Dr.K.R.Salini[1] and Dr.S.Bamini[2]
[1] Guest Lecturer, PG & Department of Mathematics, Government Arts & Science College, Kariyampatti, Tirupattur, Tamilnadu, India.
Email: shalinimanoj1985@gmail.com
[2] Research Coordinator, Assistant Professor, PG & Research Department of Mathematics, Marudhar Kesari Jain College For Women, Vaniyambadi, Tamilnadu, India.
Email: saransham07@gmail.com
Abstract
We determine the numerical solution of the specific nonlinear Voltera-Fredholm Integro-Differential Equation is proposed. The method is based on new vector forms for representation of triangular functions and its operational matrix.
Introduction
A set of basic functions and an appropriate projection method or a direct method. These methods often transform an integro-differential equation to a linear or nonlinear system of algebraic equations which can be solved by direct or iterative methods.
Consider a Volterra-Fredhlom Integro-differential equations of the form π₯′(π )+π(π )π₯(π )+π1∫π1(π ,π‘)πΉ(π₯(π‘))ππ‘10+π2∫π2(π ,π‘)πΊ(π₯(π‘))ππ‘10=π¦(π )
π₯(0)=π₯0
Where the functions πΉ(π₯(π‘)) and πΊ(π₯(π‘)) are polynomials of π₯(π‘) with constant coefficients.
For convenience, we put πΉ(π₯(π‘))=[π₯(π‘)]π1 and πΊ(π₯(π‘))=[π₯(π‘)]π2, where π1 & π2 are positive integers.
For π1,π2=1, equation (1.1) is a linear integro- differential equation.
We present new vector forms of triangular functions (TFs), operational matrix of integration, expansion of functions of one and two variables with respect to TFs and other TFs properties.
By using new representations, a nonlinear integro-differential equation can be easily reduced to a nonlinear system of algebraic equations.
Journal of the Maharaja Sayajirao University of Baroda ISSN: 0025-0422
68
Volume-56, No.1 (IV) 2022
Triangular Functions
Definition
Two π-sets of triangular functions are defined over the intercal [0,π‘) as π1π(π‘)={1−π‘−πββπβ≤π‘<(π+1)β0ππ‘βπππ€ππ π π2π(π‘)={π‘−πββπβ≤π‘<(π+1)β0ππ‘βπππ€ππ π
Where π=0,1,2…π−1 with a positive integer value for m.
Also consider β=π/π, and π1π as the ππ‘β left-handed triangular function and π2π as
the ππ‘β right-handed triangular function.
It is assumed that π=1, so TFs are defined over [0,1) and β=1/π.
From the definition of TFs, it is clear that triangular functions are disjoint, orthogonal
and complete.
We can write ∫π1π10(π‘)π1π(π‘)ππ‘=∫π2π10(π‘)π2π(π‘)ππ‘={β3π=π0π≠π
Also, ∅π(π‘)=π1π(π‘)+π2π,π=0,1,…π−1
Where ∅π(π‘) is the πth block-pulse function defined as ∅π(π‘)={1πβ≤π‘<(π+1)β0ππ‘βπππ€ππ π
Where π=0,1,…π−1.
Vector Forms
Consider the first π terms of the left-handed triangular functions and the first π terms of the right- handed triangular functions and write them concisely as π-vectors. π1(π‘)=[π10(π‘),π11(π‘),…π1π−1(π‘)]π
π2(π‘)=[π20(π‘),π21(π‘),…π2π−1(π‘)]π
Where π1(π‘) and π2(π‘) are called left-handed triangular functions (LHTF) and right-handed triangular functions (RHTF) vector, respectively.
The product of two TFs vectors are
Journal of the Maharaja Sayajirao University of Baroda ISSN: 0025-0422
69
Volume-56, No.1 (IV) 2022
π1(π‘)π1π(π‘)≅( π10(π‘)0⋯⋯⋯00π11(π‘)⋯⋯⋯0⋮⋮0⋮⋮0⋮⋮⋯⋮⋮⋯⋮⋮⋯⋮⋮π1π−1(π‘))
π2(π‘)π2π(π‘)≅( π20(π‘)0⋯⋯⋯00π21(π‘)⋯⋯⋯0⋮⋮0⋮⋮0⋮⋮⋯⋮⋮⋯⋮⋮⋯⋮⋮π2π−1(π‘)) π1(π‘)π2π(π‘)=0
π2(π‘)π1π(π‘)=0
Where 0 is the π§πππ π×π matrix.
