Computing

, Volume 54, Issue 2, pp 167–183 | Cite as

Explicit preconditioned iterative methods for solving large unsymmetric finite element systems

  • E. A. Lipitakis
  • G. A. Gravvanis
Article

Abstract

A class of Generalized Approximate Inverse Matrix (GAIM) techniques, based on the concept of LU-sparse factorization procedures, is introduced for computing explicitly approximate inverses of large sparse unsymmetric matrices of irregular structure, without inverting the decomposition factors. Explicit preconditioned iterative methods, in conjunction with modified forms of the GAIM techniques, are presented for solving numerically initial/boundary value problems on multiprocessor systems. Application of the new methods on linear boundary-value problems is discussed and numerical results are given.

AMS Subject Classification

65F10 

Key words

Approximate inverse matrix techniques explicit preconditioners parallel iterative methods unsymmetric finite element systems initial/boundary-value problems 

Explizite Präkonditionierungsverfahren zur Lösung großer unsymmetrischer Finite-Element-Systeme

Zusammenfassung

Es wird eine Methode zur Approximation der verallgemeinerten inversen Matrix (GAIM) diskutiert, die auf dem Konzept der schwachbesetzten LU-Faktorisierung basiert und explizite Inverse großer schwachbesetzter unsymmetrischer Matrizen auf irregulären Strukturen approximiert, ohne die Zerlegungsfaktoren zu invertieren. In Verbindung mit Modifikationen der GAIM-Technik werden explizite Präkonditionierungsmethoden zur numerischen Lösung von Anfangsrandwertproblemen auf Multiprozessorsystemen vorgestellt. Anwendungen der neuen Methoden auf lineare Randwertaufgaben werden diskutiert und numerische Resultate präsentiert.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. [1]
    APPEAL Library, ESPRIT CIM: Research and Development Project 1062 CAD Analysis Software. ACCORD Research Consortium, Department of Informatics, Athens University of Economics and Business, Athens, Greece, 1990.Google Scholar
  2. [2]
    Arioli, M., Duff, I. S.: Experiments in tearing large sparse systems Proc. on Advances in Numer. Computation, NPL July 1987, Oxford University Press, 1988.Google Scholar
  3. [3]
    Axelsson, O., Carey, G., Lindskog, G.: On a class of preconditioned iterative methods for parallel computers. Int. J. Numer. Methods Eng.27, 637–654 (1989).Google Scholar
  4. [4]
    Brambley, B. R.: Row projection methods for linear systems. Doctoral Thesis, Center for Supercomputing Research and Development: CSRD rep 881, University of Illinois at Urbana-Champaign, May 1989.Google Scholar
  5. [5]
    Dubois, P. F., Greenbaum, A., Rodrigue, G. H.: Approximating the inverse of a matrix for use in iterative algorithms on vector processors. Computing22, 257–268 (1979).Google Scholar
  6. [6]
    Duff, I. S., Erisman, A. M., Reid, J. K.: Direct methods for sparse matrices. Oxford: Clarendon Press 1986.Google Scholar
  7. [7]
    Evans, D. J., Lipitakis, E. A.: A normalized implicit Conjugate Gradient method for the solution of large sparse systems of linear equations. Comp. Meth. Appl. Mech. Engng.23, 1–19 (1980).Google Scholar
  8. [8]
    Fletcher, R.: Conjugate gradient methods for indefinite systems. In: Numerical analysis Dundee 1975 (Watson, G. A., ed.). Berlin Heidelberg New York: Springer 1970 (Lecture Notes in Mathematics Vol. 506).Google Scholar
  9. [9]
    Gravvanis, G. A.: A study on algorithmic techniques for the numerical solution of initial and boundary-value problems. Doctoral Thesis, Department of Informatics, Athens University of Economics and Business, Athens, Greece, 1991.Google Scholar
  10. [10]
    Johnson, O. G., Paul, G.: Vector algorithms for elliptic partial differential equations based on the Jacobi method, pp. 345–351. Proc. Elliptic Problem Solvers Conference, Los Alamos. Academic Press: New York 1981.Google Scholar
  11. [11]
    Lipitakis, E. A.: Generalized extended to the limit sparse factorization techniques for solving unsymmetric finite element systems. Computing32, 255–270 (1984).Google Scholar
  12. [12]
    Lipitakis, E. A.: Solving elliptic BV problems on parallel processors by approximate inverse matrix semi-direct methods based on the multiple explicit Jacobi iteration. Comp. Math. Appl.10, 171–184 (1984).Google Scholar
  13. [13]
    Lipitakis, E. A., Evans, D. J.: Solving non-linear elliptic difference equations by extendable sparse factorization procedures. Computing24, 325–339 (1980).Google Scholar
  14. [14]
    Lipitakis, E. A., Evans, D. J.: Explicit semi-direct methods based on approximate inverse matrix techniques for solving boundary-value problems on parallel processors. Math. Comput. Simulation29, 1–17 (1987).Google Scholar
  15. [15]
    Munksgaard, N.: Solving sparse symmetric sets of linear equations by preconditioned conjugate gradient. ACM TOMS6, 206–219, 1980.Google Scholar
  16. [16]
    Niessner, H., Reichert, K.: On computing the inverse of a sparse matrix. Int. J. Num. Meth. Eng.19, 1513–1526 (1983).Google Scholar
  17. [17]
    Pinder, G. F., Gray, W. G.: Finite element simulation in surface subsurface hydrology. New York: Academic Press 1977.Google Scholar
  18. [18]
    Porsching, T. A.: On the origins and numerical solution of some sparse non-linear systems. In: Sparse matrix computation. New York: Academic Press 1976.Google Scholar
  19. [19]
    Waugh, F. V., Dwyer, P. S.: Compact computation of the inverse of a matrix. Ann. Math. Stat.16, 259–271 (1945).Google Scholar

Copyright information

© Springer-Verlag 1995

Authors and Affiliations

  • E. A. Lipitakis
    • 1
  • G. A. Gravvanis
    • 1
  1. 1.Department of InformaticsAthens University of Economics and BusinessAthensGreece

Personalised recommendations