A Combination of Bernstein and Improved Block-Pulse Functions for Solving a System of Linear Fredholm Integral Equations

He, Ji-Huan; Taha, Mahmoud H.; Ramadan, Mohamed A.; Moatimid, Galal M.

doi:https://doi.org/10.1155/2022/6870751

Mathematical Problems in Engineering

On this page

Abstract Introduction Conclusion Data Availability Conflicts of Interest References Copyright Related Articles

Research Article | Open Access

Volume 2022 | Article ID 6870751 | https://doi.org/10.1155/2022/6870751

A Combination of Bernstein and Improved Block-Pulse Functions for Solving a System of Linear Fredholm Integral Equations

Ji-Huan He,^1,2,3Mahmoud H. Taha,⁴Mohamed A. Ramadan,⁵and Galal M. Moatimid⁴

Academic Editor: Filippo Cacace

Received17 Dec 2021

Accepted16 Feb 2022

Published16 Apr 2022

Abstract

The present paper illustrates a new numerical technique to solve a system of linear Fredholm integral equations of the second kind. The current work introduces a coupling between hybrid Bernstein functions and improved block-pulse functions (HBI). The current method transforms the linear second kind Fredholm integral equations system into an algebraic system that can be solved by using classical methods. Some numerical examples are introduced to validate the new approach. The results showed that the method is promising, powerful, ultimately accurate, and highly efficient.

1. Introduction

Integral equations are an integral part of many mathematical formulations for numerous phenomena. Integral equations can be frequently found in many applied fields, such as field theory, mechanics, biology, chemistry, physics, engineering, electrostatics, and economics; for illustration, see references [1–3]. Fredholm equations arise naturally in the theory of signal processing, for example, as the famous spectral concentration problem popularized by David Slepian. The operators involved are the same as linear filters. They also commonly arise in linear forward modeling and inverse problems. In physics, the solution of such integral equations allows for experimental spectra to be related to various underlying distributions, for instance, the mass distribution of polymers in a polymeric melt, or the distribution of relaxation times in the system. In addition, Fredholm integral equations also arise in fluid mechanics problems involving hydrodynamic interactions near finite-sized elastic interfaces. A specific application of the Fredholm equation is the generation of photo-realistic images in computer graphics, in which the Fredholm equation is used to model light transport from the virtual light sources to the image plane. The Fredholm equation is often called the rendering equation in this context.

However, most integral equations that arise in our real-life situations are difficult to solve analytically; therefore, a numerical method is required. This paper is going to work with two simultaneous different functions and combine them to create a new efficient technique. A detailed description of these two methods is thoroughly presented as follows.

The first one is known as the Bernstein functions. Many researchers have attempted to use Bernstein functions and improved block-pulse functions for solving integral equations. Bernstein functions are an important subclass of complete functions that can be found under many different definitions/names in many fields and branches of mathematics. The most common definition of Bernstein functions is given by probability: “Laplace exponents of infinite divisible nonnegative subdistributions” (or “Laplace exponents of subordinates”). By contrast, in harmonic analysis and Fourier’s analysis, they are generally named negative definite functions. Complete Bernstein functions are employed in the complex analysis under the name of Pick or Nevanlinna functions that can be found in different research papers and books. Concurrently, in the field of operator theory and matrix analysis, the more widely used name is the operator monotone function. Researchers studied the positivity of the solutions of Volterra integral equations and found that different forms of kernels associated with Bernstein functions started to appear, see reference [4]. There can be found a huge number of papers on every subclass of these classes, see reference [5]. On the other hand, only a few texts have brought to light the links between these classes and the existing techniques in other resources; for instance, see reference [6].

Bernstein polynomials have a leading role in numerous areas of mathematics. These polynomials are regularly found in analyzing several approximations theory and differential equations; for instance, see references [7–9]. Maleknejad et al. in reference [10] used Bernstein polynomials to effectively solve Volterra integral equations. Ostrovska in reference [11] studied the convergence of Bernstein polynomials. One can find a wide study on Bernstein polynomials and other similar polynomials as well in reference [12]. Mirzaee and Hoseini in reference [13] did the coupling between Bernstein polynomials and block-pulse functions which our work here depends on the way they did that. The two-dimensional Bernstein polynomials and their convergence analysis are also discussed in references [14, 15]. Bernstein polynomials also dealt with fractional integrodifferential equations with ease and it was accurate, see reference [16]. Bernstein’s polynomial is confined to the interval which will be very important within the style of Bezier curves for computer-aided geometric design (CAGD). They are used for the design of curves, and they are the starting point for several generalizations, particularly for higher dimensions and B-splines [17, 18]. Several characteristics of the Bezier curves and surfaces are inherited from those of the Bernstein polynomials. Because of the growing concerns with Bernstein polynomials, the investigation has emerged in such a way as to depict their properties as far as their coefficients are concerned after they are given inside the Bernstein basis. Farouki and Goodman [19] deduced that the calculation within the Bernstein basis is more stable than computations in other polynomial bases. Moreover, the alteration of basis algorithms is in most cases numerically unstable. Therefore, when the Bernstein basis is used to represent polynomials, it is smarter to make all estimates, utilizing only Bernstein coefficients. Formulae for multiplication, differentiation, integration, and subdivision of polynomials inside the Bernstein structure are given by Farouki and Rajan [20].

