- Research Article
57
- 10.1137/0706035
The Numerical Solution of Integral Equations on the Half-Line
- Sep 01, 1969
- SIAM Journal on Numerical Analysis
- Kendall Atkinson
The Numerical Solution of Integral Equations on the Half-Line
An application of the method of successive approximations for obtaining the solution of a nonlinear integral equation arising from a two-point boundary value problem is illustrated. In particular, we show sufficient conditions under which the convergence rate of the sequence can be improved.
The Numerical Solution of Integral Equations on the Half-Line
The Numerical Solution of Integral Equations on the Half-Line
Solutions of Integral Equations by using Successive Approximation Method
Background/Objectives: Integral transforms and iterative techniques play a central role in the theory and solution of integral equations. In this work, generalized special functions, namely Prefunctions and Extended Prefunctions are used to obtain exact and approximate solutions of Volterra integral equations of the second kind. These functions are parameterized by πΌ and extend the analytic structure of exponential, hyperbolic and trigonometric functions. These functions are effective tools in solving differential and integral equations. Methods: In this study successive approximations method is used to derive closed-form solutions for two classes of Volterra integral equations with different kernels. For the kernel πΎ(π₯, π‘) = π₯ β π‘ with π (π₯, πΌ) = ππ ππ(π₯, πΌ) the iterative sequence yields a generalized solution expressed as a series of shifted Prefunctions, reducing to π¦(π₯, 0) = 1 2 (ππ ππβ(π₯) + π ππ(π₯)) for πΌ =0 confirming the consistency of the method with known classical results. For the convolution kernel πΎ(π₯, π‘) = πβ(π₯βπ ) with π(π₯, πΌ) = ππ ππβ(π₯, πΌ) the Picard iteration process leads to the compact representation π¦(π₯, πΌ) = ππ ππβ(π₯, πΌ) + π ππ(π₯, πΌ + 1). Findings: Numerical illustrations and graphical results confirm uniform convergence of the successive approximations to the exact solution for various values of πΌ. For a large value of πΌ convergence rate is higher. Novelty: Prefunction -based methods provide generalized frameworks for solving Volterra integral equations and open further directions for applications in mathematical physics and engineering models. 2020 Mathematics Subject Classification: 45Dxx, 34A12. Keywords: Integral Equations; Prefunctions; Successive Approximation Method; Convergence; Picard Iterates
Read moreSuccessive Approximation, Variational Iteration, and Multistage-Analytical Methods for a SEIR Model of Infectious Disease Involving Vaccination Strategy
We consider a SEIR model for the spread (transmission) of an infectious disease. The model has played an important role due to world pandemic disease spread cases. Our contributions in this paper are three folds. Our first contribution is to provide successive approximation and variational iteration methods to obtain analytical approximate solutions to the SEIR model. Our second contribution is to prove that for solving the SEIR model, the variational iteration and successive approximation methods are identical when we have some particular values of Lagrange multipliers in the variational iteration formulation. Third, we propose a new multistage-analytical method for solving the SEIR model. Computational experiments show that the successive approximation and variational iteration methods are accurate for small size of time domain. In contrast, our proposed multistage-analytical method is successful to solve the SEIR model very accurately for large size of time domain. Furthermore, the order of accuracy of the multistage-analytical method can be made higher simply by taking more number of successive iterations in the multistage evolution.
Read moreΠΠ΅ΡΠ΅ΡΠΊΠΈΠ΅ ΠΈΠ½ΡΠ΅Π³ΡΠ°Π»ΡΠ½ΡΠ΅ ΡΡΠ°Π²Π½Π΅Π½ΠΈΡ ΠΠΎΠ»ΡΡΠ΅ΡΡΠ° Ρ ΠΊΡΡΠΎΡΠ½ΠΎ-Π½Π΅ΠΏΡΠ΅ΡΡΠ²Π½ΡΠΌΠΈ ΡΠ΄ΡΠ°ΠΌΠΈ: ΡΠ΅ΠΎΡΠΈΡ ΠΈ ΡΠΈΡΠ»Π΅Π½Π½ΠΎΠ΅ ΡΠ΅ΡΠ΅Π½ΠΈΠ΅
This research focuses on addressing both linear and nonlinear fuzzy Volterra integral equations that feature piecewise continuous kernels. The problem is tackled using the method of successive approximations. The study discusses the existence and uniqueness of solutions for these fuzzy Volterra integral equations with piecewise kernels. Numerical results are obtained by applying the successive approximations method to examples for both linear and nonlinear scenarios. Error analysis graphs are plotted to illustrate the accuracy of the method. Furthermore, a comparative analysis is presented through graphs of approximate solutions for different fuzzy parameter values. To highlight the effectiveness and significance of the successive approximations method, a comparison is made with the traditional homotopy analysis technique. The results indicate that the successive approximation method outperforms the homotopy analysis method in terms of accuracy and effectiveness.
