analytical and numerical methods for solving linear fuzzy ...1.6 fuzzy integral equations 20 chapter...
TRANSCRIPT
An-Najah National University
Faculty of Graduate Studies
Analytical and Numerical Methods for Solving Linear
Fuzzy Volterra Integral Equation of the Second Kind
By
Jihan Tahsin Abdal-Rahim Hamaydi
Supervisor
Prof .Naji Qatanani
This Thesis is Submitted in Partial Fulfillment of the Requirements for
the Degree of Master of Mathematics, Faculty of Graduate Studies,
An-Najah National University, Nablus, Palestine.
2016
iii
Dedication
I dedicate this thesis to my beloved homeland, to my parents, to my dear
husband Sami, to my daughter Shatha, to my sons Hothayfa and yaman, to
my sisters specially Malaak and my brothers, to my friend Aminah Alqub ,
to my school and my students, to everyone who supports and encourages
me.
iv
Acknowledgement
First I thank God to complete this thesis, and then extend my sincere thanks
to my supervisor Prof. Naji Qatanani who encouraged me all the time and
gifted me a lot of self- confidence, and I owe to his effort a great deal
because he helped me a lot to achieve my goal in this thesis. Also, l have
to thank my family who supported and encouraged me.
Finally, I would like to acknowledge my teachers at Al-Najah National
University- Mathematical Department.
vi
Table of Contents
No. Contents Page
Declaration ii
Dedication iii
Acknowledgement
iv
List of Figures viii
List of Tables ix
Abstract
x
Introduction
1
Chapter One 4
Mathematical Preliminaries 4
1.1 Crisp and Fuzzy sets 4
1.2 Operations on fuzzy sets 6
1.3 Fuzzy numbers 8
1.4 Fuzzy function 13
1.5 Fuzzy linear systems 17
1.6 Fuzzy integral equations 20
Chapter Two 24 Analytical methods for Solving Fuzzy Volterra Integral
Equation of the Second Kind 24
2.1 Fuzzy Laplace transformation method 24
2.2 Fuzzy Convolution 28
2.3 Fuzzy homotopy analysis method 33
2.4 Adomian decomposition method 45
2.5 Fuzzy differential transformation method 49
2.6 Fuzzy successive approximation method
65
Chapter Three 69
Numerical Methods for Solving Fuzzy Volterra Integral
Equation of the Second Kind 69
3.1 Taylor expansion method 69
3.2 Convergence analysis 76
vii
3.3 Trapezoidal method 80
3.4 Variation iteration method 83
Chapter Four 88
Numerical Examples and Results 88
Conclusion 123
References 124
Appendix 132
ุจ ุงูู ูุฎุต
viii
List of Figures
Figure Title Page
1.1 The support, core, band width, and height of the
fuzzy set 5
1.2 The crisp number 6
1.3 The convex and non-convex fuzzy sets 8
1.4 Triangular and trapezoidal fuzzy numbers 11
4.1(๐) The exact and approximate solutions in example
(4.1) ๐๐ก ๐ฅ = 0.5 93
4.1(๐) The absolute error between the exact and
numerical solutions in example (4.1)๐๐ก ๐ฅ = 0.5 94
4.2(๐) The exact and approximate solutions in example
(4.2)๐๐ก ๐ฅ =๐
8
100
4.2(๐) The absolute error between the exact and
numerical solutions in example (4.2)๐๐ก ๐ฅ =๐
8
101
4.3(๐) The exact and approximate solutions in example
(4.3)๐๐ก ๐ฅ = 0.84 105
4.3(๐) The absolute error between the exact and
numerical solutions in example (4.3)๐๐ก ๐ฅ = 0.84 105
4.4(๐) The exact and approximate solutions in example
(4.4)๐๐ก ๐ฅ = 0.23๐ 108
4.4(๐)
The absolute error between the exact and
numerical solutions in example (4.4)๐๐ก ๐ฅ =0.23๐
108
4.5(๐) The exact and approximate solutions in example
(4.5)๐๐ก ๐ฅ = 0.5 114
4.5(๐) The absolute error between the exact and
numerical solutions in example (4.5)๐๐ก ๐ฅ = 0.5 115
4.6(๐) The exact and approximate solutions in example
(4.6)๐๐ก ๐ฅ =๐
8
120
4.6(๐) The absolute error between the exact and
numerical solutions in example (4.6)๐๐ก ๐ฅ =๐
8
120
ix
List of Tables
Table Title Page
4.1 The exact and numerical solutions for ๐ข, and
the resulted error by algorithm (4.1)๐๐ก ๐ฅ = 0.5 95
4.2 The exact and numerical solutions for ๐ข, and
the resulted error by algorithm (4.1)๐๐ก ๐ฅ =๐
8
102
4.3
The exact and numerical solutions for ๐ข, and
the resulted error by algorithm (4.2)๐๐ก ๐ฅ =0.84
106
4.4
The exact and numerical solutions for ๐ข, and
the resulted error by algorithm (4.2)๐๐ก ๐ฅ =0.23๐
109
4.5 The exact and numerical solutions for ๐ข, and
the resulted error in example (4.3)๐๐ก ๐ฅ = 0.5 116
4.6 The absolute error between the exact and
numerical solutions in example (4.3)๐๐ก ๐ฅ =๐
8
121
x
Analytical and Numerical Methods for Solving
Linear Fuzzy Volterra Integral Equation of the Second Kind
By
Jihan Tahsin Abdel Rahim Hamaydi
Supervised
Prof. Naji Qatanani
Abstract
Integral equations, in general, play a very important role in Engineering
and technology due to their wide range of applications. Fuzzy Volterra
integral equations in particular have many applications such as fuzzy
control, fuzzy finance and economic systems.
After introducing some definitions in fuzzy mathematics, we focus our
attention on the analytical and numerical methods for solving the fuzzy
Volterra integral equation of the second kind.
For the analytical solution of the fuzzy Volterra integral equation we have
presented the following methods:
The Fuzzy Laplace Transformation Method(FLTM), Fuzzy Homotopy
Analysis Method(FHAM), Fuzzy Adomian Decomposition Method
(FADM), Fuzzy Differential Transformation Method (FDTM), and the
Fuzzy Successive Approximation Method (FSAM).
For the numerical handling of the fuzzy Volterra Integral equation we have
implemented various techniques, namely: Taylor expansion method,
Trapezoidal method, and the variation iteration method.
xi
To investigate the efficiency of these numerical techniques we have solved
some numerical examples.
Numerical results have shown to be in a close agreement with the
analytical ones.
Moreover, the variation iteration method is one of the most powerful
numerical techniques for solving Fuzzy Volterra integral equation of the
second kind in comparison with other numerical techniques.
1
Introduction
Fuzzy integral equations of the second kind have attracted the attention of
many scientists and researchers in recent years, due to their importance in
applications, such as fuzzy control, fuzzy finance, approximate reasoning
and economic systems [5].
Prior to discussing fuzzy differential equations and integral equations and
their associated numerical algorithms, it is necessary to present an
appropriate brief introduction to preliminary topics, such as fuzzy numbers
and fuzzy calculus. The concept of fuzzy sets, was originally introduced by
Zadeh [48], led to the definition of fuzzy numbers and its implementation in
fuzzy control [17], and approximate reasoning problems [49].
The concept of integral of fuzzy functions was first introduced by Dubios
and Prade [21]. Alternative approaches were later suggested by Goetschel
and Voxman [30], Kaleva [33], Nanda [40] and others. One of the first
applications of fuzzy integration was given by Wu and Ma [18] who
investigated the fuzzy Fredholm integral equation of the second kind.
Liao [34] in 1992 employed the homotopy analysis method to solve non-
linear problems. In addition, the homotopy analysis method has been used
for solving fuzzy integral equations of the second kind. Babolian etal. [9]
have solved the fuzzy integral equation by the Adomian method. There are
also numerous numerical methods which have been focusing on the solution
of integral equations [23]. Amawi [6] has investigated some analytical and
2
numerical methods for solving fuzzy Fredholm integral equation of the
second kind. Amawi and Qatanani [7] have employed the Taylor expansion
method, the Trapezoidal rule, the Adomian method and the homotopy
analysis method to solve various fuzzy Fredholm integral equations of the
second kind.
This thesis is organized as follows:
In chapter one, we introduce some basic concepts of fuzzy sets, crisp sets,
fuzzy numbers, and fuzzy integral equation.
In chapter two, we investigate some analytical methods to solve linear
fuzzy Volterra integral equation of the second kind, namely: Fuzzy Laplace
transformation method(FLTM), Fuzzy Homotopy analysis
method(FHAM), Fuzzy Adomian decomposition method(FADM), and
Fuzzy successive approximation method.
In chapter three, we employ some numerical methods to solve linear fuzzy
Volterra integral equation of the second kind, these include:
Fuzzy Taylor expansion method, Fuzzy Trapezoidal method, and Fuzzy
variational iteration method.
In chapter four, MAPLE software has been constructed to solve numerical
examples to demonstrate the efficiency of these numerical schemes
introduced in chapter three.
3
Finally, we draw a comparison between analytical and numerical solutions
for some numerical examples.
4
Chapter One
Mathematical Preliminaries
Basic Definitions and Operations.
1.1 Crisp and Fuzzy Sets
Definition )1.1( [14]: Characteristic function: Let ๐ be a set and A be a
subset of ๐ (๐ด โ ๐). Then the characteristic function of the set ๐ด in ๐,
where ๐๐จ โถ ๐ โ {0,1} is defined by:
1 if ๐ฅโA
๐๐จ(๐) =
0 if ๐ฅ โ A
Definition )1.2( [48]: Fuzzy set: a Fuzzy set is a set whose element have
degrees of membership.
It is extension of the classical notion of set.
The value 0 is used to represent complete non-membership, the value 1 is
used to represent complete membership, and values in between are used to
represent intermediate degrees of membership.
Definition )1.3( [12]: The support of fuzzy set: The support of a fuzzy
set ๐ด is defined by :
๐๐ข๐๐(๐ด) = {๐ฅ๐ ๐: ๐ด(๐ฅ) > 0}.
5
A(x)
Core
1
0.5 ----------Band width------ Height
x
Support
๐น๐๐๐ข๐๐ (1.1)
๐ถ๐๐๐ ๐๐๐ด = {๐ฅ: ๐ด(๐ฅ) = 1}
๐๐ข๐๐๐๐๐ก ๐๐ ๐ด = {๐ฅ: ๐ด(๐ฅ) > 0}
๐ต๐๐๐ ๐ค๐๐๐กโ ๐๐ ๐ด = {๐ฅ: ๐ด(๐ฅ) โฅ 0.5}
๐ป๐๐๐โ๐ก ๐๐ ๐ด = ๐ ๐ข๐๐ฅโ๐
๐ด(๐ฅ)
Definition)1.4( [25 ]: ๐ถ โ cut: An ๐ผ-level set of a fuzzy set ๐ด of ๐ is a
non-fuzzy set denoted by [๐ด] ๐ผ and is defined by:
[๐ด]๐ผ = { ๐ฅ โ ๐: ๐ด(๐ฅ) โฅ ๐ผ , ๐๐๐ผ > 0
๐๐(๐ ๐ข๐๐(๐ด)) , ๐๐ ๐ผ = 0
where ๐๐ (๐ ๐ข๐๐(๐ด)) = ๐ถ๐๐๐ ๐ข๐๐ ๐๐ ๐กโ๐ ๐ ๐ข๐๐๐๐๐ก ๐ด.
Definition )1.5 ( [ 25]: Crisp number: A crisp number a is represented by:
1 ๐๐ ๐ฅ = ๐
๐ด(๐ฅ) =
0 ๐๐ ๐ฅ โ ๐
6
a crisp interval [๐, ๐] is represented by a fuzzy set:
1 ๐๐ ๐ฅ โ [๐, ๐]
๐ต(๐ฅ) =
0 ๐๐ ๐ฅ โ [๐, ๐]
๐ด(๐ฅ) ๐ต(๐ฅ) 1
a c d X
The crisp number ๐น๐๐๐ข๐๐ (1.2)
1.2 Operations on Fuzzy Sets
As in the case of ordinary sets, the operations on fuzzy sets play a control
role in the case of fuzzy sets, Zadeh [48] defined the following operations
for fuzzy sets as generalization of crisp sets and of crisp statements.