Also ∫π110(π‘)π1π(π‘)ππ‘=∫π2(π‘)π1π(π‘)ππ‘10≅β3πΌ
∫π110(π‘)π2π(π‘)ππ‘=∫π2(π‘)π1π(π‘)ππ‘10≅β6πΌ
In which πΌ is an π×π identity matrix.
TFs Expansion
The expansion of a function π(π‘) over [0,1) with respect to TFs written as π(π‘)≅Ξ£πππ−1π=0π1π(π‘)+Ξ£πππ−1π=0π2π(π‘) π(π‘)=πππ1(π‘)+πππ2(π‘)
Where we may put ππ=π(πβ) and ππ=π((π+π)β) for π=0,1,2,..π−1.
Operational Matrix of Integration
Expressing ∫π1(π)ππ≅π1π1(π )+π2π2(π )π 0
∫π2(π)ππ≅π1π1(π )+π2π2(π )π 0
Where π1π×π and π2π×π are called operational matrices of integration in TFs domain and represented as follows π1=β2( 011⋯1001⋯10⋮00⋮00⋮0⋯⋱⋯1⋮0)
Journal of the Maharaja Sayajirao University of Baroda ISSN: 0025-0422
70
Volume-56, No.1 (IV) 2022
π2=β2( 111⋯1011⋯10⋮00⋮00⋮0⋯⋱⋯1⋮1)
So the integral of any function π(π‘) can be approximated as ∫π(π)ππ≅πππ1(π )+πππ2(π )πππ 0 ∫π(π)ππ≅(π+π)ππ1π1(π )+(π+π)ππ2π2(π )π 0
New Representation of TFs Vector Forms and Other Properties
We define a new representation of TFs vector forms. Then, some characteristics of TFs are presented using the new definition.
Definition and Expansion
Let π(π‘) be a 2π-vector defined as π(π‘)=(π1(π‘)π2(π‘)),0<π‘<1
Where π1(π‘) and π2(π‘) have been defined in (1.2).
Now, the expansion of π(π‘) with respect to TFs can be written as π(π‘)≅πΉ1ππ1(π‘)+πΉ2ππ2(π‘)
≅πΉππ(π‘)
Where πΉ1 &πΉ2 are TFs co-efficients with
πΉ1π=π(πβ)& πΉ2π=π((π+1)β),
For π=0,1,…π−1.
Also, 2π-vector πΉ is defined as πΉ=(πΉ1πΉ2)
Now, assume that π(π ,π‘) is a function of two variables. It can be expanded with respect to TFs as follows
π(π ,π‘)≅ππ(π )πΎπ(π‘)
Where π(π )& π(π‘) are 2π1&2π2 dimensional triangular functions and πΎ is a
2π1×2π2 TFs coefficient matrix.
For convenience, we put π1=π2=π. So matrix πΎ can be written as
Journal of the Maharaja Sayajirao University of Baroda ISSN: 0025-0422
71
Volume-56, No.1 (IV) 2022
πΎ=((πΎ11)π×π(πΎ12)π×π(πΎ21)π×π(πΎ22)π×π)
Where πΎ11,πΎ12,πΎ21,πΎ22 can be compted by sampling the function π(π ,π‘) at points π π & π‘π such that π π=πβ & π‘π=πβ, for π,π=0,1,…π. (πΎ11)ππ=π(π π,π‘π),π=0,1,…π−1,π=0,1,…π−1 (πΎ12)ππ=π(π π,π‘π),π=0,1,…π−1,π=0,1,…π (πΎ21)ππ=π(π π,π‘π),π=0,1,…π,π=0,1,…π−1 (πΎ22)ππ=π(π π,π‘π),π=0,1,…π,π=0,1,…π
Product Properties
Let π be an 2π-vertor which can be written as ππ=(π1π π2π) such that π1 & π2 are π-vectors.
It can be concluded from (1.3) & (1.4) that π(π‘)ππ(π‘)π=(π1(π‘)π2(π‘))(ππ(π‘)π2π(π‘))(π1π2)
≅(πππ(π1(π‘))ππ×πππ×ππππ(π2(π‘)))(π1π2)
=πππ (π(π‘))π
=πππ(π)π(π‘)
Therefore,
π(π‘)ππ(π‘)π≅π̅ π(π‘)
Where π̅=πππ(π) is an 2π×2π diagonal matrix.
Let π΅ be a 2π×2π matrix as π΅=((π΅11)π×π(π΅12)π×π(π΅21)π×π(π΅22)π×π)
So, it can be similarly concluded from equations (1.3) and (1.4) that π(π‘)π΅π(π‘)=π1π(π‘)π2π(π‘)(π΅11π΅12π΅21π΅22)(π1(π‘)π2(π‘)) ≅π1π(π‘)π΅11π1(π‘)+π2π(π‘)π΅22π2(π‘) ≅π΅̂11ππ1(π‘)+π΅̂22ππ2(π‘)
Where π΅̂11 and π΅̂22 are π−vectors with elements equal to the diagonal entries of matrices π΅11 and π΅22 respectively.
