Conjugate gradient predictor corrector method for solving large scale problems

In this paper, we give a new method for solving large scale problems. The basic idea of this method depends on implementing the conjugate gradient as a corrector into a continuation method. We use the Euler method as a predictor. Adaptive steplength control is used during the tracing of the solution curve. We present some of our experimental examples to demonstrate the efficiency of the method.

     

Finite convergence analysis and weak sharp solutions for variational inequalities

In this paper, we study the weak sharpness of the solution set of variational inequality problem (in short, VIP) and the finite convergence property of the sequence generated by some algorithm for finding the solutions of VIP. In particular, we give some characterizations of weak sharpness of the solution set of VIP without considering the primal or dual gap function. We establish an abstract result on the finite convergence property for a sequence generated by some iterative methods. We then apply such abstract result to discuss the finite termination property of the sequence generated by proximal point method, exact proximal point method and gradient projection method. We also give an estimate on the number of iterates by which the sequence converges to a solution of the VIP.     

On error estimates in the Galerkin method for hyperbolic equations

We consider the Cauchy problem in a Hilbert space for a second-order abstract quasilinear hyperbolic equation with variable operator coefficients and nonsmooth (but Bochner integrable) free term. For this problem, we establish an a priori energy error estimate for the semidiscrete Galerkin method with an arbitrary choice of projection subspaces. Also, we establish some results on existence and uniqueness of an exact weak solution. We give an explicit error estimate for the finite element method and the Galerkin method in Mikhlin form.     

Spectral gradient projection method for solving nonlinear monotone equations

An algorithm for solving nonlinear monotone equations is proposed, which combines a modified spectral gradient method and projection method. This method is shown to be globally convergent to a solution of the system if the nonlinear equations to be solved is monotone and Lipschitz continuous. An attractive property of the proposed method is that it can be applied to solving nonsmooth equations. We also give some preliminary numerical results to show the efficiency of the proposed method.     

Application to differential transformation method for solving systems of differential equations

In this paper, we present an analytical solution for different systems of differential equations by using the differential transformation method. The convergence of this method has been discussed with some examples which are presented to show the ability of the method for linear and non-linear systems of differential equations. We begin by showing how the differential transformation method applies to a non-linear system of differential equations and give two examples to illustrate the sufficiency of the method for linear and non-linear stiff systems of differential equations. The results obtained are in good agreement with the exact solution and Runge–Kutta method. These results show that the technique introduced here is accurate and easy to apply.     

Generalized equations and the generalized Newton method

We give some convergence results on the generalized Newton method (referred to by some authors as Newton's method) and the chord method when applied to generalized equations. The main results of the paper extend the classical Kantorovich results on Newton's method to (nonsmooth) generalized equations. Our results also extend earlier results on nonsmooth equations due to Eaves, Robinson, Josephy, Pang and Chan. We also propose inner-iterative schemes for the computation of the generalized Newton iterates. These schemes generalize popular iterative methods (Richardson's method, Jacobi's method and the Gauss-Seidel method) for the solution of linear equations and linear complementarity problems and are shown to be convergent under natural generalizations of classical convergence criteria. Our results are applicable to equations involving single-valued functions and also to a class of generalized equations which includes variational inequalities, nonlinear complementarity problems and some nonsmooth convex minimization problems.     

Convergence properties of the cross-entropy method for discrete optimization

We present new theoretical convergence results on the cross-entropy (CE) method for discrete optimization. We show that a popular implementation of the method converges, and finds an optimal solution with probability arbitrarily close to 1. We also give conditions under which an optimal solution is generated eventually with probability 1.     

Krasnoselski-Mann type iterative method for hierarchical fixed point problem and split mixed equilibrium problem

In this paper, we suggest and analyze a Krasnoselski-Mann type iterative method to approximate a common element of solution sets of a hierarchical fixed point problem for nonexpansive mappings and a split mixed equilibrium problem. We prove that sequences generated by the proposed iterative method converge weakly to a common element of solution sets of these problems. Further, we derive some consequences from our main result. Furthermore, we extend the considered iterative method to a split monotone variational inclusion problem and deduce some consequences. Finally, we give a numerical example to justify the main result. The method and results presented in this paper generalize and unify the corresponding known results in this area.     