Read moreNumerial Methods for Volterra Integral Equations with Singular Kernels
Numerial Methods for Volterra Integral Equations with Singular Kernels
Finding of periodic solution of the Mathieu equation with the delay
The work suggests an approach for finding periodic solution of the nonlinear delayed differential Mathieu equations applied in the theory of oscillatory processes. The application of the numerical-analytical method to finding periodic solutions of this equation is known. This idea includes reducing the equation to the system of the first order. The article proposes the use of the previously developed method for finding periodic solutions of nonlinear second-order ordinary differential equations, also used for equations with delay, without being reduced to a system. In this case, the Green's function is constructed for a self-adjoint differential operator of the second derivative, defined on functions that satisfy periodic boundary conditions. The necessary and sufficient conditions for the existence of a periodic solution of the Mathieu equation are given. The solution itself is found by the method of successive approximations. The estimates for the method's rate of convergence were obtained.
Read moreOn the singular Volterra integral equation of the boundary value problem for heat conduction in a degenerating domain
In this paper, we consider a singular Volterra type integral equation of the second kind, to which some boundary value problems of heat conduction in domains with a boundary varying with time are reduced by the method of thermal potentials. The peculiarity of such problems is that the domain degenerates into a point at the initial moment of time. Accordingly, a distinctive feature of the integral equation under study is that the integral of the kernel, as the upper limit of integration tends to the lower one, is not equal to zero. This circumstance does not allow solving this equation by the method of successive approximations. We constructed the general solution of the corresponding characteristic equation and found the solution of the complete integral equation by the CarlemanβVekua method of equivalent regularization. It is shown that the corresponding homogeneous integral equation has a nonzero solution.
Read moreThe iterative solution of a nonlinear fredholm integral equation arising from a chemical reactor problem
An iterative solution of a particular non-linear Fredholm integral equation is considered. The method is based on the reduction of the equation for the residual at any stage of the iteration to a linear integral equation for the required perturbation, using Newton' s method. This equation is then solved by using a low order expansion in terms of Chebyshev polynomials. The numerical integrations required in the basic iterative process are carried out by the powerful Clenshaw-Curtis quadrature prescription. The non-linear integral equation considered arises from the unstable two-point boundary value problem for a tubular chemical reactor. The capacity of the present method in dealing with such numerical difficulties is illustrated in representative calculations. Comparisons are made with conventional solutions of the parent differential equation and also with attempts at direct iterative solution of the integral equation by standard Neumann-Liouville series and finally with the direct algebraic approach.
Read moreA modified Pagano method for the 3D dynamic responses of functionally graded magneto-electro-elastic plates
A modified Pagano method for the 3D dynamic responses of functionally graded magneto-electro-elastic plates
Iterative methods for best approximate solutions of linear integral equations of the first and second kinds
Iterative methods for best approximate solutions of linear integral equations of the first and second kinds
A Modified MSA for Stochastic Control Problems
The classical Method of Successive Approximations (MSA) is an iterative method for solving stochastic control problems and is derived from Pontryaginβs optimality principle. It is known that the MSA may fail to converge. Using careful estimates for the backward stochastic differential equation (BSDE) this paper suggests a modification to the MSA algorithm. This modified MSA is shown to converge for general stochastic control problems with control in both the drift and diffusion coefficients. Under some additional assumptions the rate of convergence is shown. The results are valid without restrictions on the time horizon of the control problem, in contrast to iterative methods based on the theory of forward-backward stochastic differential equations.