The class of Bernstein functions is firmly identified with completely monotone functions. The idea of Bernstein functions dates back to the potential theory school of Beurling and Deny [21]. It was then led by Berg et al. [22]. Berg et al. [23] named them as completely monotone functions (or the opposite of completely monotone functions). Therefore, Laplace exponents are still preferred by probabilistic; for instance, see studies by Berge et al. [23] and Bertoin [24].

The n^th Bernstein polynomial for a mapping over the interval is given as

A linear transformation of variables can be done to the abovementioned equation; consequently, similar polynomials can be constructed on any finite interval , as an extension to the work done in this paper. It can be thoroughly proven that if is continuous on the interval , it follows that the sequence () converges uniformly to on the interval Additionally, derivatives of the Bernstein polynomial converge to derivatives of (if there are any). Furthermore, if is convex, then each Bernstein polynomial is convex; for instance, see [25].

In this paper, the Bernstein polynomial will be coupled with the improved block-pulse functions which was first introduced by Farshid Mirzaee [26] to solve a system of Fredholm integral equations of the second kind. The system of Fredholm integral equations appears in many problems in applied sciences.

Let be the set of all real-valued functions that are continuous on the interval In this paper, the fixed-point theorem is used to discuss the existence and uniqueness of the solution for the following system of Fredholm integral equations:where are unknown functions, while and the kernels are known functions.

To crystallize the paper, it is arranged as follows: The existence and uniqueness of the solutions of the system of linear Fredholm integral equations are introduced in Section 2. Section 3 describes the coupling of Bernstein polynomial and improved block pulse functions to solve a system of linear Fredholm integral equations. The function approximations are also introduced. Section 4 introduces the solution method. Section 5 investigates the problem's convergence analysis. Section 6 includes numerical results as well as application examples to demonstrate the efficiency and accuracy of the proposed method. Finally, Section 7 provides the main outline of the paper's conclusions.

2. Existence and Uniqueness

In this section, the existence and uniqueness of solutions for a system of Fredholm integral equations (2) are studied.

2.1. Notes

denotes the set of all square matrices with positive real elements, stands for the identity matrix, and refers to the zero matrix. The matrix is said to be convergent to zero if as .

Some examples of matrices that converge to zero are as follows:(a)(b)(c)

Theorem 1. Let be a nonempty set endowed with two vector-valued metrics . Let be an operator. The following is assumed:(1)There exists a matrix in such a way that (2) is a complete vector-valued metric space(3) is continuous(4) is a generalized Hardy-Rogers-type contraction, that is, there exist A, B, C in such a way that the matrices and are convergent to zero and

Proof. It can be found in reference [27].
The solution of the system is sought in equation (2) which can be rewritten in the following form:It is assumed that there exist as follows:for all .
On a nonempty set , the metrics are considered, wherefor all .
The operator is defined byfor all .
We haveThen, using the assumptions, one getswhereThe matrix converges to zero ifSo, under this condition, becomes an -contraction with respect to .
In addition, for all , it is to be noted thatBy applying Theorem 1 with , it is found that system (2) has a unique solution in .

3. Hybrid Bernstein Improved Block-Pulse Functions (HBIs)

Definition 1. (improved block-pulse functions). The IBPFs were firstly introduced by Farshid Mirzaee [26]. Farshid solved a system of integral equations by using IBPFs. The variable interval block-pulse functions are derived from the regular block-pulse functions, but with a slight change in the interval width. The variable of interval block pulse is a agreed of functions defined over the interval as follows:where is a positive constant that represents the number of subintervals to be decided for the accuracy needed for solving the problem. The IBPFs are disjoint aswhere and those functions are orthogonal to each otherwhere .
The first terms of the IBPF can be written in a vector formEquation (17) gives

Definition 2. (HBI, [28]). is a combination of Bernstein polynomials and improved block-pulse functions, where all are orthogonal and complete. Subsequently, the set is a complete orthogonal system. Let where have two arguments, and representing the order of IBPFs and the degree of Bernstein polynomials, respectively. The function on the interval may be defined as follows:Therefore, the new basis is given by and the function can be approximated to the base function, where is an arbitrary positive integer and .
The next section deals with the approximating properties of such functions.

