On using approximate finite differences in matrix-free Newton-Krylov methods

被引:13
作者
Brown, Peter N. [1 ]
Walker, Homer F. [2 ]
Wasyk, Rebecca [2 ]
Woodward, Carol S. [1 ]
机构
[1] Lawrence Livermore Natl Lab, Ctr Appl Sci Comp, Livermore, CA 94551 USA
[2] Worcester Polytech Inst, Dept Math Sci, Worcester, MA 01609 USA
关键词
matrix-free Newton-Krylov methods; Newton-GMRES methods; Newton-Arnoldi methods; Newton's method; Krylov subspace methods; GMRES method; Arnoldi method;
D O I
10.1137/060652749
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
A Newton-Krylov method is an implementation of Newton's method in which a Krylov subspace method is used to solve approximately the linear systems that characterize steps of Newton's method. Newton-Krylov methods are often implemented in "matrix-free" form, in which the Jacobian-vector products required by the Krylov solver are approximated by finite differences. Here we consider using approximate function values in these finite differences. We first formulate a finite-difference Arnoldi process that uses approximate function values. We then outline a Newton Krylov method that uses an implementation of the GMRES or Arnoldi method based on this process, and we develop a local convergence analysis for it, giving sufficient conditions on the approximate function values for desirable local convergence properties to hold. We conclude with numerical experiments involving particular function-value approximations suitable for nonlinear diffusion problems. For this case, conditions are given for meeting the convergence assumptions for both lagging and linearizing the nonlinearity in the function evaluation.
引用
收藏
页码:1892 / 1911
页数:20
相关论文
共 19 条
[1]  
[Anonymous], SERIES AUTOMATIC COM
[3]   HYBRID KRYLOV METHODS FOR NONLINEAR-SYSTEMS OF EQUATIONS [J].
BROWN, PN ;
SAAD, Y .
SIAM JOURNAL ON SCIENTIFIC AND STATISTICAL COMPUTING, 1990, 11 (03) :450-481
[5]   Inexact perturbed Newton methods and applications to a class of Krylov solvers [J].
Catinas, E .
JOURNAL OF OPTIMIZATION THEORY AND APPLICATIONS, 2001, 108 (03) :543-570
[6]   INEXACT NEWTON METHODS [J].
DEMBO, RS ;
EISENSTAT, SC ;
STEIHAUG, T .
SIAM JOURNAL ON NUMERICAL ANALYSIS, 1982, 19 (02) :400-408
[7]   Choosing the forcing terms in an inexact Newton method [J].
Eisenstat, SC ;
Walker, HF .
SIAM JOURNAL ON SCIENTIFIC COMPUTING, 1996, 17 (01) :16-32
[8]  
Freund RW., 1992, ACTA NUMER, V1, P57
[9]  
Golub G. H., 1996, MATRIX COMPUTATIONS
[10]   SUNDIALS: Suite of nonlinear and differential/algebraic equation solvers [J].
Hindmarsh, AC ;
Brown, PN ;
Grant, KE ;
Lee, SL ;
Serban, R ;
Shumaker, DE ;
Woodward, CS .
ACM TRANSACTIONS ON MATHEMATICAL SOFTWARE, 2005, 31 (03) :363-396