A regularized projection method for complementarity problems with non-Lipschitzian functions

We consider complementarity problems involving functions which are not Lipschitz continuous at the origin. Such problems arise from the numerical solution for differential equations with non-Lipschitzian continuity, e.g. reaction and diffusion problems. We propose a regularized projection method to find an approximate solution with an estimation of the error for the non-Lipschitzian complementarity problems. We prove that the projection method globally and linearly converges to a solution of a regularized problem with any regularization parameter. Moreover, we give error bounds for a computed solution of the non-Lipschitzian problem. Numerical examples are presented to demonstrate the efficiency of the method and error bounds.

     

Nonlinear semidefinite programming: sensitivity, convergence, and an application in passive reduced-order modeling

We consider the solution of nonlinear programs with nonlinear semidefiniteness constraints. The need for an efficient exploitation of the cone of positive semidefinite matrices makes the solution of such nonlinear semidefinite programs more complicated than the solution of standard nonlinear programs. This paper studies a sequential semidefinite programming (SSP) method, which is a generalization of the well-known sequential quadratic programming method for standard nonlinear programs. We present a sensitivity result for nonlinear semidefinite programs, and then based on this result, we give a self-contained proof of local quadratic convergence of the SSP method. We also describe a class of nonlinear semidefinite programs that arise in passive reduced-order modeling, and we report results of some numerical experiments with the SSP method applied to problems in that class.     

A matrix-splitting method for symmetric affine second-order cone complementarity problems

The affine second-order cone complementarity problem (SOCCP) is a wide class of problems that contains the linear complementarity problem (LCP) as a special case. The purpose of this paper is to propose an iterative method for the symmetric affine SOCCP that is based on the idea of matrix splitting. Matrix-splitting methods have originally been developed for the solution of the system of linear equations and have subsequently been extended to the LCP and the affine variational inequality problem. In this paper, we first give conditions under which the matrix-splitting method converges to a solution of the affine SOCCP. We then present, as a particular realization of the matrix-splitting method, the block successive overrelaxation (SOR) method for the affine SOCCP involving a positive definite matrix, and propose an efficient method for solving subproblems. Finally, we report some numerical results with the proposed algorithm, where promising results are obtained especially for problems with sparse matrices.     

An iterative method for linear discrete ill-posed problems with box constraints

Many questions in science and engineering give rise to linear discrete ill-posed problems. Often it is desirable that the computed approximate solution satisfies certain constraints, e.g., that some or all elements of the computed solution be nonnegative. This paper describes an iterative method of active set-type for the solution of large-scale problems of this kind. The method employs conjugate gradient iteration with a stopping criterion based on the discrepancy principle and allows updates of the active set by more than one index at a time.     

Noniterative solution of inverse problems by the linear least square method

In this paper, a noniterative linear least-squares error method developed by Yang and Chen for solving the inverse problems is re-examined. For the method, condition for the existence of a unique solution and the error bound of the resulting inverse solution considering the measurement errors are derived. Though the method was shown to be able to give the unique inverse solution at only one iteration in the literature, however, it is pointed out with two examples that for some inverse problems the method is practically not applicable, once the unavoidable measurement errors are included. The reason behind this is that the so-called reverse matrix for these inverse problems has a huge number of 1-norm, thus, magnifying a small measurement error to an extent that is unacceptable for the resulting inverse solution in a practical sense. In other words, the method fails to yield a reasonable solution whenever applied to an ill-conditioned inverse problem. In such a case, two approaches are recommended for decreasing the very high condition number: (i) by increasing the number of measurements or taking measurements as close as possible to the location at which the to-be-estimated unknown condition is applied, and (ii) by using the singular value decomposition (SVD).     