3.1. Function Approximation

In terms of the HBI basis, any function can be expressed as follows:where is the column vector defined asand is the row matrix defined asand we have

Or, it can be rewritten aswhere the inner product of the right hand of equation (26) is computed component-wise and the hybrid Bernstein improved block-pulse coefficients are given bythenwhere is the standard inner product and the matrix is the matrix that is commonly called a dual matrix and defined asand is defined as follows:and the function can be approximated as follows:where is an matrix; it can be obtained as follows:

4. Outline of the Solution of a System of Linear Fredholm Integral Equations

This section presents the algorithm of the method for solving a system of linear Fredholm equations as given in (2).

Step 1. The functions maybe approximated aswhere is an unknown vector and are given by (23).

Step 2. The functions and are also expanded into the HBI function as follows:where is a known -matrix and is a known vector.

Step 3. Substituting approximation (34)–(36) into equation (2). For , one getsSo, system (2) can be written in a matrix form as follows:where and are vectors in the following form:Also, is matrix as follows:where and is matrix. Therefore, the equation can be written as follows:where is the identity matrix. After solving this linear system, the solution can be approximated asTo validate our approach, some numerical examples will be introduced in the next section.

5. Convergence Analysis

We assume the following conditions on and , where are known functions and are unknown functions, in such a way that(a)(b) is continuous on and satisfies Lipschitz condition

The error of term of the solution obtained by the HBI method is denoted by , where and are the exact and approximate solution of (2), respectively. is given as

Theorem 2. The solution to Fredholm integral equations system (2) using the HBI method converges if .

Proof. We haveThis implies that
Hence, by choosing when , it implies that .

6. Numerical Examples

In this part, some mathematical models will be introduced to show the exactness of our methodology. Most of these models are taken from different references. In this way, one can reevaluate the obtained mathematical outcomes by the proposed technique for obtaining the exact solution. All results are computed by using a program implemented in MATLAB R2018b. The absolute errors are defined aswhere and are the approximate and the exact solutions of the integral equation system, respectively.

Example 1. Consider the system of two linear Fredholm integral equations that was given in references [29, 30].with exact solution .
method is applied to solve Example 1. The numerical results are tabulated, as can be seen in Table 1 for . Table 2 shows the comparison of the absolute error among the approximate solutions obtained by the proposed method, the block-pulse function method which works with equally spaced points intervals while taking nonequal spaced points in the improved block-pulse method with different intervals [30], and the iterative triangular function method for [31] at different values of .

Example 2. Consider the system of linear Fredholm integral equations (Tables 3 and 4):with exact solution .

Example 3. Consider the following system of two linear Fredholm integral equations,with exact solution .
method is applied to solve Example 3. The numerical results are tabulated, as seen in Table 5 for . Table 6 shows the comparison of the absolute error between the approximate solutions obtained by the proposed method and the iterative triangular function method for [31] at different values of .

Example 4. Consider the system of two linear Fredholm integral equations [31, 32].with exact solution .
Table 7 shows the comparison among the results yielded by our presented method, iterative triangular function method [31], and the decomposition method proposed with [32] as far as Example 4 is concerned. Furthermore, in Table 8, the comparison of absolute error, among the approximate solutions obtained by the proposed method, iterative triangular function method for [31] and the decomposition method proposed in [32], at different values of is shown.

6.1. Work Extension

We intend to extend our work to apply our method to solve weakly singular Fredholm integral equation [33]. Also, fractional Fredholm integral equations will be investigated in future work [34–36] as well as two-dimensional Fredholm integral equations [37].

7. Conclusion

The hybrid Bernstein and improved block-pulse functions method proposed by Ramadan and Osheba [28] is used in this paper to achieve a numerical solution for a system of linear Fredholm integral equations of the second kind. Illustrative examples to test the accuracy and the applicability of the proposed are given. The obtained numerical results indicate that the accuracy of the proposed hybrid method is more satisfactory than the usage of triangular functions, iterative triangular functions, block-pulse function, and the decomposition method. Error obtained by the present method is reduced at lower iterations, that is why the new hybrid method is highly recommended. The software package used to obtain those results is the MATLAB R2018b program.