Therefore,
ππ(π‘)π΅π(π‘)≅π΅̂ ππ(π‘)
Journal of the Maharaja Sayajirao University of Baroda ISSN: 0025-0422
72
Volume-56, No.1 (IV) 2022
In which π΅̂ is a 2π- vector with elements equal to the diagonal entries of matrix π΅.
It is immediately concluded from equation (1.5) that ∫π10(π‘)ππ(π‘)ππ‘=∫(π1(π‘)π2(π‘))π1π(π‘)π2π(π‘)ππ‘10
=∫(π1(π‘)π1π(π‘)π1(π‘)π2π(π‘)π2(π‘)π1π(π‘)π2(π‘)π2π(π‘))ππ‘10
≅(β3πΌπ×πβ6πΌπ×πβ6πΌπ×πβ3πΌπ×π)
Therefore,
∫π10(π‘)ππ(π‘)ππ‘≅π·
Where π· is the following 2π×2π matrix. π·=β3( 10⋯012⁄0⋯001⋯0012⁄⋯0⋮012⁄0⋮0⋮0012⁄⋮0⋱⋯⋯⋯⋱⋯⋮100⋮12⁄⋮010⋮0⋮001⋮0⋱…⋯⋯⋱⋯⋮12⁄00⋮1)
Operational Matrix
Expressing ∫π(π)πππ 0 in terms of π(π ), we can write ∫π(π)πππ 0=∫(π1(π)π2(π))πππ 0
≅(π1π1(π )+π2π2(π )π1π1(π )+π2π2(π ))
=(π1π2π1π2)(π1(π )π2(π ))
So, ∫π(π)πππ 0≅ππ(π )
Where π2π×2π, operational matrix of π(π ) is
π=(π1π2π1π2) (
Journal of the Maharaja Sayajirao University of Baroda ISSN: 0025-0422
73
Volume-56, No.1 (IV) 2022
Where π1 & π2 are given by
The integral of any function π(π‘) can be approximated as ∫π(π)πππ 0≅∫πΉππ(π)πππ 0
≅πΉπππ(π )
Solving Nonlinear Integro-Differential Equation
Consider the following nonlinear Volterra-Fredhlom integro-differential equation
{π₯′(π )+π(π )π₯(π )+π1∫π1(π ,π‘)[π₯(π‘)]π1ππ‘+π 0π2∫π2(π ,π‘)[π₯(π‘)]π2ππ‘=π¦(π )π 0π₯(0)=π₯0, 0≤π <1,π1,π2≥1
Where the parameters π1 and π2 and β2 functions π(π ),π¦(π ),π1(π ,π‘),π2(π ,π‘) are known but π₯(π ) is not.
The appearance of initial condition equation. This is necessary to ensure the existence of a solution.
Approximating functions
π₯(π ),π₯′(π ),π(π ),π¦(π ),[π₯(π‘)]π1,[π₯(π‘)]π2,π1(π ,π‘),π2(π ,π‘)
with respect to TFs, π₯(π )≅π₯π(π )π π₯′(π )≅π′ππ(π )=πππ′
π(π )≅ πππ(π )=ππ(π )π
π¦(π )≅ πππ(π )=ππ(π )π [π₯(π‘)]π1≅ππ1ππ(π )=ππ(π )ππ1 [π₯(π‘)]π2≅ππ2ππ(π )=ππ(π )ππ2 π1(π ,π‘)≅πππΎ1π(π‘)
π2(π ,π‘)≅πππΎ2π(π‘)
Where 2π-vectors π,π′,π,π,ππ1,ππ2 and 2π×2π matrices πΎ1 & πΎ2 are TFs coefficients of
π₯(π ),π₯′(π ),π(π ),π¦(π ),[π₯(π‘)]π1,[π₯(π‘)]π2,π1(π ,π‘),π2(π ,π‘) respectively.
Lemma
Let 2π-vectors π and ππ be TFs coefficients of π₯(π ) and [π₯(π )]π respectively. If
π=(π1ππ2π)π=(π10,π11,…,π1π−1,π20,π21…π2π−1)π, then
ππ=(π10,π11,…,π1π−1,π20,π21…π2π−1)π
Where π≥1 is a positive integer.
Proof:
Journal of the Maharaja Sayajirao University of Baroda ISSN: 0025-0422
74
Volume-56, No.1 (IV) 2022
When π=1, , follows at once from [π₯(π )]π=π₯(π ).