Definition (1.6): The intersection: the membership function of the
intersection of two fuzzy sets ๐ด and ๐ต is defined as:
(๐ด โฉ ๐ต) (๐ฅ) = ๐๐๐ {๐ด(๐ฅ), ๐ต(๐ฅ)}, โ ๐ฅ โ ๐.
or, in abbreviated form: (๐ด โฉ ๐ต) (๐ฅ) = ๐ด (๐ฅ) โ ๐ต(๐ฅ)
7
Definition (1.7): The union: the membership function of the union of two
fuzzy sets ๐ด and B is defined as: (๐ด โช ๐ต) (๐ฅ) = ๐๐๐ฅ {๐ด(๐ฅ), ๐ต(๐ฅ)}, โ๐ฅ โ
๐.
or, in abbreviated form: (๐ด โช ๐ต) (๐ฅ) = ๐ด (๐ฅ) โ ๐ต(๐ฅ)
Definition (1.8): The complement: the membership function of
complement is defined as : ๐ด(๐ฅ)๐ = 1 โ ๐ด(๐ฅ), ๐๐๐ ๐๐ฃ๐๐๐ฆ ๐ฅ๐ ๐.
Definition (1.9): The containment: A fuzzy set ๐ด is contained in a fuzzy
set ๐ต (๐ด โ ๐ต) if and only if ๐ด (๐ฅ) โค ๐ต(๐ฅ), โ๐ฅ โ ๐.
In symbols, ๐ด โ ๐ต โบ ๐ด (๐ฅ) โค ๐ต(๐ฅ).
Definition (1.10): The equality: The two fuzzy sets ๐ด and ๐ต are equal if
and only if
๐ด (๐ฅ) = ๐ต(๐ฅ), โ๐ฅ โ ๐.
In symbols, ๐ด = ๐ต โบ ๐ด (๐ฅ) = ๐ต(๐ฅ), โ๐ฅ โ ๐.
Definition (1.11): The empty fuzzy set: A fuzzy set ๐ด is empty if and only
if its membership function ๐ด(๐ฅ) = 0, โ๐ฅ โ ๐.
Some Properties of Fuzzy Sets
(A โช B)C = AC โฉ BC
De Morganโs laws
(A โฉ B)C = AC โช BC
8
๐ถ โฉ (A โช B) = ( ๐ถ โฉ A ) โช ( ๐ถ โฉ A ) Distributive laws
๐ถ โช (A โฉ B) = ( ๐ถ โช A ) โฉ ( ๐ถ โช A ) 1- Max{๐ด(๐ฅ), ๐ต(๐ฅ)}= Min {1 โ ๐ด(๐ฅ), 1 โ ๐ต(๐ฅ)} 1-Min{๐ด(๐ฅ), ๐ต(๐ฅ)}= Max {1 โ ๐ด(๐ฅ), 1 โ ๐ต(๐ฅ)}
1.3 Fuzzy Numbers
Definition (1.12) [48]: Normal fuzzy set: A fuzzy set ๐ด is called normal
if there exists ๐๐ ๐๐๐๐๐๐๐ก ๐ฅ โ ๐ ๐ ๐ข๐โ ๐กโ๐๐ก ๐ด(๐ฅ) = 1, Otherwise ๐ด is
subnormal.
Definition (1.12) [4]: Convex fuzzy set: A fuzzy set ๐ด is convex if:
๐ด (๐๐ฅ1 + (1 โ ๐) ๐ฅ2 ) โฅ ๐๐๐ {๐ด (๐ฅ1), ๐ด (๐ฅ2)}, โ ๐ฅ1, ๐ฅ2๐ ๐ , ๐๐๐
๐ ๐ [0,1]. ๐ด is convex if all its ๐ผ โ ๐๐ข๐ก๐ are convex.
๐ด (๐ ๐ฅ1 + (1 โ ๐)๐ฅ2)
1 1
๐ด(๐ฅ1) ๐ด(๐ฅ2) ๐ด(๐ฅ1) ๐ด(๐ฅ2)
๐ฅ1 ๐ฅ2 ๐ฅ1 ๐ฅ2
Convex fuzzy set non-Convex fuzzy set
๐น๐๐๐ข๐๐ (1.3)
9
Definition (1.13) [47]: Upper semi continuous: A function f: ๐ธ โถ ๐ is
said to be upper semi continuous at ๐ฅ 0 โ ๐ธ if :
โํ > 0, โ๐ฟ > 0 ๐ . ๐ก ๐(๐ฅ) = ๐(๐ฅ0) + ํ, Where ๐ฅ โ ๐ธ โฉ ๐ต๐ฟ(๐ฅ0).
Definition (1.14) [28]: Fuzzy number: A fuzzy number is a fuzzy set
๐ด: ๐ โ [0, 1] such that:
(๐) ๐ด is upper semi continuous,
(๐๐) ๐ด(๐ฅ) = 0 outside some interval [๐, ๐],
(๐๐๐) There are real numbers ๐, ๐: ๐ โค ๐ โค ๐ โค ๐, for which:
(1) ๐ด(๐ฅ) is monotonically increasing on [๐, ๐],
(2) ๐ด(๐ฅ) is monotonically decreasing on [๐, ๐],
(3) ๐ด(๐ฅ) = 1, ๐ โค ๐ฅ โค ๐.
Definition (1.15) [50]: Convex normalized fuzzy set: A fuzzy number ๐ด
is a convex normalized if:
1. there exists exactly one ๐ฅ0 โ ๐ : ๐ด (๐ฅ0) = 1 (๐ฅ0 is called the mean
value of A)
2. ๐ด (๐ฅ) is piecewise continuous.
Definition (1.16) [50]: The cardinality: For finite fuzzy set ๐ด, the
cardinality of A is defined as :
10
|๐จ| = โ๐จ(๐)
๐โ๐ฟ
โ๐จโ = |๐จ|
|๐ฟ| is called the relative cardinality of ๐ด .
Definition (1.17) [24]: Triangular fuzzy number: Triangular a fuzzy
number represented with three points as follows ๐ด = ( ๐1, ๐2, ๐3 )
This representation is interpreted as membership functions and holds the
following conditions:
(๐) ๐1 to ๐2 is increasing function.
(๐๐) ๐2 to ๐3is decreasing function.
(๐๐๐)๐1 โค ๐2 โค ๐3. 0 ๐๐๐ ๐ฅ < ๐1
๐ฅ โ ๐1
๐2 โ ๐1 ๐๐ ๐1 โค ๐ฅ โค ๐2
๐ด(๐ฅ) =
๐3 โ ๐ฅ
๐3 โ ๐2 ๐๐ ๐2 โค ๐ฅ โค ๐3
0 ๐๐๐ ๐ฅ > ๐3
Definition (1.18) [10]: Trapezoidal fuzzy number: We can define
trapezoidal fuzzy number ๐ด as ๐ด = ( ๐1, ๐2, ๐3, ๐4 )The membership of
this fuzzy number will be interpreted as follows:
0 ๐๐๐ ๐ฅ < ๐1
๐ฅ โ ๐2
๐2 โ ๐1 ๐๐ ๐1 โค ๐ฅ โค ๐2
๐ด(๐ฅ) = 1 ๐๐ ๐2 โค ๐ฅ โค ๐3
๐4 โ ๐ฅ
๐4 โ ๐3 ๐๐ ๐3 โค ๐ฅ โค ๐4
0 ๐๐๐ ๐ฅ > ๐4
11
a1 a2 a3 a1 a2 a3 a4
Triangular fuzzy number Trapezoidal fuzzy number
๐น๐๐๐ข๐๐ (1.4)
Definition (๐. ๐๐): Fuzzy metric: Given two fuzzy sets ๐ด, ๐ต. we try to
find a fuzzy number d*(๐ด, ๐ต) that should satisfy the following:
(๐) ๐โ(๐ด, ๐ต) โฅ 0, ๐๐๐ ๐โ(๐ด, ๐ต) โบ ๐ด = ๐ต
(๐๐) ๐โ(๐ด, ๐ต) = ๐โ(๐ต, ๐ด)
(๐๐๐) ๐โ(๐ด, ๐ถ) โค ๐โ(๐ด, ๐ต) + ๐โ(๐ต, ๐ถ), ๐๐๐ ๐๐๐ฆ ๐ ๐๐ก๐ ๐ด, ๐ต, ๐ถ
Definition (1.20) [2]: Parametric form of fuzzy number:
A fuzzy number ๐ฅ is a pair (๐ฅ, ๏ฟฝฬ ๏ฟฝ) of functions ๐ฅ (๐ผ), ๏ฟฝฬ ๏ฟฝ( ๐ผ); 0 โค ๐ผ โค 1
which satisfies the following:
๐. ๐ฅ(๐ผ)is a bounded left-continuous non-decreasing function over [0, 1]
๐๐. ๐ฅ(๐ผ)is a bounded left-continuous non-increasing function over [0, 1]
๐๐๐. ๐ฅ (๐ผ) โค ๐ฅ ฬ (๐ผ); 0 โค ๐ผ โค 1
12
A crisp number ๐ผ is simply represented by ๐ฅ(๐ผ) = ๏ฟฝฬ ๏ฟฝ(๐ผ) = ๐ผ,
0 โค ๐ผ โค 1.
arithmetic operations of arbitrary fuzzy numbers ๐ฅ = (๐ฅ, ๏ฟฝฬ ๏ฟฝ), ๐ฆ = (๐ฆ, ๐ฆ),
and ๐ โ ๐ , can be defined as [43]:
1. ๐ฅ = ๐ฆ ๐๐๐ ๐ฅ(๐ผ) = ๐ฆ(๐ผ)๐๐๐ ๐ฅ(๐ผ) = ๐ฆ(๐ผ),
2. ๐ฅ + ๐ฆ = [ ๐ฅ(๐ผ) + ๐ฆ(๐ผ), ๐ฅ(๐ผ) + ๐ฆ(๐ผ)],
3. ๐ฅ โ ๐ฆ = [๐ฅ(๐ผ) โ ๐ฆ(๐ผ), ๐ฅ(๐ผ) โ ๐ฆ(๐ผ)],
4. ๐๐ฅ = {๐๐ฅ(๐ผ), ๐๐ฅ(๐ผ), ๐ โฅ 0
๐๐ฅ(๐ผ), ๐๐ฅ(๐ผ), ๐ < 0
Definition (1.21) [3]: The distance: For arbitrary fuzzy numbers ๐ฅ =
(๐ฅ, ๐ฅ)๐๐๐ ๐ฆ = (๐ฆ, ๐ฆ) we define the distance between ๐ฅ and ๐ฆ by
๐(๐ฅ, ๐ฆ) = sup0โค๐ผโค1
{max [ |๐ฅ โ ๐ฆ| , |๐ฅ โ ๐ฆ| ]}
Let d: ๐ธ ร ๐ธ โถ ๐ ,๐คโ๐๐๐ ๐ ๐๐ ๐กโ๐ ๐ ๐๐๐๐ ๐๐ ๐๐ข๐ง๐ง๐ฆ ๐๐ข๐๐๐๐.
The following are satisfied:
1. ๐(๐ฅ + ๐ง, ๐ฆ + ๐ง) = ๐(๐ฅ, ๐ฆ), โ๐ฅ, ๐ฆ, ๐ง โ ๐ธ.
2. ๐(๐๐ฅ, ๐๐ฆ) = |๐|๐(๐ฅ, ๐ฆ), ๐ โ ๐ , ๐ฅ, ๐ฆ โ ๐ธ.
3. ๐(๐ฅ + ๐ฆ, ๐ง + ๐ค) โค ๐(๐ฅ, ๐ง) + ๐(๐ฆ,๐ค), โ๐ฅ, ๐ฆ, ๐ง, ๐ค โ ๐ธ.
the function ๐ (๐ฅ , ๐ฆ) is a metric on ๐ธ.
