Field-of-values analysis of preconditioned iterative methods for nonsymmetric elliptic problems

Summary. The convergence rate of Krylov subspace methods for the solution of nonsymmetric systems of linear equations, such as GMRES or FOM, is studied. Bounds on the convergence rate are presented which are based on the smallest real part of the field of values of the coefficient matrix and of its inverse. Estimates for these quantities are available during the iteration from the underlying Arnoldi process. It is shown how these bounds can be used to study the convergence properties, in particular, the dependence on the mesh-size and on the size of the skew-symmetric part, for preconditioners for finite element discretizations of nonsymmetric elliptic boundary value problems. This is illustrated for the hierarchical basis and multilevel preconditioners which constitute popular preconditioning strategies for such problems. Received May 3, 1996     

A Criterion for Spectrum Decomposition

We obtain a necessary and sufficient condition for the decomposition of the spectrum of an arbitrary nonsymmetric potential whose least value is attained at finitely many points.     

STRONG LAW OF LARGE NUMBERS AND ASYMPTOTIC EQUIPARTITION PROPERTY FOR NONSYMMETRIC MARKOV CHAIN FIELDS ON CAYLEY TREES

Some strong laws of large numbers for the frequcncies of occurrence of states and ordered couples of states for nonsymmetric Markov chain fields(NSMC)on Cayley trees are studied.In the proof,a new technique for the study of strong liinit theorems of Markov chains is extended to the case of Markov chain fields.The asymptotic equiparti- tion properties with almost everywhere(a.e.)convergence for NSMC on Cayley trees are obtained.     

Liquid crystalline polymers V. Thermotropic liquid crystalline poly(azomethine-ether)s containing a cycloalkanone moiety in the polymer backbone

A new homologous series of thermotropic liquid crystalline poly(azomethine-ether)s was synthesized by solution polycondensation of various diformyl-alpha,omega-diphenoxyalkanes, I-VIII with 2,5-bis(m-aminobenzylidene)cyclopentanone, XI and 2,6-bis(m-aminobenzylidene)cyclohexanone, XII. The inherent viscosities of the polymers were in the range 0.34-0.75 dI g^-1. The mesomorphic properties were studied as a function of the diphenoxyalkane spacer length. Analyses by DSC and optical polarized microscopy demonstrated that the poly(azomethine-ether)s form nematic mesophases over wide temperature ranges. The poly(azomethine-ether)s based on the cyclohexanone moiety had better thermal stabilities than the corresponding poly(azomethine-ether)s based on cyclopentanone.     

Matrix Continued Fraction for the Resolvent Function of the Band Operator

The aim of this paper is the expansion of a matrix function in terms of a matrix-continued fraction as defined by Sorokin and Van Iseghem. The function under study is the Weyl function or resolvent function of an operator, given in the standard basis by a bi-infinite band matrix, with p subdiagonals and q superdiagonals, where the p + q – 1 intermediate diagonals are zero. The main goal of this paper is to find, for the moments, an explicit form in terms of the coefficients of the continued fraction, called genetic sums, which lead to a generalization of the notion of a Stieltjes continued fraction. These results are extension of some results already found for the vector case (p = 1) and are a step in the direction towards the solution of the direct and inverse spectral problem. The actual computation of the approximants of the given function as the convergents of the continued fraction is shown to be effective. Some possible extensions are considered.     

一类增算子不动点定理及其应用

利用锥理论和非对称迭代方法,讨论了不具有连续性和紧性条件的增算子方程解的存在唯一性.作为其应用着重讨论了非增算子方程解的存在唯一性,并给出了迭代序列收敛于解的误差估计,改进和推广了某些已知结果.     

Effect of numerical integration in the DGFEM for nonlinear convection‐diffusion problems

This paper is concerned with the effect of numerical integration applied to the discontinuous Galerkin finite element discretization of nonlinear convection‐diffusion problems in 2D. In the space semidiscretization the volume and line integrals are evaluated by numerical quadratures. Our goal is to estimate the error caused by the numerical integration and to show what numerical quadratures guarantee that the accuracy of the method with exact integration is preserved. © 2007 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq, 2007     

Linking-Preserving Perturbations of Symmetric Functionals

We consider paths of functionals starting with one which is invariant under the action of an arbitrary group of symmetries. We give conditions for the existence of an unbounded sequence of critical values of the non-symmetric functional at the end of the path in terms of the growth of the critical values of the symmetric one. We apply this to obtain a multiplicity result for a system of elliptic equations whose symmetries are perturbed by a linear term and a non-homogeneous boundary condition.     

A simpler GMRES

The generalized minimal residual (GMRES) method is widely used for solving very large, nonsymmetric linear systems, particularly those that arise through discretization of continuous mathematical models in science and engineering. By shifting the Arnoldi process to begin with Ar₀ instead of r₀, we obtain simpler Gram–Schmidt and Householder implementations of the GMRES method that do not require upper Hessenberg factorization. The Gram–Schmidt implementation also maintains the residual vector at each iteration, which allows cheaper restarts of GMRES(m) and may otherwise be useful.     

Truncated Trust Region Methods Based on Preconditioned Iterative Subalgorithms for Large Sparse Systems of Nonlinear Equations

This paper is devoted to globally convergent methods for solving large sparse systems of nonlinear equations with an inexact approximation of the Jacobian matrix. These methods include difference versions of the Newton method and various quasi-Newton methods. We propose a class of trust region methods together with a proof of their global convergence and describe an implementable globally convergent algorithm which can be used as a realization of these methods. Considerable attention is concentrated on the application of conjugate gradient-type iterative methods to the solution of linear subproblems. We prove that both the GMRES and the smoothed COS well-preconditioned methods can be used for the construction of globally convergent trust region methods. The efficiency of our algorithm is demonstrated computationally by using a large collection of sparse test problems.