Suppose that (4.16), holds for π, we shall deduce it for π+1.
Since [π₯(π )]π+1≅(πππ(π )).(π1ππ(π ))
=πππ(π )ππ(π )ππ =ππππ̃π(π )
we obtain ππππ̃=(π10π+1,π11π+1,……,π1π−1π+1,π20π+1,π21π+1,……,π2π−1π+1)π
Equation holds for π+1.
Components of πΏπ can be Computed in terms of Components of Unknown Vector πΏ
We substitute (1.15) into (1.14), we have πππ(π )≅π′ππ(π )+πππ(π )ππ(π )π+π1ππ(π )πΎ1∫π(π‘)ππ(π‘)ππ1ππ‘+π 0π2ππ(π )πΎ2∫π(π‘)ππ(π‘)ππ2ππ‘π 0
it follows that πππ(π )≅π′ππ(π )+(π̃ π(π ))ππ+π1ππ(π )πΎ1π̃π1∫π(π‘)ππ‘+π2ππ(π )πΎ2π·ππ2π 0
Using Operational matrix π,
πππ(π )≅π′ππ(π )+πππ̃ π(π )+π1ππ(π )πΎ1π̃π1ππ(π )+π2(πΎ2π·ππ2)ππ(π )
In which π1πΎ1π̃π1π is a 2π×2π matrix
ππ(π )π1πΎ1π̃π1ππ(π )≅ππ1π̂π(π )
Where π̃π1 is an 2π-vector with components equal to the diagonal entries of the matrix π1πΎ1π̃π1π.
πππ(π )≅π′ππ(π )+πππ̃ π(π )+ππ1π̂π(π )+π2(πΎ2π·ππ2)ππ(+π̃ π+ππ1̂+ π2πΎ2π·ππ2≅π
Where π̃ is a diagonal matrix, so π̃ π=π̃.
π′ must be computed in terms of π. π₯(π )−π₯(0)=∫π₯′(π)πππ 0
≅∫π′ππ(π)πππ 0 ≅π′πππ(π ) π₯(π )≅π′πππ(π )+π0ππ(π )
Where π0 is the 2π- vector of the form
Journal of the Maharaja Sayajirao University of Baroda ISSN: 0025-0422
75
Volume-56, No.1 (IV) 2022
π0=[π₯0,π₯0,…π₯0]π
Consequently,
π≅πππ′+π0
Now, combining (1.19) and (1.20) and replacing ≅ with =, it follows that
(πΌ+πππ̃)π+ππππ1̂+π2πππΎ2π·ππ2≅πππ+π0
Equation is a nonlinear system of 2π algebraic equations for the 2π unknowns π10,π11,…π1π−1,π20,π21,…π2π−1.
Components of ππ=(π1π π2π) can be obtainbed by an iterative method.
Hence, an approximate solution π₯(π )≅πππ(π )
π₯(π )≅π1ππ1(π )+π2ππ2(π )
Can be computed for equation (1.14) without using any projection method.
Conclusion
We investigated the numerical solution of nonlinear and singularly perturbed for Vulture integro-differential equations. Also, numerical solution of high-order and fractional nonlinear Volterra-Fredholm integro-differential equations and its error analysis are discussed. Finally, some numerical results of integro-differential equations are presented to illustrated the efficiency and accuracy of the proposed methods.
Reference
1. G.Yu.Mekhtieva, V.R.Ibrahimov “On one numerical method for solution of integro-differential equations”, Vestnik BSU, ser.phys.-math. Scien., 2003, No3, pp.21-27.(Russian).
2. H.Brunner, “Collection method for Volterra integral and Relation Functional Equations”, Cambridge University Press, Cambridge,2004.
3. P.Darania, E.Abadian, “A method for the numerical solution of the integro-differential equations”, Appl.Math. and Comput. 188(2007) 657-668.
4. B. Ahmad, S. Siva Sundaram, Existence Results For Nonlinear Impulsive Hybrid Boundary Value Problems involving Fractional Differential Equations, Non linear Analysis, hybrid System, 3(2009).
5. Z.B. Bai, H.S. Lii, Positive Solutions Of Boundary Valuve problems Of Nonlinear Fractional Differential Equations, J. Math Anal.Appl.311(2005)\
6. v. Lakshmikantham, D.D. Bainov And P.S. Simeonov, Theory Of Impulsive Differential Equations, modern Applied Mathematics (1989).
7. K.S. Miller, B.Ross, An Introduction To The Fractional Caculus And Fractional Differential Equations, Wiley, New york, 1993
No comments:
Post a Comment