Data Availability

No data were associated with this research.

Conflicts of Interest

The authors declare no conflicts of interest.

References

H. Esmaeili and D. Moazami, “A stable kernel-based technique for solving linear Fredholm integral equations of the second kind and its applications,” Engineering Analysis with Boundary Elements, vol. 116, pp. 48–63, 2020.
View at: Publisher Site | Google Scholar
M. Mohammad, “A numerical solution of Fredholm integral equations of the second kind based on tight framelets generated by the oblique extension principle,” Symmetry, vol. 11, no. 7, Article ID 854, 2019.
View at: Publisher Site | Google Scholar
M. A. Ramadan and R. Ali Mohamed, “An efficient hybrid method for solving Fredholm integral equations using triangular functions,” New Trends in Mathematical Sciences, vol. 5, no. 1, pp. 213–224, 2017.
View at: Publisher Site | Google Scholar
S. Bhattacharya and N. Mandal, “Use of Bernstein polynomials in numerical solutions of Volterra integral equations,” Applied Mathematical Sciences, vol. 2, pp. 1773–1787, 2008.
View at: Google Scholar
R. A. Alharbey and K. Sultan, “Evolutionary algorithm based solution of rössler chaotic system using Bernstein polynomials,” Journal of Computational and Theoretical Nanoscience, vol. 17, no. 7, pp. 2932–2939, 2020.
View at: Publisher Site | Google Scholar
D. Miclăuş, “An approximation of the surface areas using the classical Bernstein quadrature formula,” Mathematical Methods in the Applied Sciences, vol. 42, no. 16, pp. 5317–5330, 2019.
View at: Google Scholar
M. I. Bhatti and P. Bracken, “Solutions of differential equations in Bernstein polynomial basis,” Journal of Computer and Applications in Mathematics, vol. 205, pp. 272–280, 2017.
View at: Google Scholar
A. Il’inskii and S. Ostrovska, “Convergence of generalized Bernstein polynomials,” Journal of Approximation Theory, vol. 116, no. 1, pp. 100–112, 2002.
View at: Google Scholar
N. Mahmudov, “The moments for q-Bernstein operators in the case 0,” Numerical Algorithms, vol. 53, no. 4, pp. 439–450, 2010.
View at: Publisher Site | Google Scholar
K. Maleknejad, E. Hashemizadeh, and R. Ezzati, “A new approach to the numerical solution of Volterra integral equations by using Bernstein’s approximation,” Communications in Nonlinear Science and Numerical Simulation, vol. 16, no. 2, pp. 647–655, 2011.
View at: Publisher Site | Google Scholar
S. Ostrovska, “The convergence ofq-Bernstein polynomials,” Mathematische Nachrichten, vol. 282, no. 2, pp. 243–252, 2009.
View at: Publisher Site | Google Scholar
G. M. Phillips, Interpolation and Approximation by Polynomials, Springer, New York, NY, USA, 2003.
F. Mirzaee and S. F. Hoseini, “Hybrid functions of Bernstein polynomials and block-pulse functions for solving optimal control of the nonlinear Volterra integral equations,” Indagationes Mathematicae, vol. 27, no. 3, pp. 835–849, 2016.
View at: Publisher Site | Google Scholar
F. Mirzaee, S. Alipour, and N. Samadyar, “A numerical approach for solving weakly singular partial integro-differential equations via two-dimensional-orthonormal Bernstein polynomials with the convergence analysis,” Numerical Methods for Partial Differential Equations, vol. 35, no. 1, pp. 615–637, 2018.
View at: Publisher Site | Google Scholar
F. Mirzaee and N. Samadyar, “Convergence of 2D-orthonormal Bernstein collocation method for solving 2D-mixed Volterra-Fredholm integral equations,” Transactions of A. Razmadze Mathematical Institute, vol. 172, no. 3, pp. 631–641, 2018.
View at: Publisher Site | Google Scholar
F. Mirzaee and N. Samadyar, “Application of orthonormal Bernstein polynomials to construct a efficient scheme for solving fractional stochastic integro-differential equation,” Optik, vol. 132, pp. 262–273, 2017.
View at: Publisher Site | Google Scholar
Ç. Dişibüyük and H. Oruç, “A generalization of rational Bernstein–Bezier curves,” BIT Numerical Mathematics, vol. 47, pp. 313–323, 2007.
View at: Google Scholar
G. Farin, Curves and Surfaces for Computer-Aided Geometric Design, Academic Press, Cambridge, MA, USA, 1996.