13
1.4 Fuzzy function
Definition (1.22) [42]: Fuzzy function: Let ๐ ๐๐๐ ๐ are universal sets and
๐: ๐ โถ ๐ be a function. Let moreover ๐น(๐), ๐น(๐) be respective universes
of fuzzy sets, identified with their membership functions, ๐. ๐ ๐น(๐) =
{๐ด: ๐ โ [0,1]} and similarly ๐น(๐). By extension principle, ๐ induces a
function ๐โ: ๐น(๐) โถ ๐น(๐) such that for all ๐ดโ ๐น(๐)
0 ๐๐ ๐โ1(๐ฆ) = โ ๐โ(๐ด)(๐ฆ) =
๐ ๐ข๐{๐ด(๐ฅ): ๐ฅ โ ๐โ1(๐ฆ) } ๐๐ ๐โ1(๐ฆ) โ โ
Definition (1.23) [20]: Continuous fuzzy function
A function ๐ โถ [๐ , ๐] โ ๐ธ is said to be continuous fuzzy function if for
arbitrary fixed ๐ฅ0 โ [๐ , ๐] , ๐ฟ > 0 such that:
|๐ฅ โ ๐ฅ0| < ๐ฟ โน ๐(๐(๐ฅ), ๐(๐ฅ0)) < ํ
Definition (1.24) [46]: Differential of a fuzzy function:
Let ๐ข: (๐, ๐) โถ ๐ธ be a fuzzy function and ๐ฅ0 โ (๐, ๐). ๐ข is differentiable
at ๐ฅ0 if two forms were sustained as follows:
It exists an element ๐ขโฒ(๐ฅ0) โ ๐ธ such that , for all โ > 0 sufficiently
near to 0, there are ๐ข(๐ฅ0 + โ) โ ๐ข(๐ฅ0), ๐ข(๐ฅ0) โ ๐ข(๐ฅ0 โ โ)and the
limits :
14
limโโ0+
๐ข(๐ฅ0 + โ) โ ๐ข(๐ฅ0)
โ= lim
โโ0+
๐ข(๐ฅ0) โ ๐ข(๐ฅ0 โ โ)
โ= ๐ขโฒ(๐ฅ0)
It exists an element ๐ขโฒ(๐ฅ0) โ ๐ธ such that , for all h < 0 sufficiently
near to 0, there are ๐ข(๐ฅ0 + โ) โ ๐ข(๐ฅ0), ๐ข(๐ฅ0) โ ๐ข(๐ฅ0 โ โ) and the
limits :
limโโ0โ
๐ข(๐ฅ0 + โ) โ ๐ข(๐ฅ0)
โ= lim
โโ0โ
๐ข(๐ฅ0) โ ๐ข(๐ฅ0 โ โ)
โ= ๐ขโฒ(๐ฅ0)
Let ๐ข: (๐, ๐) โ ๐ธ be a fuzzy function and denote:
[๐ข(๐ฅ)]๐ผ = [ ๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)], โ๐ผ โ [0, 1] ๐๐๐ ๐ฅ โ (0,1)
If f is differentiable, then ๐ข(๐ฅ, ๐ผ)๐๐๐ ๐ข(๐ฅ, ๐ผ)are differentiable functions
and
[๐ขโฒ(๐ฅ)]๐ผ = [ ๐ขโฒ(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)]
Definition (1.25) [13]: Strongly generalized differentiable
Let ๐ข: (๐, ๐) โ ๐ธ and ๐ฅ0 โ (๐, ๐) , We say that ๐ข is strongly generalized
differentiable at ๐ฅ0 if there exists an element ๐ขโฒ(๐ฅ0) โ ๐ธ such that:
(i) โ โ > 0 Sufficiently small, โ๐ข(๐ฅ0 + โ) โ ๐ข(๐ฅ0), ๐ข(๐ฅ0) โ ๐ข(๐ฅ0 โ โ)
and the limits (in the metric ๐)
๐๐๐โโ0
๐ข(๐ฅ0 + โ) โ ๐ข(๐ฅ0)
โ= ๐๐๐
โโ0
๐ข(๐ฅ0) โ ๐ข(๐ฅ0 โ โ)
โ= ๐ขโฒ(๐ฅ0),
๐๐
(ii) โ โ > 0 Sufficiently small, โ๐ข(๐ฅ0) โ ๐ข(๐ฅ0 + โ), ๐ข(๐ฅ0 โ โ) โ ๐ข(๐ฅ0)
and limits
15
๐๐๐โโ0
๐ข(๐ฅ0) โ ๐ข(๐ฅ0 + โ)
โ= ๐๐๐
โโ0
๐ข(๐ฅ0 โ โ) โ ๐ข(๐ฅ0)
โ= ๐ขโฒ(๐ฅ0),
๐๐
(iii) โ โ > 0 Sufficiently small, โ๐ข(๐ฅ0 + โ) โ ๐ข(๐ฅ0), ๐ข(๐ฅ0 โ โ) โ
๐ข(๐ฅ0) and the limits
๐๐๐โโ0
๐ข(๐ฅ0) โ ๐ข(๐ฅ0 + โ)
โโ= ๐๐๐
โโ0
๐ข(๐ฅ0) โ ๐ข(๐ฅ0 โ โ)
โโ= ๐ขโฒ(๐ฅ0),
๐๐
(iv) โ โ > 0 Sufficientlysmall,โ๐ข(๐ฅ0) โ ๐ข(๐ฅ0 + โ), ๐ข(๐ฅ0) โ ๐ข(๐ฅ0 โ โ)
and the limits
๐๐๐โโโ0
๐ข(๐ฅ0) โ ๐ข(๐ฅ0 + โ)
โโ= ๐๐๐
โโ0
๐ข(๐ฅ0) โ ๐ข(๐ฅ0 โ โ)
โ= ๐ขโฒ(๐ฅ0)
We will define the integral of a fuzzy function using the Riemann integral
concept.
Definition (1.26) [2]: Integral of a fuzzy function: Let ๐ข โถ [๐, ๐] โ ๐ธ.
for each partition ๐ = {๐ฅ0, ๐ฅ1, โฆ , ๐ฅ๐} of [a, b] and for arbitrary ๐ ๐ โถ
๐ฅ๐โ1 โค ๐ ๐ โค ๐ฅ๐ , 1 โค ๐ โค ๐, ๐๐๐ก ๐ = max1โค๐โค๐
{|๐ฅ๐ โ ๐ฅ๐โ1|} ๐๐๐
๐ ๐ = โ๐ข(
๐
๐=1
๐ ๐)(๐ฅ๐ โ ๐ฅ๐โ1)
the definite integral of ๐ข(๐ฅ) over [๐, ๐] is:
โซ๐ข(๐ฅ)๐๐ฅ =
๐
๐
lim๐โ0
๐ ๐
16
provided that this limit exists in the metric ๐.
If the fuzzy function ๐ข(๐ฅ) is continuous in the metric ๐, its definite integral
exists, furthermore.
(โซ๐ข(๐ฅ, ๐ผ)๐๐ฅ
๐
๐
) = โซ๐ข(๐ฅ, ๐ผ)๐๐ฅ
๐
๐
(โซ๐ข(๐ฅ, ๐ผ)๐๐ฅ
๐
๐
) = โซ๐ข(๐ฅ, ๐ผ)๐๐ฅ
๐
๐
where (๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)) is the parametric form of ๐ข(๐ฅ). It should be noted
that the fuzzy integral can be also defined using the Lebesgue- type
approach. However, if ๐ข (๐ฅ) is continuous, both approaches yield the same
value. Moreover, the representation of the fuzzy integral is more
convenient for numerical calculations.
Properties of the fuzzy integral [๐๐]
Let ๐น, ๐บ: ๐ โถ ๐ธ be integrable and ๐ โ ๐ , then the following are satisfied:
(1) โซ(๐น(๐ฅ) + ๐บ(๐ฅ)
๐
)๐๐ฅ = โซ๐น(๐ฅ)๐๐ฅ + โซ๐บ(๐ฅ)๐๐ฅ
๐๐
(2) โซ๐๐น(๐ฅ)๐๐ฅ = ๐ โซ๐น(๐ฅ)๐๐ฅ
๐๐
(3) ๐(๐น, ๐บ) integrable
(4) ๐ (โซ๐น(๐ฅ)๐๐ฅ
๐
, โซ๐บ
๐
(๐ฅ)๐๐ฅ) โค โซ๐(๐น, ๐บ)๐๐ก
๐
17
1.5 Fuzzy Linear Systems
Systems of simulations linear equations play major role in various areas
such as mathematics, statistics, and social sciences. Since in many
applications, at least some of the system's parameters and measurements
are represented by fuzzy rather than crisp numbers, therefore, it's important
to develop mathematical models and numerical procedures that would
appropriately treat general fuzzy linear systems and solve them.
The system of linear equations ๐๐ = ๐ถ where the coefficient matrix ๐ is
crisp, while ๐ถ is a fuzzy number vector, is called a fuzzy system of linear
equations (FSLE).
Definition (1.27) [1]: Fuzzy linear system: The ๐ ร ๐ linear system of
equations
๐ก11 ๐ฅ1 + ๐ก12 ๐ฅ2 + โฏ+ ๐ก1๐๐ฅ๐ = ๐1 ๐ก21 ๐ฅ1 + ๐ก22 ๐ฅ2 + โฏ+ ๐ก2๐ ๐ฅ๐ = ๐2 โฎ (1.1) ๐ก๐1 ๐ฅ1 + ๐ก๐2 ๐ฅ2 + โฏ+ ๐ก๐๐ ๐ฅ๐ = ๐๐
or, in matrix form:
๐๐ = ๐ถ (1.2)
where the coefficient matrix ๐ = ๐ก๐๐ , 1 โค ๐, ๐ โค ๐ is a crisp ๐ ร ๐ matrix
and ๐๐ โ ๐ธ, 1 โค ๐ โค ๐ .This system is called a fuzzy linear system (FLS).
18
Definition (1.28) [3]: Solution of fuzzy linear system:
A fuzzy vector ๐ = (๐ฅ1, ๐ฅ2, โฆ , ๐ฅ๐)๐ก, given by ๐ฅ๐ (๐ผ) = [๐ฅ๐(๐) , ๐ฅ๐(๐)], is
called the solution of (1.1) if:
โ๐๐๐ ๐๐
๐
๐=๐
= โ๐๐๐ ๐๐
๐
๐=๐
= ๐๐ , โ๐๐๐ ๐๐
๐
๐=๐
= โ๐๐๐ ๐๐
๐
๐=๐
= ๐๐,
we introduce the notations below [8]:
๐ฅ = (๐ฅ1, ๐ฅ2, โฆ , ๐ฅ๐, โ๐ฅ1, โ๐ฅ2, โฆ ,โ๐ฅ๐)๐ก= (๐ฅ,โ๐ฅ)
๐ก
(1.3)
๐ = (๐1, ๐2, โฆ , ๐๐, โ๐1, โ๐2, โฆ ,โ๐๐)๐ก= (๐,โ๐)
๐ก
๐ = (๐ง๐๐), 1 โค ๐, ๐ โค 2๐, ๐คโ๐๐๐ ๐ง๐๐ are determined as follows:
๐ก๐๐ โฅ 0 โน ๐ง๐๐ = ๐ก๐๐ , ๐ง๐+๐,๐+๐ = ๐ก๐๐ ,
๐ก๐๐ < 0 โน ๐ง๐,๐+ ๐ = โ๐ก๐๐ , ๐ง๐+๐,๐ = โ๐ก๐๐
and any ๐ง๐๐ which is not determined by (1.3) is zero. Using matrix notation
we have ๐๐ = ๐ถ (1.4)
the structure of Z implies that ๐ง๐๐ โฅ 0 and that
๐ = [๐ ๐๐ ๐
] (1.5)
where M contains the positive elements of ๐,๐ contains the absolute value
of the negative elements of ๐ and ๐ = ๐ โ ๐.
If matrix ๐ is nonsingular, then matrix ๐ may be singular.
19
Theorem (1.2) [40]: If ๐โ1exists it must have the same structure as ๐, ๐. ๐.
๐โ1 = (๐ท ๐ธ๐ธ ๐ท
) (1.6)
assuming that ๐ is nonsingular we obtain:
๐ = ๐โ1 ๐ถ (1.7)
Proof: see [40]
Theorem (1.3) [43]: The unique solution ๐ of equation (1.7) is a fuzzy
vector for arbitrary b if and only if ๐โ1 is nonnegative, ๐. ๐.