Read moreJoint Rate and Power Control in Wireless Network: A Novel Successive Approximations Method
The conventional method of joint rate and power control (JRPC) relies on high signal-to-interference ratio (SIR) assumption which achieves only suboptimal results. By using a novel successive convex approximations method, we can attain the global optimal source rates and link powers in a distributed fashion exploiting message passing. Through simulations, our method converges faster than the previous work based on logarithm successive convex approximations.
Read moreThe Method of Successive Approximation in Descriptive Ecology
The Method of Successive Approximation in Descriptive Ecology
Problems on elastic equilibrium of a wedge with cracks on the axis of symmetry
We use the Wiener-Hopf method to obtain exact solutions of plane deformation problems for an elastic wedge whose lateral sides are stress free and which has rectilinear cracks on its axis of symmetry. In problem 1, a finite crack issues from the wedge apex edge; in problem 2, a half-infinite crack originates at a certain distance from the wedge apex edge; and in problem 3, the wedge contains an internal finite crack. Earlier, many authors obtained approximate solutions to these problems [1β10] and exact solutions to problem 1 [11β17] and homogeneous problem 2 [18β20] (see also [21]). The method of approximate conformal transformation [1, 2] and that of integral equations [3, 4] were used to study a specific case of problem 1 about equilibrium of an elastic half-plane with a boundary crack perpendicular to the half-plane boundary. The solution to problem 1 was obtained in [11] by reducing the problem to the Riemann problem for analytic functions and in [6, 7], by the Wiener-Hopf method. In [11], the problem coefficient was factorized in terms of Cauchy-type integrals, but no further calculations were presented, and in [6, 7], an approximate factorization was performed by approximating the factorized function. In [8], problems 1 and 2 were reduced to Fredholm integral equations of the second kind, which were solved numerically. In problem 1, values of the integral equation density were calculated, and the normal displacements of the crack edges were expressed in terms of this density in the form of Abel integrals; no calculations were given in problem 2. The exact values of the stress intensity factors in problem 1 were obtained in [12β17] by the Wiener-Hopf method. The paper [12] considered the case of linear normal stresses given on the crack edges in the absence of tangential stresses was considered, the papers [13β16], the case of concentrated forces acting on the crack edges, and the paper [17], the case of concentrated moments applied to the wedge apex. Solutions to the homogeneous problem 2, where the crack edges are stress free and the principal vector and the principal moment of stresses are given at infinity, were constructed by the Wiener-Hopf method in [18β20]. Problem 3 was studied in [6, 7, 9] by different approximate methods: using the asymptotic solution of the integral equation, solving the Fredholm integral equation of the second kind by the method of successive approximations after approximating the Fourier transform of the difference kernel of the original integral equation, and using another approximation of the kernel to reduce the problem to a singular integral equation admitting a solution in closed form. Here, the asymptotic method can be used only in the case of a crack relatively distant from the wedge apex, and the approximation of the kernel transform gives satisfactory results if the angle at the wedge apex exceeds Ο. The authors present the numerical results for the stress intensity factors and the displacement jump at an internal point of the crack in the case where the wedge is a half-plane. In [6β9], the cases of rigid clamping and hinged support of the faces of an elastic wedge were investigated. Problem 3 was numerically solved by the method of singular integral equations in [10], where the case of fixed sides of the wedge was also considered. In what follows, problems 1 and 2 whose integral equations are given on a half-infinite interval and have distinct kernels are solved by the Wiener-Hopf method [22], and problem 3 is solved by the generalized Wiener-Hopf method, developed in [23β25] for solving integral equations with difference kernel on a finite interval. The coefficient of the functional Wiener-Hopf equation is factorized in terms of infinite products. We also present numerical results for the stress intensity factors, the normal stress distribution on the crack continuation line, and the normal displacements of the crack edges.
Read moreMethods of Estimating Travel Times and Epicentres
Summary We compare the method of successive approximation (MSA) with the formal method of solving the matrix of normal equations in travel time studies and show that if the matrix is ill-conditioned the MSA may be so slow to converge that effectively there is no convergence. Consequently as the true confidence limits which would reveal the ill-conditioning cannot be computed, P travel times estimated using the MSA may have systematic errors that will pass unnoticed. The formal method of solution has the advantage that ill-conditioning will be revealed by wide confidence limits on the estimates. Use of the formal method rather than the MSA for estimating the epicentres of groups of earthquakes seems to have similar advantages.
Read more