Numerische Mathematik

, Volume 39, Issue 1, pp 119–137

Partitioned variable metric updates for large structured optimization problems

  • A. Griewank
  • Ph. L. Toint
Article

Summary

This paper presents a minimization method based on the idea of partitioned updating of the Hessian matrix in the case where the objective function can be decomposed in a sum of convex “element” functions. This situation occurs in a large class of practical problems including nonlinear finite elements calculations. Some theoretical and algorithmic properties of the update are discussed and encouraging numerical results are presented.

Subject Classifications

AMS(MOS): 65H10 CR: 5.15 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Axelsson, O.: Solution of linear systems of equations: Iterative Methods. In: Sparse matrix techniques. V.A. Baker ed. Copenhagen 1976. Springer Verlag Berlin 1976Google Scholar
  2. 2.
    Brandt, A.: Multi-level adaptative solutions to boundary value problems. Math. Comput.31, 333–390 (1977)Google Scholar
  3. 3.
    Davidon, W.C.: Variable metric method for minimization. Technical Report #ANL-5990 (Rev.), Argonne National Laboratory, Research and Development 1959Google Scholar
  4. 4.
    Dennis, J.E., Moré, J.J.: Quasi-Newton methods: Motivation and theory. SIAM Rev.19, 46–89 (1977)Google Scholar
  5. 5.
    Dixon, L.C.W.: The solution of the Navier-Stokes equations via finite elements and optimization on a parallel processor. Presented at the CEC/CREST International Summer School on Numerical Algorithms for Parallel Processors. Bergamo, Italy June 1981Google Scholar
  6. 6.
    Ekeland, I., Temam, R.: Convex analysis and variational problems. North-Holland: Amsterdam 1976Google Scholar
  7. 7.
    Fletcher, R., Powell, M.J.D.: On the modification ofLDL 1 factorisations. Math. Comput.28, 1067–1087 (1974)Google Scholar
  8. 8.
    Gill, P., Murray, W.: Conjugate gradient methods for large scale nonlinear optimization. Technical Report SOL 79-15, Department of Operations Research, Stanford University, Stanford 1979Google Scholar
  9. 9.
    Greenstadt, J.: Variations upon variable metric methods. Math. Comput.24, 1–18 (1970)Google Scholar
  10. 10.
    Griewank, A., Toint, Ph.L.: On the unconstrained optimization of partially separable functions. In: Nonlinear optimization, M.J.D. Powell ed. Academic Press: New York (1981)Google Scholar
  11. 11.
    Gustafsson, I.: Stability and rate of convergence of modified incomplete factorisation methods. Research Report 79.02 R Department of Computer Science, University of Göteborg (1979)Google Scholar
  12. 12.
    Huang, H.Y.: Unified Approach to quadratically convergent algorithms for function minimization. J. Optimization Theory Appl.5, 405–423 (1970)CrossRefGoogle Scholar
  13. 13.
    Kamat, M.P., Vanden Brink, D.J., Watson, L.T.: Non-linear structural analysis using quasi-Newton minimization algorithms that exploit sparsity. Presented at the International Conference on Numerical Methods for Nonlinear Problems, Swansea, U.K. (1980)Google Scholar
  14. 14.
    Marwil, E.: Exploiting sparsity in Newton-like methods. PhD thesis, Cornell University, Ithaca, New York (1978)Google Scholar
  15. 15.
    Oren, S.S., Spedicato, E.: Optimal conditioning of self-scaling variable metric algorithms. Math. Programming10, 70–90 (1976)Google Scholar
  16. 16.
    Powell, M.J.D.: Restart procedures for the conjugate gradient method. Math. Programming12, 241–254 (1977)Google Scholar
  17. 17.
    Powell, M.J.D.: Some global convergence properties of a variable metric algorithm for minimization without exact line searches. SIAM-AMS Proceedings9, 53–72 (1976)Google Scholar
  18. 18.
    Shanno, D.F.: On variable metric methods for sparse hessians. Math. Comput.34, 499–514 (1980)Google Scholar
  19. 19.
    Shanno, D.F., Phua, K.H.: Matrix conditionning and nonlinear optimization. Math. Programming14 (1978)Google Scholar
  20. 20.
    Stetter, H.J.: Analysis of discretization methods for ordinary differential equations. Springer tracts on Natural Philosophy23 (1973)Google Scholar
  21. 21.
    Toint, Ph.L.: On sparse and symmetric matrix updating subject to a linear equation. Math. Comput.31, 954–961 (1977)Google Scholar
  22. 22.
    Toint, Ph.L.: A note on sparsity exploiting quasi-Newton methods. Math. Programming21, 172–181 (1981)Google Scholar
  23. 23.
    Toint, Ph.L.: Towards an efficient sparsity exploiting Newton method. In: Sparse matrices and their uses. I.S. Duff ed. Academic Press: New York 1981Google Scholar
  24. 24.
    Toint, Ph.L., Strodiot, J.J.: An algorithm for unconstrained minimization of large scale problems by Decomposition in subspaces. Technical Report 76/1. Department of Maths, FUN Namur, Belgium 1976Google Scholar

Copyright information

© Springer-Verlag 1982

Authors and Affiliations

  • A. Griewank
    • 1
    • 2
  • Ph. L. Toint
    • 1
    • 2
  1. 1.Department of Applied Mathematics and Theoretical PhysicsUniversity of CambridgeCambridgeUK
  2. 2.Department of MathematicsFacultés Universitaires de NamurNamurBelgium

Personalised recommendations