Atlas Mathematical Conference Abstracts || Conferences | Abstracts | for Organizers | About AMCA

5th IMACS Conference on Iterative Methods in Scientific Computing
May 28-31, 2001
Foundation for Research and Technology - Hellas (FORTH)
Heraklion, Crete, Greece

Organizers
Apostolos Hadjidimos, Elias Houstis, Emmanuel Vavalis

View Abstracts
Conference Homepage

Conjugate gradients with indefinite constraint preconditioning
by
Miroslav Rozloznik
Institute of Computer Science, Czech Academy of Sciences, Prague
Coauthors: Valeria Simoncini

CONJUGATE GRADIENTS WITH INDEFINITE CONSTRAINT PRECONDITIONING

CONJUGATE GRADIENTS WITH INDEFINITE CONSTRAINT PRECONDITIONING

MIROSLAV ROZLOZNÍK

Institute of Computer Science, Academy of Sciences of the Czech Republic
Pod vodárenskou vezí 2, CZ-182 07 Prague 8, Czech Republic
and Technical University of Liberec, Department of Modelling of Processes
Faculty of Mechatronics and Interdisciplinary Studies, Liberec
e-mail: miro@cs.cas.cz

Abstract

In this contribution we analyze the null-space projection (or constraint) indefinite preconditioner applied to the solution of large-scale (and indefinite) saddle point problems. Since it was shown that the behavior of most of nonsymmetric Krylov subspace methods can be in this case related to the convergence of preconditioned conjugate gradient method (PCG) we study in detail its theoretical properties and propose simple procedures for correcting its possible misconvergence. The numerical behavior of the scheme is discussed and the maximum attainable accuracy of the approximate solution computed in finite precision arithmetic is estimated. This is a joint work with V. Simoncini and it was supported by the Grant Agency of the Czech Republic under grant No. 101/00/1035.

http://www.cs.cas.cz/~miro

Date received: February 8, 2001


Copyright © 2001 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Mathematical Conference Abstracts. Document # cagm-12.