(๐โ1)๐๐ โฅ 0, 1 โค ๐, ๐ โค 2๐ (1.8)
Proof: see [43]
Definition (1.29) [2]: Strong fuzzy solution
Let ๐ = {(๐ฅ๐(๐ผ), ๐ฅ๐(๐ผ)), 1 โค ๐ โค ๐} denotes the unique solution
of (1.1). the fuzzy number vector ๐ = {(๐๐(๐ผ), ๐๐(๐ผ)) , 1 โค ๐ โค ๐}
defined by:
๐๐(๐ผ) = ๐๐๐{๐ฅ๐(๐ผ), ๐ฅ๐(๐ผ), ๐ฅ๐(1)}
(1.9)
๐๐(๐ผ) = max{๐ฅ๐(๐ผ), ๐ฅ๐(๐ผ), ๐ฅ๐(1)}
is called the fuzzy solution of ๐๐ = ๐ถ. Moreover if
20
(๐ฅ๐(๐ผ), ๐ฅ๐(๐ผ)), 1 โค ๐ โค ๐, are all of fuzzy numbers then we have ๐ฅ๐(๐ผ) =
๐๐(๐), ๐ฅ๐(๐ผ) = ๐๐(๐ผ), 1 โค ๐ โค ๐, and ๐ is called a strong fuzzy solution.
Otherwise, ๐ is a week fuzzy solution.
necessary and sufficient conditions for the existence of a strong solution
can be described by the following theorem:
Theorem (1.4) [8]:
Let ๐ = [๐ ๐๐ ๐
] be a nonsingular matrix. The system (1.4) has a strong
solution if and only if:
(๐ + ๐)โ1(๐ โ ๐) โค 0 (1.10)
Proof: see [8]
Theorem (1.5) [8]: The (๐น๐ฟ๐) (1.1) has a unique strong solution if and
only if the following conditions hold:
1) The matrices ๐ = ๐ โ ๐ ๐๐๐ ๐ + ๐ ๐re both nonsingulars.
2) (๐ + ๐)โ1(๐ โ ๐) โค 0 .
1.6 Fuzzy integral equations
Definition (1.30): Integral equation: An integral equation is the equation
in which the unknown function appears under an integral sign.
In this section, the fuzzy integral equation of second kind is introduced.
the Fredholm integral equation of the second kind is given by [32]:
๐ข(๐ฅ) = ๐(๐ฅ) + ๐ โซ ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐ก๐
๐ (1.11)
where ๐ > 0, ๐(๐ฅ, ๐ก)is an arbitrary kernal function over the square
21
๐ โค ๐ก, ๐ฅ โค ๐ and ๐(๐ฅ)is a function of ๐ฅ: ๐ โค ๐ฅ โค ๐.
If the kernel function satisfies ๐( ๐ฅ, ๐ก ) = 0, t > ๐ฅ, we obtain the Volterra
integral equation:
๐ข(๐ฅ) = ๐(๐ฅ) + ๐ โซ ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐ก๐ฅ
๐
(1.12)
if ๐ (๐ฅ) is a fuzzy function, we have Volterra fuzzy integral of the second
kind.
Definition (1.30) [27]: General form of Volterra integral equation of
the second kind: The second kind fuzzy Volterra integral equation
system is in the form
๐ข๐(๐ฅ) = ๐๐(๐ฅ) + โ ๐๐๐ โซ ๐1๐(๐ฅ, ๐ก)๐ข๐(๐ก)๐๐ก ๐ฅ
๐๐๐=1 (1.13)
๐ โค ๐ก โค ๐ฅ โค ๐ ๐๐๐ ๐๐๐ โ 0 (๐๐๐ ๐, ๐ = 1,2,3,โฆ ,๐) are real constants.
Moreover, in system (1.13), the fuzzy function ๐๐(๐ฅ) and kernel ๐_๐๐ (๐ฅ, ๐ก)
are given and assumed to be sufficiently differentiable with respect to all
their arguments on the interval ๐ โค ๐ก, ๐ฅ โค ๐, and we assume that the kernel
function ๐๐๐(๐ฅ, ๐ก) โ ๐ฟ2([๐, ๐] ร [๐, ๐]), and
๐ข(๐ฅ) = [๐ข1, ๐ข2, โฆ , ๐ข๐]๐กis the solution to be determined.
now,(๐๐(๐ฅ, ๐ผ), ๐๐(๐ฅ, ๐ผ))and (๐ข๐(๐ฅ, ๐ผ), ๐ข๐(๐ฅ, ๐ผ)), (0 โค ๐ผ โค 1, ๐ โค ๐ฅ โค ๐)
be parametric form of ๐๐(๐ฅ) and ๐ข๐(๐ฅ), respectively. To simplify, we
assume that ๐๐๐ > 0(๐๐๐ ๐, ๐ = 1,2,โฆ ,๐).
In order to design a numerical scheme for solving (1.16), we write the
parametric form of the given fuzzy Volterra integral equations system as
follows:
22
๐ข๐(๐ฅ, ๐ผ) = ๐๐(๐ฅ, ๐ผ) + โ( ๐๐๐ โซ ๐
๐๐(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
)
๐
๐=1
๐ = 1,2,โฆ ,๐ (1.14)
๐ข๐(๐ฅ, ๐ผ) = ๐๐(๐ฅ, ๐ผ) + โ( ๐๐๐ โซ ๐๐๐(๐ก, ๐ผ)๐๐ก ๐ฅ
๐
)
๐
๐=1
where
๐๐๐(๐ก, ๐ผ) =
๐๐๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ), ๐๐๐(๐ฅ, ๐ก) โฅ 0
๐๐๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ), ๐๐๐(๐ฅ, ๐ก) < 0
(1.15)
and
๐๐๐(๐ก, ๐ผ) =
๐๐๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ), ๐๐๐(๐ฅ, ๐ก) โฅ 0
๐๐๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ), ๐๐๐(๐ฅ, ๐ก) < 0
(1.16)
Fuzzy integro-differential equations
The linear fuzzy integro-differential equation is given by [38]:
๐ขโฒ(๐ฅ) = ๐(๐ฅ) + ๐ โซ ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐กโ(๐ฅ)
๐, ๐ข(๐ฅ0) = ๐ข0
(1.17)
where ๐ > 0, the kernel function ๐(๐ฅ, ๐ก) โ ๐ฟ2 [๐, ๐], ๐ โค ๐ก, ๐ฅ โค ๐, and
๐(๐ฅ)is a given function of ๐ฅ โ [๐, ๐].If ๐ข is a fuzzy function, ๐(๐ฅ)is a
given fuzzy function of ๐ฅ โ [๐, ๐] and ๐ขโฒ is the first fuzzy derivative of ๐ข.
This equation may be possessing fuzzy solution.
If the kernel function satisfies ๐( ๐ฅ, ๐ก ) = 0, โ(๐ฅ)is a variable, we obtain
the Voltera integro-differential equation:
23
๐ขโฒ(๐ฅ) = ๐(๐ฅ) + ๐ โซ ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐ก๐ฅ
๐ (1.18)
If โ(๐ฅ) = ๐ is a constant then the equation (1.20) is fuzzy Fredholm
integro-differential equation of the second kind is given by:
๐ขโฒ(๐ฅ) = ๐(๐ฅ) + ๐ โซ ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐ก๐
๐ (1.19)
Let ๐ข(๐ฅ) = (๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)) is a fuzzy solution of equation (1.17). using
definitions (1.17), ๐๐๐ (1.18) we have the equivalent system:
{
๐ขโฒ(๐ฅ) = ๐(๐ฅ) + ๐โซ๐
๐ฅ
๐
(๐ก, ๐ผ)๐๐ก, ๐ข(๐ฅ0) = ๐ข 0
๐ขโฒ(๐ฅ) = ๐(๐ฅ) + ๐โซ๐
๐ฅ
๐
(๐ก, ๐ผ)๐๐ก, ๐ข(๐ฅ0) = ๐ข0
(1.20)
for each 0โค ๐ผ โค 1๐๐๐ ๐ , ๐ฅ โ [๐, ๐], where
(๐ ๐ก, ๐ผ) =
๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ), ๐(๐ฅ, ๐ก) โฅ 0
๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ), ๐(๐ฅ, ๐ก) < 0
and (1.21)
๐(๐ก, ๐ผ) =
๐(๐ , ๐ก)๐ข(๐ก, ๐ผ), ๐(๐ฅ, ๐ก) โฅ 0
๐(๐ , ๐ก)๐ข(๐ก, ๐ผ), ๐(๐ฅ, ๐ก) < 0
this possesses a unique solution. (๐ข, ๐ข) โ ๐ต
which is a fuzzy function, ๐. ๐. ๐๐๐ ๐๐๐โ ๐ฅ, (๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)) is a fuzzy
number, therefore each solution of equation(1.17) is a solution of system
(1.20) and conversely.
24
Chapter Two
Analytical Methods For Solving Fuzzy Volterra
Integral Equation of the second Kind
In this chapter we will investigate analytical solutions for the linear
Volterra fuzzy integral equation of the second kind , namely; fuzzy Laplace
transform method, homotopy analysis method, the Adomian decomposition
method and differential transformation method.
2.1 Fuzzy Laplace Transformation Method
In this section, we will introduce some basic definitions for the fuzzy
Laplace transform and fuzzy convolution, then solve fuzzy convolution
Volterra integral equation of the second kind by using fuzzy Laplace
transform method.
Theorem (2.1) [๐๐]: Suppose that ๐ข(๐ฅ) is a fuzzy valued function
on[ ๐,โ) represented by the parametric form (๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)), for any
number ๐ผ โ [0,1]. Assume that ๐ข(๐ฅ, ๐ผ) ๐๐๐ ๐ข(๐ฅ, ๐ผ) are Riemann-
Integrable on [๐, ๐], ๐๐๐ ๐๐ฃ๐๐๐ฆ ๐ โฅ ๐. Also we assume that
๐(๐ผ) ๐๐๐ ๐(๐ผ) are positive functions, such that:
โซ ๐ข(๐ฅ, ๐ผ)
๐
๐
๐๐ฅ โค ๐(๐ผ) ๐๐๐ โซ|๐ข(๐ฅ, ๐ผ)|
๐
๐
๐๐ฅ โค ๐(๐ผ), ๐๐๐ ๐๐ฃ๐๐๐ฆ ๐ โฅ ๐,
25
Then ๐ข(๐ฅ) is improper fuzzy Riemann-Integrable on [ ๐,โ) which is a
fuzzy number, we have:
โซ๐ข(๐ฅ, ๐ผ)๐๐ฅ = (โซ๐ข(๐ฅ, ๐ผ)๐๐ฅ,โซ ๐ข(๐ฅ, ๐ผ)๐๐ฅ
๐
๐
๐
๐
)
๐
๐
Proposition (2.1) [๐๐]: Let ๐ข(๐ฅ) ๐๐๐ ๐ฃ(๐ฅ) are two fuzzy valued functions
and fuzzy Riemann-integrable on [ ๐,โ ), then ๐ข(๐ฅ) + ๐ฃ(๐ฅ) is fuzzy
Riemann-integrable on[ ๐,โ ).
Moreover, we have:
โซ(๐ข(๐ฅ) + ๐ฃ(๐ฅ))
โ
๐
๐๐ฅ = โซ ๐ข(๐ฅ)
โ
๐
๐๐ฅ + โซ ๐ฃ(๐ฅ)
โ
๐
๐๐ฅ
Definition )2.1 ( [ 19 ]: Fuzzy Laplace transform:
Let ๐ข(๐ฅ) be a fuzzy valued function and ๐ is a real parameter, then the
fuzzy Laplace transform of the function ๐ข with respect to ๐ก denoted by
๐(๐ ) is defined by:
๐(๐ ) = ๐ฟ(๐ข(๐ฅ)) = โซ ๐โ๐ ๐ฅ๐ข(๐ฅ)๐๐ฅโ
0
= lim๐โโ
โซ๐โ๐ ๐ฅ๐ข(๐ฅ)๐๐ฅ,
๐
0
(2.1)
using theorem (2.1), we obtain:
๐(๐ ) = lim๐โโ
โซ๐โ๐ ๐ฅ๐ข(๐ฅ)๐๐ฅ
๐
0
, lim๐โโ
โซ๐โ๐ ๐ฅ ๐ข(๐ฅ)๐๐ฅ
๐
0
where the limits exist.