R. T. Farouki and T. N. T. Goodman, “On the optimal stability of the Bernstein basis,” Mathematics of Computation, vol. 65, no. 216, pp. 1553–1566, 1996.
View at: Publisher Site | Google Scholar
R. T. Farouki and V. T. Rajan, “Algorithms for polynomials in Bernstein form,” Computer Aided Geometric Design, vol. 5, no. 1, pp. 1–26, 1988.
View at: Publisher Site | Google Scholar
A. Beurling and J. Deny, “Espaces de dirichlet: I. Le casélémentaire,” Acta Mathematica, vol. 99, 1958.
View at: Publisher Site | Google Scholar
C. Berg, J. P. R. Christensen, and P. Ressel, Harmonic Analysis on Semigroups, Springer, New York, NY, USA, 1984.
C. Berg, K. Boyadzhiev, and R. deLaubenfels, “Generation of generators of holomorphic semigroups,” Journal of the Australian Mathematical Society. Series A. Pure Mathematics and Statistics, vol. 55, no. 2, pp. 246–269, 1993.
View at: Publisher Site | Google Scholar
J. Bertoin, Lévy Processes, Cambridge University Press, New York, NY, USA, 1996.
P. Benard, Lectures on Probability and Statistics, Ecole d’Eté de Probabilités de Saint-Flour XXVII-1997, Springer, New York, NY, USA, 1999.
F. Mirzaee, “Numerical solution of system of linear integral equations via improvement of block-pulse functions,” Journal of Mathematical Modeling, vol. 4, no. 2, pp. 133–159, 2016.
View at: Google Scholar
A.-D. Filip and A. Petruşel, “Existence and uniqueness of the solution for a general system of Fredholm integral equations,” Mathematical Methods in the Applied Sciences, pp. 1–10, 2020.
View at: Publisher Site | Google Scholar
M. Ramadan and H. S. Osheba, “A new hybrid orthonormal Bernstein and improved block-pulse functions method for solving mathematical physics and engineering problems,” Alexandria Engineering Journal, vol. 59, no. 5, pp. 3643–3652, 2020.
View at: Publisher Site | Google Scholar
H. Almasieh and M. Roodaki, “Triangular functions method for the solution of Fredholm integral equations system,” Ain Shams Engineering Journal, vol. 3, no. 4, pp. 411–416, 2012.
View at: Publisher Site | Google Scholar
K. Maleknejad, M. Shahrezaee, and H. Khatami, “Numerical solution of integral equations system of the second kind by Block-Pulse functions,” Applied Mathematics and Computation, vol. 166, no. 1, pp. 15–24, 2005.
View at: Publisher Site | Google Scholar
M. Ramadan, H. Oshaba, and R. Kharabsheh, “Triangular functions based method for the solution of system of linear Fredholm integral equations via an efficient finite iterative algorithm,” Journal of Intelligent and Fuzzy Systems, vol. 38, no. 3, pp. 2847–2858, 2020.
View at: Publisher Site | Google Scholar
E. Babolian, J. Biazar, and A. R. Vahidi, “The decomposition method applied to systems of Fredholm integral equations of the second kind,” Applied Mathematics and Computation, vol. 148, no. 2, pp. 443–452, 2004.
View at: Publisher Site | Google Scholar
P. Assari, “Solving weakly singular integral equations utilizing the meshless local discrete collocation technique,” Alexandria Engineering Journal, vol. 57, no. 4, pp. 2497–2507, 2018.
View at: Publisher Site | Google Scholar
C. Y. Gu, G. C. Wu, and B. Shiri, “An inverse problem approach to determine possible memory length of fractional differential equations,” Fractional Calculus and Applied Analysis, vol. 24, no. 6, pp. 1919–1936, 2021.
View at: Publisher Site | Google Scholar
G. Yang, B. Shiri, and H. Kong, “Intermediate value problems for fractional differential equations,” Computations and Applied Mathematics, vol. 40, no. 195, 2021.
View at: Publisher Site | Google Scholar
B. Shiri, G. C. Wu, and D. Baleanu, “Terminal value problems for the nonlinear systems of fractional differential equations,” Applied Numerical Mathematics, vol. 170, pp. 162–178, 2021.
View at: Publisher Site | Google Scholar
P. Assari, “On the numerical solution of two-dimensional integral equations using a meshless local discrete Galerkin scheme with error analysis,” Engineering with Computers, vol. 35, no. 3, pp. 893–916, 2019.
View at: Publisher Site | Google Scholar

Copyright

Copyright © 2022 Ji-Huan He et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

PDF Download Citation

Download other formats

Order printed copies