On the approximate solutions for system of fractional integro-differential equations using Chebyshev pseudo-spectral method

In this paper, we implement Chebyshev pseudo-spectral method for solving numerically system of linear and non-linear fractional integro-differential equations of Volterra type. The proposed technique is based on the new derived formula of the Caputo fractional derivative. The suggested method reduces this type of systems to the solution of system of linear or non-linear algebraic equations. We give the convergence analysis and derive an upper bound of the error for the derived formula. To demonstrate the validity and applicability of the suggested method, some test examples are given. Also, we present a comparison with the previous work using the homotopy perturbation method.     

Three-point finite-difference schemes, Padé and the spectral Galerkin method. I. One-sided impedance approximation

A method for calculating special grid placement for three-point schemes which yields exponential superconvergence of the Neumann to Dirichlet map has been suggested earlier. Here we show that such a grid placement can yield impedance which is equivalent to that of a spectral Galerkin method, or more generally to that of a spectral Galerkin-Petrov method. In fact we show that for every stable Galerkin-Petrov method there is a three-point scheme which yields the same solution at the boundary. We discuss the application of this result to partial differential equations and give numerical examples. We also show equivalence at one corner of a two-dimensional optimal grid with a spectral Galerkin method.

     

A mathematical programming problem with singular mixed pointwise-integral constraints

We investigate an infinite dimensional optimization problem which constraints are singular integral-pointwise ones. We give some partial results of existence for a solution in some particular cases. However, the lack of compactness, even in L¹ prevents to conclude in the general case. We give an existence result for a weak solution (as a measure) that we are able to describe. The regularity of such a solution is still an open problem.     

Observations on an adaptive moving grid method for one-dimensional systems of partial differential equations

Recently a scheme has been proposed for choosing a moving mesh based on minimizing the time rate of change of the solution in the moving coordinates for one-dimensional systems of PDEs. In this paper we show how to apply this idea to systems where the time derivatives cannot be solved for explicitly, writing the moving mesh equations in an implicit form. We give a geometrical interpretation of the scheme which exposes some of its weaknesses, and suggest some modifications based on this interpretation which increase the efficiency of the scheme. Finally, we present some numerical experiments which illustrate how well the resulting method works.     

On a method of calculation of hypersingular integrals

We propose a method for calculation of one-dimensional and two-dimensional hypersingular integrals, which allows to obtain a solution in the analytic form in a series of cases. Based on this method, we construct an approximate method for calculation of one-dimensional and twodimensional hypersingular integrals having singularities on rather complicated manifolds. We give estimates of errors of suggested algorithms.     

The nullspace method for the three-dimensional Stokes problem

In this paper, we present the preconditioned nullspace method for the iterative solution of the three-dimensional Stokes problem. In the nullspace method, the original saddle point system is reduced to a positive definite problem by representing the solution with respect to a basis of discretely divergence free vectors. The exact, explicit computation of such a basis typically has non-optimal (storage and computational) complexity. There exist some algorithms that exploit the sparsity of the matrix and work well for two dimensional problems but fail for three dimensions. Here, we will exploit an implicit representation of the nullspace basis which can be computed efficiently also in a three-dimensional setting, possibly only as an approximation. We will present some numerical results to illustrate the performance of the resulting solution method. (© 2008 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim)     

An expanded mixed covolume method for elliptic problems

We consider the mixed covolume method combining with the expanded mixed element for a system of first‐order partial differential equations resulting from the mixed formulation of a general self‐adjoint elliptic problem with a full diffusion tensor. The system can be used to model the transport of a contaminant carried by a flow in porous media. We use the lowest order Raviart‐Thomas mixed element space. We show the first‐order error estimate for the approximate solution in L² norm. We show the superconvergence both for pressure and velocity in certain discrete norms. We also get a finite difference scheme by using proper approximate integration formulas. Finally we give some numerical examples. © 2004 Wiley Periodicals, Inc. Numer Methods Partial Differential Eq, 2005