26
the ๐ผ โ ๐๐ข๐ก representation of ๐(๐ ) is given by:
๐(๐ , ๐ผ) = ๐ฟ(๐ข(๐ฅ, ๐ผ)) = [๐ฟ (๐ข(๐ฅ, ๐ผ)) , ๐ฟ(๐ข(๐ฅ, ๐ผ))],
where
๐ฟ (๐ข(๐ฅ, ๐ผ)) = โซ ๐โ๐ ๐ฅ๐ข(๐ฅ, ๐ผ)๐๐ฅโ
0
= lim๐โโ
โซ๐โ๐ ๐ฅ๐ข(๐ฅ, ๐ผ)๐๐ฅ
๐
0
, 0 โค ๐ผ โค 1
๐ฟ(๐ข(๐ฅ, ๐ผ)) = โซ ๐โ๐ ๐ฅ๐ข(๐ฅ, ๐ผ)๐๐ฅโ
0
= lim๐โโ
โซ๐โ๐ ๐ฅ๐ข(๐ฅ, ๐ผ)๐๐ฅ
๐
0
, 0 โค ๐ผ โค 1
Theorem(2.2)[๐๐]: Suppose that ๐ข(๐ฅ) ๐๐๐ ๐ฃ(๐ฅ) are continuous fuzzy
valued functions, and suppose ๐1, ๐2 are constants, then:
๐ฟ{๐1 ๐ข(๐ฅ) + ๐2๐ฃ(๐ฅ)} = ๐1 ๐ฟ{๐ข(๐ฅ)} + ๐2 ๐ฟ{๐ฃ(๐ฅ)} = ๐1๐(๐ ) + ๐2๐(๐ )
Lemma(2.1) [๐๐]: Suppose that ๐ข(๐ฅ) be a fuzzy valued function
on [ 0,โ) and ๐ โ ๐ , then
๐ฟ{๐ ๐ข(๐ฅ)} = ๐ ๐ฟ{๐ข(๐ฅ)} = ๐๐(๐ )
Lemma(2.2) [๐๐]: Suppose that ๐ข(๐ฅ) is a continuous fuzzy valued
function and ๐ฃ(๐ฅ) โฅ 0 is a real valued function such that (๐ข(๐ฅ), ๐ฃ(๐ฅ))๐โ๐ ๐ฅ
is improper fuzzy Riemann-integrable on [ 0,โ ) then for fixed ๐ผ โ [0,1],
we have
โซ(๐ข(๐ฅ, ๐ผ)๐ฃ(๐ฅ, ๐ผ))
โ
0
๐โ๐ ๐ฅ๐๐ฅ
= โซ ๐ข(๐ฅ, ๐ผ)๐ฃ(๐ฅ, ๐ผ)๐โ๐ ๐ฅ๐๐ฅ
โ
0
, โซ ๐ข(๐ฅ, ๐ผ)๐ฃ(๐ฅ, ๐ผ)๐โ๐ ๐ฅ๐๐ฅ
โ
0
27
Theorem(2.3)[๐๐]: (First shifting theorem): Let ๐ข(๐ฅ) be a continuous
fuzzy valued function and ๐ฟ{๐ข(๐ฅ, ๐ผ)} = ๐(๐ ); ๐กโ๐๐:
๐ฟ{๐๐๐ฅ๐ข(๐ฅ)} = ๐(๐ โ ๐), ๐ โ ๐ > 0.
Proof:
๐ฟ{๐๐๐ฅ๐ข(๐ฅ)} = โซ ๐โ๐ ๐ฅ๐๐๐ฅ๐ข(๐ฅ)๐๐ฅ
โ
0
= โซ ๐โ๐ ๐ฅ+๐๐ฅ๐ข(๐ฅ)๐๐ฅ =
โ
0
โซ ๐โ(๐ โ๐)๐ฅ๐ข(๐ฅ)๐๐ฅ = ๐(๐ โ ๐)
โ
0
hence,
๐ฟ{๐๐๐ฅ๐ข(๐ฅ)} = ๐(๐ โ ๐).
Theorem(2.4)[๐๐]: Let ๐ข(๐ฅ) be a continuous fuzzy valued function and
๐ฟ{๐ข(๐ฅ)} = ๐(๐ ). Suppose that ๐ is a constant, then:
๐ฟ{๐ข(๐๐ฅ)} =1
๐๐ (
๐
๐)
Proof:
๐ฟ{๐ข(๐๐ฅ)} = โซ ๐โ๐ ๐ฅ๐ข(๐๐ฅ)๐๐ฅ
โ
0
๐๐ข๐ก ๐ฆ = ๐๐ฅ โน๐๐ฆ
๐๐ฅ= ๐ โน ๐๐ฅ =
๐๐ฆ
๐
now,
๐ฟ{๐ข(๐๐ฅ)} = โซ ๐โ๐ ๐ฆ๐๐ข(๐ฆ)
๐๐ฆ
๐
โ
0
=1
๐โซ ๐โ๐
๐ฆ๐๐ข(๐ฆ)๐๐ฆ
โ
0
=1
๐๐ (
๐
๐)
28
hence,
๐ฟ{๐ข(๐๐ฅ)} =1
๐๐ (
๐
๐).
2.2 Fuzzy Convolution
In this section we will introduce the concept of fuzzy convolution, and
solve fuzzy convolution Volterra integral equation of the second kind.
Definition (๐. ๐) [ 42 ]: (๐ฑ๐๐๐ ๐ ๐๐๐๐๐๐๐๐๐๐๐๐): Suppose that ๐ข(๐ฅ) is a
fuzzy-valued function, ๐ข is said to have a jump discontinuity at ๐ฅ0 if the
following conditions are satisfied:
lim๐ฅโ๐ฅ0โ
๐ข(๐ฅ)๐๐๐ lim๐ฅโ๐ฅ0+
๐ข(๐ฅ) ๐๐ฅ๐๐ ๐ก
๐ข(๐ฅ0โ) โ ๐ข(๐ฅ0
+)
Definition (๐. ๐)[42]: (๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐๐): Let ๐ข(๐ฅ) be a fuzzy-
valued function,then ๐ข is a piecewise continuous on [ 0,โ ) if :
lim๐ฅโ0+
๐ข(๐ฅ) = ๐ข(0+)
๐ข is continuous on every finite interval (0,๐) except at finite
number of points ๐1, ๐2, โฆ , ๐๐ in the interval (0,๐) at which ๐ข has
a jump discontinuity.
Definition (๐. ๐) [42 ]: (๐๐๐๐๐๐๐๐๐๐๐ ๐๐๐ ๐๐): Suppose that ๐ข(๐ฅ) is a
fuzzy-valued function on [ 0,โ ), then ๐ข has an exponential order ๐ if
there exists constants ๐ > 0 ๐๐๐ ๐, such that |๐ข(๐ฅ)| โค ๐๐๐๐ฅ, ๐ฅ โฅ ๐ฅ0, for
some ๐ฅ0 โฅ 0.
29
Definition (๐. ๐)[ 45 ]: Fuzzy convolution: Let ๐(๐ฅ) ๐๐๐ ๐(๐ฅ) are two
piece-wise continuous fuzzy-valued functions on [ 0,โ ), then the
convolution of ๐ ๐๐๐ ๐ denoted by ๐ โ ๐ is defined as:
(๐ โ ๐)(๐ฅ) = โซ๐(๐)๐(๐ฅ โ ๐)๐๐ . (2.2)
๐ฅ
0
Properties of fuzzy convolution:
( ๐ ) (๐ โ ๐) = (๐ โ ๐) (๐๐๐๐๐ข๐ก๐๐ก๐๐ฃ๐ ๐๐๐๐๐๐๐ก๐ฆ)
( ๐๐ ) ๐(๐ โ ๐) = (๐๐) โ ๐ = ๐ โ (๐๐), ๐ ๐๐ ๐๐๐๐ ๐ก๐๐๐ก
(๐๐๐)๐ โ (๐ โ โ) = (๐ โ ๐) โ โ (๐๐ ๐ ๐๐๐๐๐ก๐๐ฃ๐ ๐๐๐๐๐๐๐ก๐ฆ)
Proof:
(๐) ๐ ๐ข๐๐ ๐ก๐๐ก๐ข๐ก๐๐๐ ๐ = ๐ฅ โ ๐ ๐๐ ๐กโ๐ ๐๐๐ข๐๐ก๐๐๐ (2.2),
๐ค๐ ๐๐๐ก:
(๐ โ ๐)(๐ฅ) = โซ๐(๐)๐(๐ฅ โ ๐)๐๐ฅ = (๐ โ ๐)(๐ฅ)
๐ฅ
0
(๐๐) ๐ก๐๐๐ฃ๐๐๐
(๐๐๐) [๐ โ (๐ โ โ)](๐ฅ) = โซ๐(๐). (๐ โ โ)(๐ฅ โ ๐)๐๐
๐ฅ
0
30
= โซ๐(๐)(โซ ๐(๐)โ(๐ฅ โ ๐ โ ๐)๐๐
๐ฅโ๐
0
)๐๐
๐ฅ
0
= โซ(โซ๐(๐)
๐
0
๐(๐ โ ๐)๐๐)โ(๐ฅ โ ๐)
๐ฅ
0
๐๐
= [(๐ โ ๐) โ โ](๐ฅ)
Theorem(๐. ๐)[๐๐]: Convolution theorem:
Suppose that ๐(๐ฅ) ๐๐๐ ๐(๐ฅ) are piece-wise continuous fuzzy-valued
functions on [ 0,โ) of exponential order ๐ with fuzzy Laplace transforms
๐(๐ ) ๐๐๐ ๐(๐ ) respectively, then:
๐ฟ{(๐ โ ๐)(๐ฅ)} = ๐ฟ{๐(๐ฅ)}. ๐ฟ{๐(๐ฅ)} = ๐(๐ ). ๐(๐ ), ๐ > ๐
Proof: See [45]
Definition (๐. ๐)[45]: Fuzzy convolution Volterra integral equation of
the second kind: The fuzzy convolution Volterra integral equation of the
second kind is defined as:
๐ข(๐ฅ) = ๐(๐ฅ) + ๐โซ๐(๐ฅ โ ๐ก)๐ข(๐ก)๐๐ก
๐ฅ
0
, ๐ฅ โ [0, ๐], ๐ < โ (2.3)
where ๐(๐ฅ โ ๐ก)is an arbitrary given fuzzy-valued convolution kernel
function, and ๐(๐ฅ)is a continuous fuzzy-valued function.
31
now, by taking fuzzy Laplace transform on both sides of equation (2.3), we
get:
L{๐ข(๐ฅ)} = ๐ฟ{๐(๐ฅ)} + ๐ฟ ๐โซ๐(๐ฅ โ ๐ก)๐ข(๐ก)๐๐ก
๐ฅ
0
, ๐ฅ โ [0, ๐], ๐ < โ
then, by using fuzzy convolution and definition of fuzzy Laplace transform,
we get:
๐ฟ{ ๐ข(๐ฅ, ๐ผ)} = ๐ฟ { ๐(๐ฅ, ๐ผ)} + ๐๐ฟ{๐(๐ฅ, ๐ก)} ๐ฟ{๐ข(๐ฅ, ๐ผ)}
๐ฟ{ ๐ข(๐ฅ, ๐ผ)} = ๐ฟ{ ๐(๐ฅ, ๐ผ)} + ๐๐ฟ{๐(๐ฅ, ๐ก)} ๐ฟ{๐ข(๐ฅ, ๐ผ)} (2.4)
For solving the equation (2.4), we will discuss the following cases:
๐๐๐ ๐(1): ๐๐ ๐(๐ฅ, ๐ก) > 0 , ๐กโ๐๐ ๐ค๐ ๐๐๐ก:
๐ฟ{ ๐ข(๐ฅ, ๐ผ)} = ๐ฟ { ๐(๐ฅ, ๐ผ)} + ๐ ๐ฟ{ ๐(๐ฅ, ๐ก)}๐ฟ{ ๐ข(๐ฅ, ๐ผ)}
๐ฟ{๐ข(๐ฅ, ๐ผ)} = ๐ฟ{ ๐(๐ฅ, ๐ผ)} + ๐๐ฟ{ ๐(๐ฅ, ๐ก)}๐ฟ{ ๐ข(๐ฅ, ๐ผ)}
we obtain explicit formula:
๐ฟ{ ๐ข(๐ฅ, ๐ผ)} =๐ฟ { ๐(๐ฅ, ๐ผ)}
1 โ ๐ ๐ฟ{ ๐(๐ฅ, ๐ก)}
๐ฟ{ ๐ข(๐ฅ, ๐ผ)} =๐ฟ{ ๐(๐ฅ, ๐ผ)}
1 โ ๐๐ฟ{ ๐(๐ฅ, ๐ก)}
take the inverse of fuzzy Laplace transform, we get the solution:
๐ข(๐ฅ, ๐ผ) = ๐ฟโ1 ๐น(๐ )
1 โ ๐๐พ(๐ )
32
๐ข(๐ฅ, ๐ผ) = ๐ฟโ1 ๐น(๐ )
1 โ ๐๐พ(๐ )
where F(s) and K(s) are fuzzy Laplace transformations f(x)and k(x, t)
respectively.
