Majorizing sequences for Newton’s method from initial value problems

The most restrictive condition used by Kantorovich for proving the semilocal convergence of Newton’s method in Banach spaces is relaxed in this paper, providing we can guarantee the semilocal convergence in situations that Kantorovich cannot. To achieve this, we use Kantorovich’s technique based on majorizing sequences, but our majorizing sequences are obtained differently, by solving initial value problems.     

The convergence of shooting methods for singular boundary value problems

We investigate the convergence properties of single and multiple shooting when applied to singular boundary value problems. Particular attention is paid to the well-posedness of the process. It is shown that boundary value problems can be solved efficiently when a high order integrator for the associated singular initial value problems is available. Moreover, convergence results for a perturbed Newton iteration are discussed.

     

Cubic spline functions and initial value problems

A numerical process is presented which provides a cubic spline function approximation for the solution of initial value problems in ordinary differential equations. With interpolate cubic spline functions we can achieveO(h ⁴) convergence.     

On the process of successive approximation in initial value problems

Summary The convergence of the process of the successive approximations for the initial value problem dw/dz=f(z, w), w(0)=0, is studied in the complex field. The resulting z-domains of convergence are more favorable than those supplied by the classical approach. The improvement is made possible by the use of a ? principle of subordination ?, which allows a ? comparison test ? relative to the real field.     

A reliable treatment of singular Emden-Fowler initial value problems and boundary value problems

In this paper, the variational iteration method (VIM) is used to study the singular Emden-Fowler initial value problems and boundary value problems arising in physics and astrophysics. The VIM overcomes the singularity at the origin. The Lagrange multipliers for all cases of the equations are determined. The work is supported by analyzing few initial value problems and boundary value problems where the convergence of the results is emphasized.     

Stability and convergence of discrete kinetic approximations to an initial-boundary value problem for conservation laws

We present some new convergence results for a discrete velocities BGK approximation to an initial boundary value problem for a single hyperbolic conservation law. In this paper we show stability and convergence toward a unique entropy solution in the general framework without any restriction either on the data of the limit problem or on the set of velocity of the BGK model.

     

Parameter-robust numerical method for a system of singularly perturbed initial value problems

In this work we study a system of M( ≥ 2) first-order singularly perturbed ordinary differential equations with given initial conditions. The leading term of each equation is multiplied by a distinct small positive parameter, which induces overlapping layers. A maximum principle does not, in general, hold for this system. It is discretized using backward Euler difference scheme for which a general convergence result is derived that allows to establish nodal convergence of O(N ^− 1ln N) on the Shishkin mesh and O(N ^− 1) on the Bakhvalov mesh, where N is the number of mesh intervals and the convergence is robust in all of the parameters. Numerical experiments are performed to support the theoretical results.     

Solution of monotone nonlinear elliptic boundary value problems

LetL be a linear uniformly elliptic second order operator. The boundary value problem is solved for the nonlinear elliptic equationLu=f(x, u) wheref(x, u) is a monotone increasing function ofu for each pointx in the domain. A descent technique based on Newton's method is shown to yield a sequence of iterates which converges uniformly and quadratically to the solution. The convergence is independent of the choice for the initial iterate. Numerical results in two dimensions are presented.     

Solving linear initial value problems by Faber polynomials

In this paper we use the theory of Faber polynomials for solving N‐dimensional linear initial value problems. In particular, we use Faber polynomials to approximate the evolution operator creating the so‐called exponential integrators. We also provide a consistence and convergence analysis. Some tests where we compare our methods with some Krylov exponential integrators are finally shown. Copyright © 2002 John Wiley & Sons, Ltd.     

Probabilistic and deterministic convergence proofs for software for initial value problems

The numerical solution of initial value problems for ordinary differential equations is frequently performed by means of adaptive algorithms with user-input tolerance τ. The time-step is then chosen according to an estimate, based on small time-step heuristics, designed to try and ensure that an approximation to the local error commited is bounded by τ. A question of natural interest is to determine how the global error behaves with respect to the tolerance τ. This has obvious practical interest and also leads to an interesting problem in mathematical analysis. The primary difficulties arising in the analysis are that: (i) the time-step selection mechanisms used in practice are discontinuous as functions of the specified data; (ii) the small time-step heuristics underlying the control of the local error can break down in some cases. In this paper an analysis is presented which incorporates these two difficulties. For a mathematical model of an error per unit step or error per step adaptive Runge–Kutta algorithm, it may be shown that in a certain probabilistic sense, with respect to a measure on the space of initial data, the small time-step heuristics are valid with probability one, leading to a probabilistic convergence result for the global error as τ→0. The probabilistic approach is only valid in dimension m>1 this observation is consistent with recent analysis concerning the existence of spurious steady solutions of software codes which highlights the difference between the cases m=1 and m>1. The breakdown of the small time-step heuristics can be circumvented by making minor modifications to the algorithm, leading to a deterministic convergence proof for the global error of such algorithms as τ→0. An underlying theory is developed and the deterministic and probabilistic convergence results proved as particular applications of this theory. This revised version was published online in June 2006 with corrections to the Cover Date.     

