, 1996, 67 (01) :15-41
[8]  
DONGARRA JJ, 1998, SOFTWATE ENV TOOLS, V7
[9]   Analysis of acceleration strategies for restarted minimal residual methods [J].
Eiermann, M ;
Ernst, OG ;
Schneider, O .
JOURNAL OF COMPUTATIONAL AND APPLIED MATHEMATICS, 2000, 123 (1-2) :261-292
[10]  
Farhat C., 1994, Computational Mechanics Advances, V2, P1