๐๐๐ ๐(2): ๐๐ ๐(๐ฅ, ๐ก) < 0, ๐กโ๐๐ ๐ค๐ ๐๐๐ก:
๐ฟ{๐ข(๐ฅ, ๐ผ)} = ๐ฟ { ๐(๐ฅ, ๐ผ)} + ๐ ๐ฟ{๐(๐ฅ, ๐ก)}๐ฟ{ ๐ข(๐ฅ, ๐ผ)}
๐ฟ{๐ข(๐ฅ, ๐ผ)} = ๐ฟ{ ๐(๐ฅ, ๐ผ)} + ๐๐ฟ{๐(๐ฅ, ๐ก)}๐ฟ{ ๐ข(๐ฅ, ๐ผ)}
we obtain explicit formula:
๐ฟ{ ๐ข(๐ฅ, ๐ผ)} =๐ฟ { ๐(๐ฅ, ๐ผ)} โ ๐๐ฟ{ ๐(๐ฅ, ๐ก)}๐ฟ{ ๐(๐ฅ, ๐ผ)}
1 โ ๐2๐ฟ{ ๐(๐ฅ, ๐ก)} ๐ฟ{ ๐(๐ฅ, ๐ก)}
๐ฟ{ ๐ข(๐ฅ, ๐ผ)} =๐ฟ{ ๐(๐ฅ, ๐ผ)} โ ๐๐ฟ{๐(๐ฅ, ๐ก)}๐ฟ { ๐(๐ฅ, ๐ผ)}
1 โ ๐2๐ฟ{ ๐(๐ฅ, ๐ก)} ๐ฟ{ ๐(๐ฅ, ๐ก)}
take the inverse of fuzzy Laplace transform, we get the solution:
๐ข(๐ฅ, ๐ผ) = ๐ฟโ1 ๐น(๐ ) โ ๐๐พ(๐ )๐น(๐ )
1 โ [๐ ๐พ(๐ )]2
๐ข(๐ฅ, ๐ผ) = ๐ฟโ1 ๐น(๐ ) โ ๐๐พ(๐ )๐น(๐ )
1 โ [๐ ๐พ(๐ )]2
where F(s) and K(s) are fuzzy Laplace transformations f(x)and k(x, t)
respectively.
33
Example (2.1): consider the following fuzzy Volterra integral equation
๐ข(๐ฅ) = (๐ผ + 1,3 โ ๐ผ) + โซ (๐ฅ โ ๐ก)
๐ฅ
0
๐ข(๐ก)๐๐ก, ๐ฅ โ [0, ๐], ๐ < โ (2.5)
by taking fuzzy Laplace transform on both sides of equation(2.5), we get:
๐ฟ{๐ข(๐ฅ)} = ๐ฟ{(๐ผ + 1,3 โ ๐ผ)} + ๐ฟ{๐ฅ}. ๐ฟ{๐ข(๐ก)}
๐. ๐
๐ฟ{๐ข(๐ฅ, ๐ผ)} = ๐ฟ{(๐ผ + 1) } + ๐ฟ{๐ฅ}. ๐ฟ{๐ข(๐ฅ, ๐ผ)}, 0โค ๐ผ โค 1
๐ฟ{๐ข(๐ฅ, ๐ผ)} = ๐ฟ{(3 โ ๐ผ)} + ๐ฟ{๐ฅ}. ๐ฟ{๐ข(๐ฅ, ๐ผ)}, 0 โค ๐ผ โค 1
hence, we get:
๐ฟ{๐ข(๐ฅ, ๐ผ)} = (๐ผ + 1)1
๐ +
1
๐ 2๐ฟ{๐ข(๐ฅ, ๐ผ)}, 0 โค ๐ผ โค 1
๐ฟ{๐ข(๐ฅ, ๐ผ)} = (3 โ ๐ผ)1
๐ +
1
๐ 2๐ฟ{๐ข(๐ฅ, ๐ผ)}, 0 โค ๐ผ โค 1
Finally, by taking the inverse of fuzzy Laplace transform on both sides of
above equations, we get:
{๐ข(๐ฅ, ๐ผ) = (๐ผ + 1) cosh(๐ฅ)
๐ข(๐ฅ, ๐ผ) = (3 โ ๐ผ) cosh(๐ฅ) 0 โค ๐ผ โค 1
2.3 Fuzzy Homotopy Analysis Method
In this section, we find the approximate analytical solutions of fuzzy
Volterra integral equations of the second kind by using fuzzy homotopy
analysis method. The main advantage of this method is that it can be used
34
directly without using assumptions or transformations. The (FHAM)
solutions depend on an auxiliary parameter which provides a suitable way
of controlling the convergence region of series solutions. we give an
example reveal the efficiency of this method.
Consider the following differential equation:
โณ(๐ข(๐ฅ, ๐ผ)) = 0, (2.6)
where โณ is a nonlinear operator, ๐ฅ denotes independent variable and
๐ข(๐ฅ, ๐ผ)is the unknown function.
Definition (๐. ๐) [31]: Homotopy operator: We define the homotopy
operator โ as:
โ(ฮฆ, ๐) โก (1 โ ๐)โ(ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐ข0(๐ฅ, ๐ผ)) โ ๐ ๐ ๐(๐ฅ, ๐ผ) โณ(ฮฆ(๐ฅ, ๐, ๐ผ))
(2.7)
where ๐ โ [0, 1] is the embedding parameter, ๐ โ 0 is a non- zero auxiliary
parameter (convergence control parameter), ๐(๐ฅ, ๐ผ) โ 0 is an auxiliary
function, ๐ข0(๐ฅ, ๐ผ) denotes the initial approximation of ๐ข(๐ฅ, ๐ผ) in the
equation (2.6)ฮฆ(๐ฅ, ๐, ๐ผ) is an unknown function, and โ is an auxiliary
linear operator with property โ(0) = 0.
In equation (2.7) if we put โ(ฮฆ, ๐, ๐ผ) = 0, we will construct the so-
called ๐ง๐๐๐ order deformation equation
(1 โ ๐)โ(ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐ข0(๐ฅ, ๐ผ))
= ๐ ๐ ๐(๐ฅ, ๐ผ) โณ(ฮฆ(๐ฅ, ๐, ๐ผ)) (2.8)
35
obviously, for ๐ = 0 we get
โ(ฮฆ(๐ฅ, 0, ๐ผ) โ ๐ข0(๐ฅ, ๐ผ)) = 0 โน ฮฆ(๐ฅ, 0, ๐ผ) = ๐ข0(๐ฅ, ๐ผ)
whereas, for ๐ = 1,we get:
โณ(ฮฆ(๐ฅ, 1, ๐ผ)) = 0 โน ฮฆ(๐ฅ, 1, ๐ผ) = ๐ข(๐ฅ, ๐ผ), where ๐ข(๐ฅ) is
the solution of the equation (2.6)
thus, as ๐ increases from 0 ๐ก๐ 1, the solution ฮฆ(๐ฅ, ๐, ๐ผ) changes
continuously from the initial approximation ๐ข0(๐ฅ, ๐ผ) to the exact solution
๐ข(๐ฅ, ๐ผ).
by expanding ฮฆ(๐ฅ, ๐, ๐ผ) into the Maclaurin series with respect to ๐,
we have:
ฮฆ(๐ฅ, ๐, ๐ผ) = ๐ข0(๐ฅ, ๐ผ) + โ ๐ข๐(๐ฅ, ๐ผ)๐๐
โ
๐=1
(2.9)
where
๐ข๐(๐ฅ, ๐ผ) = ๐ท๐(ฮฆ) = 1
๐! ๐๐โณ(ฮฆ(๐ฅ, ๐, ๐ผ))
๐๐๐ ๐=0
, ๐ = 1,2, โฆ (2.10)
if the series (2.9) convergent for ๐ = 1, then we have
๐ข(๐ฅ, ๐ผ) = ๐ข0(๐ฅ, ๐ผ) + โ ๐ข๐(๐ฅ, ๐ผ) (2.11)
โ
๐=1
In [๐๐] by using equation (2.9), we can write equation (2.8) as:
36
(1 โ ๐)โ(ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐ข0(๐ฅ, ๐ผ)) = (1 โ ๐)โ โ ๐ข๐(๐ฅ, ๐ผ)๐๐
โ
๐=1
= ๐ ๐ ๐(๐ฅ)โณ(ฮฆ(๐ฅ, ๐, ๐ผ)) (2.12)
this implies
โ โ ๐ข๐(๐ฅ, ๐ผ)๐๐
โ
๐=1
โ ๐โ โ ๐ข๐(๐ฅ, ๐ผ)๐๐
โ
๐=1
= ๐ ๐ ๐(๐ฅ)โณ(ฮฆ(๐ฅ, ๐, ๐ผ))
(2.13)
by differentiating the both sides of equation (2.13)๐ ๐ก๐๐๐๐ with respect to
๐, we get:
๐! โ[๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐โ1(๐ฅ, ๐ผ)] = ๐ ๐(๐ฅ)๐ ๐๐โ1โณ(ฮฆ(๐ฅ, ๐, ๐ผ))
๐๐๐โ1 ๐=0
hence, the ๐th-order deformation equation (๐ > 0) is:
โ[๐ข๐(๐ฅ, ๐ผ) โ ๐๐๐ข๐โ1(๐ฅ, ๐ผ)] = ๐ ๐(๐ฅ) ฮค๐(๏ฟฝโ๏ฟฝ ๐โ1(๐ฅ, ๐ผ)) (2.14)
๐คโ๐๐๐ ๏ฟฝโ๏ฟฝ ๐โ1 = {๐ข0(๐ฅ, ๐ผ), ๐ข1(๐ฅ, ๐ผ),โฆ , ๐ข๐โ1(๐ฅ, ๐ผ)},
๐๐ = {0, ๐ โค 11, ๐ > 1
(2.15)
๐๐๐
ฮค๐(๏ฟฝโ๏ฟฝ ๐โ1(๐ฅ, ๐ผ)) = 1
(๐ โ 1)!
๐๐โ1
๐๐๐โ1โณ(ฮฆ(๐ฅ, ๐, ๐ผ))
๐=0
(2.16)
To ensure the convergence of the series, we must concentrate that ๐ข0(๐ฅ, ๐ผ)
is the initial approximation, ๐ โ 0 is a non-zero auxiliary parameter, โ is an
37
auxiliary linear operator, ๐ is the embedding parameter, and ๐(๐ฅ, ๐ผ) is an
auxiliary function.
In this part, we will rewrite the fuzzy Volterra integral equations of the
second kind, and then solve them by using homotopy analysis method.
we know that the parametric form of fuzzy Volterra integral equation of
second kind is:
{
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐ โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
(2.17)
now, assume that ๐(๐ฅ, ๐ก) โฅ 0 , ๐ โค ๐ โค ๐, ๐๐๐ ๐(๐ฅ, ๐ก) < 0, ๐ โค ๐ โค ๐ฅ
with above assumptions, the equation (2.17) becomes:
{
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐ โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐)๐๐ก
๐
๐
+ ฮป โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐
๐
+ ๐ โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
(2.18)
we note that (2.18) is a linear Fredholm-Volterra integral equations in
crisp case for each 0 โค ๐ผ โค 1.