Numerical solution of random differential initial value problems: Multistep methods

This paper deals with the construction of numerical methods of random initial value problems. Random linear multistep methods are presented and sufficient conditions for their mean square convergence are established. Main statistical properties of the approximations processes are computed in several illustrative examples. Copyright © 2010 John Wiley & Sons, Ltd.     

Runge-Kutta methods without order reduction for linear initial boundary value problems

Summary. It is well-known the loss of accuracy when a Runge–Kutta method is used together with the method of lines for the full discretization of an initial boundary value problem. We show that this phenomenon, called order reduction, is caused by wrong boundary values in intermediate stages. With a right choice, the order reduction can be avoided and the optimal order of convergence in time is achieved. We prove this fact for time discretizations of abstract initial boundary value problems based on implicit Runge–Kutta methods. Moreover, we apply these results to the full discretization of parabolic problems by means of Galerkin finite element techniques. We present some numerical examples in order to confirm that the optimal order is actually achieved. Received July 10, 2000 / Revised version received March 13, 2001 / Published online October 17, 2001     

On the stability and convergence of discretizations of initial value p.d.e.s

This paper examines the stability and convergence of discretizationsof initial value p.d.e.s using spatial discretization followedby time integration with an explicit one-step method. A Cauchyintegral representation is used to bound the growth in the discretesolution. New results are obtained regarding sufficient conditionsfor both algebraic and strong stability. Sufficient conditionsare also derived for convergence on a finite time interval.     

Convergence of linear multistep and one-leg methods for stiff nonlinear initial value problems

To prove convergence of numerical methods for stiff initial value problems, stability is needed but also estimates for the local errors which are not affected by stiffness. In this paper global error bounds are derived for one-leg and linear multistep methods applied to classes of arbitrarily stiff, nonlinear initial value problems. It will be shown that under suitable stability assumptions the multistep methods are convergent for stiff problems with the same order of convergence as for nonstiff problems, provided that the stepsize variation is sufficiently regular.     

Fourier regularization for a final value time-fractional diffusion problem

The evolution process of fractional order describes some phenomenon of anomalous diffusion and transport dynamics in complex system. The equation containing time-fractional derivative provides a suitable mathematical model for describing such a process. The backward problem for this system, which means to recover the initial state for some slow diffusion process from its present status, is very hard to solve due to the nonlocal property of fractional derivative and the irreversibility of time. For this ill-posed problem, we construct a regularizing solution using the Fourier transform method. Both the a-priori choice strategy and the a-posteriori choice strategy for the regularizing parameter are given, with the convergence analysis on the regularizing solution. Numerical implementations are presented to show the validity of the proposed scheme.     

A parametrization method for solving nonlinear two-point boundary value problems

A sharper version of the local Hadamard theorem on the solvability of nonlinear equations is proved. Additional parameters are introduced, and a two-parameter family of algorithms for solving nonlinear two-point boundary value problems is proposed. Conditions for the convergence of these algorithms are given in terms of the initial data. Using the right-hand side of the system of differential equations and the boundary conditions, equations are constructed from which initial approximations to the unknown parameters can be found. A criterion is established for the existence of an isolated solution to a nonlinear two-point boundary value problem. This solution is shown to be a continuous function of the data specifying the problem.     

An initial value method for the computation of the characteristic values and functions of an integral operator-II

A convergence proof is given for an initial value method for finding the spectral properties of an integral operator. The algorithm, which is based on a Cauchy system for the Fredholm determinants, is related to the Nyström method and results of Anselone and Atkinson become applicable. The proof is also shown to work for a modification of the procedure due to Kalaba and Scott.     

Consistency,convergence and stability of general discretizations of the initial value problem

Discretizations of nonlinear operators in Banach space are described and the concept of an inverse discretization introduced. In the main part of the paper, the very general formalism of BUTCHER for the initial value problem for ordinary differential equations is examined and the sufficiency of conditions for its stability and convergence is demonstrated. The order of convergence of these methods is discussed, and an example is given.     

Finite-difference method for a nonlocal boundary value problem for a third-order pseudoparabolic equation

We consider a nonlocal boundary value problem for a third-order pseudoparabolic equation with variable coefficients. For its solution, in the differential and finite-difference settings, we derive a priori estimates that imply the stability of the solution with respect to the initial data and the right-hand side on a layer as well as the convergence of the solution of the difference problem to that of the differential problem.     

Newton's method for convex nonlinear elliptic boundary value problems

Summary Newton's method is applied to solving the boundary value problem for the equationLu=f(x,u) whereL is a linear second order uniformly elliptic operator andf(x,u) is a convex monotone increasing function ofu for each pointx in the domainD. The Newton iterates are shown to converge uniformly, quadratically and monotonically downward to the solution of the problem. The convergence is independent of the choice for the initial Newton iterate. Numerical results are presented for several problems of physical interest.