Breakdowns and stagnation in iterative methods
One of the disadvantages of Krylov subspace iterative methods is the possibility of breakdown. This occurs when it is impossible to get the next approximation of the solution to the linear system of equationsAu=f. There are two different situations: lucky breakdown, when we have found the solution and hard breakdown, when the next Krylov subspace cannot be generated and/or the next approximate solution (iterate) cannot be computed. We show that some breakdowns depend on the chosen method of generating the basis vectors. Another undesirable feature of the iterative methods is stagnation. This occurs when the error does not change for several iterative steps. We investigate when iterative methods can stagnate and describe conditions which characterize stagnation. We show that in some cases stagnation can imply breakdown.
AMS subject classification65F10
Key wordsIterative methods Krylov subspace biconjugate gradient conjugate gradient GMRES generalized conjugate residual
Unable to display preview. Download preview PDF.
- 9.R., Fletcher,Conjugate gradient methods for indefinite systems, in Proceedings Dundee Conference on Numerical Analysis, 1975, G. Watson, ed., Springer-Verlag, 1976, Lecture Notes in Mathematics 506, pp. 73–89.Google Scholar
- 18.M. Hestenes,A History of Scientific Computing, Addison-Wesley, Reading, MA, 1990, Chap. Conjugacy and gradients, pp. 167–179.Google Scholar
- 22.W. Joubert and T. Manteuffel,Iterative methods for nonsymmetric linear systems, in Iterative Methods for Large Linear Systems, L. Hayes and D. Kincaid, eds., Academic Press, 1990, pp. 149–171.Google Scholar
- 25.Z. Leyk,Modified generalized conjugate residual method for nonsymmetric systems of linear equations, in Proceedings of the 6th Conference on Computational Techniques and Applications: CTAC93, D. Steward, H. Gardner, and D. Singleton, eds., World Scientific, Singapore, 1994, pp. 338–344.Google Scholar