To solve the system (2.18) by means of homotopy analysis method, we
choose the linear operator [๐๐]:
38
โ[ฮฆ(๐ฅ, ๐, ๐ผ)] = ฮฆ(๐ฅ, ๐, ๐ผ) (2.19)
we now define a nonlinear operator โณ(ฮฆ(๐ฅ, ๐, ๐)) as:
โณ (ฮฆ(๐ฅ, ๐, ๐ผ)) = ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐(๐ฅ, ๐ผ) โ ๐โซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, ๐, ๐ผ)๐๐ก
๐
๐
โ๐ โซ ๐(๐ฅ, ๐ก)ฮฆ(๐ก, ๐, ๐ผ)๐๐ก๐ฅ
๐,
and
โณ (ฮฆ(๐ฅ, ๐, ๐ผ)) = ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐(๐ฅ, ๐ผ) โ ๐โซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, ๐, ๐ผ)๐๐ก
๐
๐
โฮป โซ ๐(๐ฅ, ๐ก)ฮฆ(๐ก, ๐, ๐ผ)๐๐ก๐ฅ
๐. (2.20)
using the two equations (2.8) and (2.20), we construct the ๐ง๐๐๐๐กโ-๐๐๐๐๐
deformation equation:
{(1 โ ๐)โ[ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐ข0(๐ฅ, ๐ผ)] = ๐ ๐ ๐(๐ฅ, ๐ผ)โณ (ฮฆ(๐ฅ, ๐, ๐ผ))
(1 โ ๐)โ[ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐ข0(๐ฅ, ๐ผ)] = ๐ ๐ ๐(๐ฅ, ๐ผ)โณ (ฮฆ(๐ฅ, ๐, ๐ผ)) (2.21)
Substituting equation (2.19) ๐๐๐ (2.20) ๐๐๐ก๐ (2.21), we obtain:
39
๐๐๐
(1 โ ๐)[ ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐ข0(๐ฅ, ๐ผ)] = ๐ ๐ ๐(๐ฅ, ๐ผ)[ ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐(๐ฅ, ๐ผ)
โ๐โซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, ๐, ๐ผ)๐๐ก
๐
๐
โฮปโซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, ๐, ๐ผ)๐๐ก],
๐ฅ
๐
(1 โ ๐)[ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐ข0(๐ฅ, ๐ผ)] = ๐ ๐ ๐(๐ฅ, ๐ผ)ฮฆ(๐ฅ, ๐, ๐ผ) โ ๐(๐ฅ, ๐ผ)
โ๐โซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, ๐, ๐ผ)๐๐ก
๐
๐
โฮปโซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, ๐, ๐ผ)๐๐ก
๐ฅ
๐
(2.22)
take ๐(๐ฅ, ๐ผ) = 1,when ๐ = 0, the zero-order deformation (2.22) becomes:
{ฮฆ(๐ฅ, 0, ๐ผ) = ๐ข0(๐ฅ, ๐ผ)
ฮฆ(๐ฅ, 0, ๐ผ) = ๐ข0(๐ฅ, ๐ผ) (2.23)
when ๐ = 1, the ๐ง๐๐๐๐กโ ๐๐๐๐๐ deformation (2.22) becomes:
{
ฮฆ(๐ฅ, 1, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, 1, ๐ผ)๐๐ก
๐
๐
+ ฮปโซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, 1, ๐ผ)๐๐ก
๐ฅ
๐
ฮฆ(๐ฅ, 1, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, 1, ๐ผ)๐๐ก
๐
๐
+ ฮปโซ๐(๐ฅ, ๐ก)ฮฆ(๐ก, 1, ๐ผ)๐๐ก
๐ฅ
๐
(2.24)
notice that this equation is the same as equation (2.18). Thus, as ๐
increases from 0 to 1, the analytical solution (ฮฆ(๐ฅ, ๐, ๐ผ),ฮฆ(๐ฅ, ๐, ๐ผ))
changes from the initial approximation (๐ข0(๐ฅ, ๐ผ), ๐ข0(๐ฅ, ๐ผ)) to the exact
solution (๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)).
40
Expanding ฮฆ(๐ฅ, ๐, ๐ผ), ๐๐๐ ฮฆ(๐ฅ, ๐, ๐ผ) into the Maclaurin series with
respect to ๐, we have:
{ฮฆ(๐ฅ, ๐, ๐ผ) = ๐ข0(๐ฅ, ๐ผ) + โ ๐ข๐(๐ฅ, ๐ผ)๐๐โ
๐=1
ฮฆ(๐ฅ, ๐, ๐ผ) = ๐ข0(๐ฅ, ๐ผ) + โ ๐ข๐(๐ฅ, ๐ผ)๐๐โ๐=1
(2.25)
where
๐ข๐(๐ฅ, ๐ผ) = 1
๐! ๐๐ฮฆ(๐ฅ, ๐, ๐ผ)
๐๐๐ ๐=0
๐ข๐(๐ฅ, ๐ผ) = [1
๐! ๐๐ฮฆ(๐ฅ,๐,๐ผ)
๐๐๐ ]๐=0
(2.26)
Differentiating equation (2.22) ๐ times with respect to ๐, we obtain [๐๐]:
๐๐๐
๐๐ฮฆ(๐ฅ, ๐, ๐ผ)
๐๐๐ โ
๐๐โ1ฮฆ(๐ฅ, ๐, ๐ผ)
๐๐๐โ1 = ๐ [
๐๐โ1ฮฆ(๐ฅ, ๐, ๐ผ)
๐๐๐โ1 โ ๐(๐ฅ, ๐ผ)
โ๐โซ๐(๐ฅ, ๐ก)๐๐โ1ฮฆ(๐ก, ๐, ๐ผ)
๐๐๐โ1 ๐๐ก
๐
๐
โ๐โซ๐(๐ฅ, ๐ก)๐๐โ1ฮฆ(๐ก, ๐, ๐ผ)
๐๐๐โ1 ๐๐ก
๐ฅ
๐
],
๐๐ฮฆ(๐ฅ, ๐, ๐ผ)
๐๐๐ โ
๐๐โ1ฮฆ(๐ฅ, ๐, ๐ผ)
๐๐๐โ1 = ๐ [
๐๐โ1ฮฆ(๐ฅ, ๐, ๐ผ)
๐๐๐โ1 โ ๐(๐ฅ, ๐ผ)
โฮปโซ๐(๐ฅ, ๐ก)๐๐โ1ฮฆ(๐ก, ๐, ๐ผ)
๐๐๐โ1
๐
๐
๐๐ก
โ๐โซ๐(๐ฅ, ๐ก)๐๐โ1ฮฆ(๐ก, ๐, ๐ผ)
๐๐๐โ1 ๐๐ก
๐ฅ
๐
].
(2.27)
now, dividing the equation(2.27) by ๐!, and setting ๐ = 0, we get the ๐th-
order deformation equation:
41
๐ข๐(๐ฅ, ๐ผ) = ๐๐ ๐ข๐โ1(๐ฅ, ๐ผ) + s[ ๐ข๐โ1(๐ฅ, ๐ผ) โ ํ๐๐(๐ฅ, ๐ผ)
โ ๐โซ๐(๐ฅ, ๐ก)๐ข๐โ1(๐ก, ๐, ๐ผ)๐๐ก
๐
๐
โ ๐โซ๐(๐ฅ, ๐ก)๐ข๐โ1(๐ก, ๐, ๐ผ)๐๐ก]
๐ฅ
๐
(2.27)
๐ข๐(๐ฅ, ๐ผ) = ๐๐ ๐ข๐โ1 (๐ฅ, ๐ผ) + s[ ๐ข๐โ1(๐ฅ, ๐ผ) โ ํ๐๐(๐ฅ, ๐ผ)
โ ๐โซ๐(๐ฅ, ๐ก)๐ข๐โ1(๐ก, ๐, ๐ผ)๐๐ก
๐
๐
โ ๐โซ๐(๐ฅ, ๐ก)๐ข๐โ1(๐ก, ๐, ๐ผ)๐๐ก]
๐ฅ
๐
where ๐ โฅ 1, ๐๐๐
๐๐ = {0, ๐ = 11, ๐ โ 1
,
ํ๐ = {0, ๐ โ 11, ๐ = 1
,
If we take ๐ข0(๐ฅ, ๐ผ) = ๐ข0(๐ฅ, ๐ผ) = 0, for ๐ โฅ 2, we have:
๐ข๐(๐ฅ, ๐ผ) = (๐ + 1)๐ข๐โ1(๐ฅ, ๐ผ) โ ๐ ๐(โซ๐(๐ฅ, ๐ก)๐ข๐โ1(๐ก, ๐, ๐ผ)๐๐ก
๐
๐
+ โซ๐(๐ฅ, ๐ก)๐ข๐โ1(๐ก, ๐, ๐ผ)๐๐ก ),
๐ฅ
๐
๐ข๐(๐ฅ, ๐ผ) = (๐ + 1)๐ข๐โ1(๐ฅ, ๐ผ) โ ๐ ๐(โซ๐(๐ฅ, ๐ก)๐ข๐โ1(๐ก, ๐, ๐ผ)๐๐ก
๐
๐
+โซ๐(๐ฅ, ๐ก)๐ข๐โ1(๐ก, ๐, ๐ผ)๐๐ก ).
๐ฅ
๐
(2.29)
hence, the solution of the system (2.18) in series form is:
42
๐ข(๐ฅ, ๐ผ) = lim๐โ1
ฮฆ(๐ฅ, ๐, ๐ผ) = โ ๐ข๐
โ
๐=1
(๐ฅ, ๐ผ),
๐ข(๐ฅ, ๐ผ) = lim๐โ1
ฮฆ(๐ฅ, ๐, ๐ผ) = โ ๐ข๐
โ
๐=1
(๐ฅ, ๐ผ).
(2.30)
We denote the ๐๐กโ โorder approximation to the solutions
๐ข(๐ฅ, ๐ผ), ๐๐๐ ๐ข(๐ฅ, ๐ผ) respectively with:
๐ข๐(๐ฅ, ๐ผ) = โ ๐ข๐
๐
๐=1
(๐ฅ, ๐ผ),
๐ข๐(๐ฅ, ๐ผ) = โ ๐ข๐
๐
๐=1
(๐ฅ, ๐ผ).
Example (๐. ๐):
Consider the fuzzy Volterra integral equation
๐ข(๐ฅ) = (๐ผ + 1,3 โ ๐ผ) + โซ (๐ฅ โ ๐ก)
๐ฅ
0
๐ข(๐ก, ๐ผ)๐๐ก, ๐ฅ โ [0, ๐], ๐ < โ
Solution:
In this example, ๐(๐ฅ, ๐ก) โฅ 0 for each 0โค ๐ก โค ๐ฅ, in this case ๐ = ๐ฅ,
๐ = 0, ๐ = 1. By equation (2.27), the first terms of homotopy analysis
method series are:
๐ข0(๐ฅ, ๐ผ) = 0,
๐ข1(๐ฅ, ๐ผ) = โ๐ ๐(๐ฅ, ๐ผ) = โ๐ (๐ผ + 1)
43
๐ข2(๐ฅ, ๐ผ) = (1 + ๐ )๐ข1(๐ฅ, ๐ผ) โ ๐ โซ(๐ฅ โ ๐ก) ๐ข1(๐ก, ๐ผ)๐๐ก
๐ฅ
0
= โ๐ (๐ผ + 1) 1 + ๐ (1 โ๐ฅ2
2)
๐ข3(๐ฅ, ๐ผ) = (1 + ๐ ) ๐ข2(๐ฅ, ๐ผ) โ ๐ โซ (๐ฅ โ ๐ก) ๐ข2(๐ก, ๐ผ)๐๐ก ๐ฅ
0
= โ๐ (๐ผ + 1) 1 + ๐ (2 โ ๐ฅ2) + ๐ 2 (1 โ ๐ฅ2 +๐ฅ4
24)
and
๐ข0(๐ฅ, ๐ผ) = 0
๐ข1(๐ฅ, ๐ผ) = โ๐ ๐(๐ฅ, ๐ผ) = โ๐ (3 โ ๐ผ)
๐ข2(๐ฅ, ๐ผ) = (1 + ๐ )๐ข1(๐ฅ, ๐ผ) โ ๐ โซ(๐ฅ โ ๐ก) ๐ข1(๐ก, ๐ผ)๐๐ก
๐ฅ
0
= โ๐ (3 โ ๐ผ) 1 + ๐ (1 โ๐ฅ2
2)
๐ข3(๐ฅ, ๐ผ) = (1 + ๐ ) ๐ข2(๐ฅ, ๐ผ) โ ๐ โซ (๐ฅ โ ๐ก)๐ข2 (๐ก, ๐ผ)๐๐ก ๐ฅ
0
= โ๐ (3 โ ๐ผ) 1 + ๐ (2 โ ๐ฅ2) + ๐ 2 (1 โ ๐ฅ2 +๐ฅ4
24)
then we approximate ๐ข(๐ฅ, ๐ผ) with
๐ข3(๐ฅ, ๐ผ) = โ ๐ข๐
3
๐=1
(๐ฅ, ๐ผ)
= โ๐ (๐ผ + 1) 3 + ๐ (3 โ3
2๐ฅ2) + ๐ 2 (1 โ ๐ฅ2 +
๐ฅ4
24)
44
and ๐ข(๐ฅ, ๐ผ) with
๐ข(๐ฅ, ๐ผ) = โ ๐ข๐
3
๐=1
(๐ฅ, ๐ผ)
= โ๐ (3 โ ๐ผ) 3 + ๐ (3 โ3
2๐ฅ2) + ๐ 2 (1 โ ๐ฅ2 +
๐ฅ4
24)
The exact solution of system is given by:
๐ข(๐ฅ, ๐ผ) = lim๐โ1
ฮฆ(๐ฅ, ๐, ๐ผ) = โ ๐ข๐(๐ฅ, ๐ผ),
โ
๐=1
and
๐ข(๐ฅ, ๐ผ) = lim๐โ1
ฮฆ(๐ฅ, ๐, ๐ผ) = โ ๐ข๐(๐ฅ, ๐ผ).
โ
๐=1
hence, the solution is:
๐ข(๐ฅ, ๐ผ) = โ๐ (๐ผ + 1) 3 + ๐ (3 โ3
2๐ฅ2) + ๐ 2 (1 โ ๐ฅ2 +
๐ฅ4
24) + โฏ ,
๐ข(๐ฅ, ๐ผ) = โ๐ (3 โ ๐ผ) 3 + ๐ (3 โ3
2๐ฅ2) + ๐ 2 (1 โ ๐ฅ2 +
๐ฅ4
24) + โฏ
note that the solution series contains the auxiliary parameter ๐ , which
provides a convenient way to controlling the convergence region of series
solution. we choose ๐ = โ1, we have:
๐ข(๐ฅ, ๐ผ) = (๐ผ + 1) 1 +๐ฅ2
2+
๐ฅ4
24+ โฏ = (๐ผ + 1)โ
๐ฅ2๐
(2๐)!
โ
๐=0
๐ข(๐ฅ, ๐ผ) = (3 โ ๐ผ) 1 +๐ฅ2
2+
๐ฅ4
24+ โฏ = (3 โ ๐ผ)โ
๐ฅ2๐
(2๐)!
โ
๐=0
45
as ๐ โ โ, the solution series converges to:
{๐ข(๐ฅ, ๐ผ) = (๐ผ + 1) cosh(๐ฅ)
๐ข(๐ฅ, ๐ผ) = (3 โ ๐ผ) cosh(๐ฅ) 0 โค ๐ผ โค 1
2.4 Adomian Decomposition Method (๐๐๐)
The Adomian decomposition method has been applied by scientists and
engeneers since the beginning of the 1980(s)[15]. this method gives the
solution as infinite series usually converges to the closed form solution.
The Adomian decomposition method is a special case of homotopy
analysis method, if we take ๐ = โ1 in a homotopy frame then it's
converted to the Adomian decomposition method. Consider the fuzzy
Volterra integral equation:
{
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)(๐ข(๐ก, ๐ผ))
๐ฅ
๐
๐๐ก
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)(๐ข(๐ก, ๐ผ))
๐ฅ
๐
๐๐ก
(2.31)
where ๐ โค ๐ก โค ๐ฅ, ๐ฅ โ [๐, ๐], 0 โค ๐ผ โค 1, ๐๐๐
๐(๐ฅ, ๐ก)๐ข(๐ฅ, ๐ผ) = {๐(๐ฅ, ๐ก)๐ข(๐ฅ, ๐ผ), ๐(๐ฅ, ๐ก) โฅ 0,
๐(๐ฅ, ๐ก)๐ข(๐ฅ, ๐ผ), ๐(๐ฅ, ๐ก) < 0.
๐(๐ฅ, ๐ก)๐ข(๐ฅ, ๐ผ) = {๐(๐ฅ, ๐ก)๐ข(๐ฅ, ๐ผ), ๐(๐ฅ, ๐ก) โฅ 0,
๐(๐ฅ, ๐ก)๐ข(๐ฅ, ๐ผ), ๐(๐ฅ, ๐ก) < 0.
for each 0 โค ๐ผ โค 1, and ๐ โค ๐ฅ โค ๐.
46
In order to solve equation(2.31) by the Adomian decomposition method,
we express the solution for the unknown functions
[ ๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)] in series form[15]:
{
๐ข(๐ฅ, ๐ผ) = โ๐ข๐(๐ฅ, ๐ผ)
โ
๐=0
๐ข(๐ฅ, ๐ผ) = โ๐ข๐(๐ฅ, ๐ผ)
โ
๐=0
(2.32)
now, setting (2.32) ๐๐๐ก๐ (2.31) yields:
{
โ ๐ข๐(๐ฅ, ๐ผ) =
โ
๐=0
๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก) (โ ๐ข๐(๐ก, ๐ผ)
โ
๐=0
) ๐๐ก
๐ฅ
๐
โ ๐ข๐(๐ฅ, ๐ผ) =
โ
๐=0
๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก) (โ ๐ข๐(๐ก, ๐ผ)
โ
๐=0
) ๐๐ก
๐ฅ
๐
(2.33)
or
{
๐ข0(๐ฅ, ๐ผ) + โฏ = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)(๐ข0(๐ก, ๐ผ) + ๐ข1(๐ก, ๐ผ) + โฏ )๐๐ก
๐ฅ
๐
๐ข0(๐ฅ, ๐ผ) + โฏ = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)(๐ข0(๐ก, ๐ผ) + ๐ข1(๐ก, ๐ผ) + โฏ )๐๐ก
๐ฅ
๐
(2.34)
The ๐ง๐๐๐๐กโ ๐๐๐๐๐๐๐๐๐ก can be identified by all the terms not included
under the integral sign, that is:
๐ข0(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ)
๐ข0(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) (2.35)
47
whereas the ๐๐กโ โ ๐๐๐๐๐๐๐๐๐ก๐ , ๐ โฅ 1 are given by the recurrence
relation:
๐ข๐+1(๐ฅ, ๐ผ) = ๐โซ๐(๐ฅ, ๐ก)( ๐ข๐(๐ก, ๐ผ) )๐๐ก,
๐ฅ
๐
(2.36)
๐ข๐+1(๐ฅ, ๐ผ) = ๐โซ๐(๐ฅ, ๐ก)( ๐ข๐(๐ก, ๐ผ) )๐๐ก, ๐ โฅ 0
๐ฅ
๐
we approximate ๐ข(๐ฅ, ๐ผ) = [ ๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ) ] by [45]:
๐๐ = โ ๐ข๐(๐ฅ, ๐ผ)
๐โ1
๐=0
๐๐๐ ๐๐
= โ ๐ข๐(๐ฅ, ๐ผ)
๐โ1
๐=0
where
lim๐โโ
๐๐ = ๐ข(๐ฅ, ๐ผ)
and
lim๐โโ
๐๐
= ๐ข(๐ฅ, ๐ผ)
Example (๐. ๐):
Consider the fuzzy Volterra integral equation:
๐ข(๐ฅ) = (๐ผ + 1,3 โ ๐ผ) + โซ (๐ฅ โ ๐ก)
๐ฅ
0
๐ข(๐ก)๐๐ก, ๐ฅ โ [0, ๐], ๐ < โ
48
Solution:
In this example ๐(๐ฅ, ๐ก) โฅ 0, ๐๐๐ ๐๐๐โ 0 โค ๐ก โค ๐ฅ.
the first terms of Adomain decomposition series are:
๐ข0(๐ฅ, ๐ผ) = (๐ผ + 1)
๐ข1(๐ฅ, ๐ผ) = โซ(๐ฅ โ ๐ก)
๐ฅ
0
๐ข0(๐ก, ๐ผ)๐๐ก = โซ(๐ฅ โ ๐ก)
๐ฅ
0
(๐ผ + 1)๐๐ก = (๐ผ + 1)(๐ฅ2
2)
๐ข2(๐ฅ, ๐ผ) = โซ(๐ฅ โ ๐ก)
๐ฅ
0
๐ข1(๐ก, ๐ผ)๐๐ก = โซ(๐ฅ โ ๐ก) (๐ก2
2)
๐ฅ
0
(๐ผ + 1)๐๐ก
= (๐ผ + 1) (๐ฅ4
24)
๐ข3(๐ฅ, ๐ผ) = โซ(๐ฅ โ ๐ก)
๐ฅ
0
๐ข2(๐ก, ๐ผ)๐๐ก = โซ(๐ฅ โ ๐ก) (๐ก4
24)
๐ฅ
0
(๐ผ + 1)๐๐ก
= (๐ผ + 1) (๐ฅ6
720)
and
๐ข0(๐ฅ, ๐ผ) = (3 โ ๐ผ)
๐ข1(๐ฅ, ๐ผ) = โซ(๐ฅ โ ๐ก)
๐ฅ
0
๐ข0(๐ก, ๐ผ)๐๐ก = โซ(๐ฅ โ ๐ก)
๐ฅ
0
(3 โ ๐ผ)๐๐ก = (3 โ ๐ผ)(๐ฅ2
2)
๐ข2(๐ฅ, ๐ผ) = โซ(๐ฅ โ ๐ก)
๐ฅ
0
๐ข1(๐ก, ๐ผ)๐๐ก = โซ(๐ฅ โ ๐ก) (๐ก2
2)
๐ฅ
0
(3 โ ๐ผ)๐๐ก
= (3 โ ๐ผ)(๐ฅ4
24)
49
๐ข3(๐ฅ, ๐ผ) = โซ(๐ฅ โ ๐ก)
๐ฅ
0
๐ข2(๐ก, ๐ผ)๐๐ก = โซ(๐ฅ โ ๐ก) (๐ก4
24)
๐ฅ
0
(3 โ ๐ผ)๐๐ก
= (3 โ ๐ผ)(๐ฅ6
720)
then, we approximate ๐ข(๐ฅ, ๐ผ) = [ ๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)] by:
๐4 = โ๐ข๐(๐ฅ, ๐ผ)
3
๐=0
= (๐ผ + 2) (1 +๐ฅ2
2!+
๐ฅ4
4!+
๐ฅ6
6!)
and
๐4= โ๐ข๐(๐ฅ, ๐ผ)
3
๐=0
= (3 โ ๐ผ) (1 +๐ฅ2
2!+
๐ฅ4
4!+
๐ฅ6
6!)
๐ข(๐ฅ, ๐ผ) = lim๐โโ
๐๐ =(๐ผ + 1) lim๐โโ
(1 +๐ฅ2
2!+
๐ฅ4
4!+
๐ฅ6
6!+ โฏ)
= (๐ผ + 1) cosh(๐ฅ),
๐ข(๐ฅ, ๐ผ) = lim๐โโ
๐๐
=(3 โ ๐ผ) lim๐โโ
(1 +๐ฅ2
2!+
๐ฅ4
4!+
๐ฅ6
6!+ โฏ)
= (3 โ ๐ผ) cosh(๐ฅ).
2.5 Fuzzy Differential Transformation Method
In this section FDTM is applied to fuzzy Volterra integral equation of the
second kind, an example is shown to illustrate the superiority of this
method.
50
As special case of definition (1.23) if ๐ข(๐ฅ) is a fuzzy-valued function, we
get the following result.
Definition(๐. ๐)[๐๐]: (๐๐ญ๐ซ๐จ๐ง๐ ๐ฅ๐ฒ ๐ ๐๐ง๐๐ซ๐๐ฅ๐ข๐ณ๐๐ ๐๐ข๐๐๐๐ซ๐๐ง๐ญ๐ข๐๐๐ฅ๐ ๐จ๐ ๐ญ๐ก๐
๐ โ ๐๐ ๐จ๐ซ๐๐๐ซ ๐๐ญ ๐๐)
Let ๐ข: (๐, ๐) โ ๐ธ and ๐ฅ0 โ (๐, ๐). we say that ๐ข is strongly generalized
differential of the ๐๐กโ ๐๐๐๐๐ at ๐ฅ0. If there exists an element ๐ข(๐)(๐ฅ0) โ
๐ธ, โ๐ = 1,2,โฆ , ๐, such that:
(๐) โ โ > 0 Sufficiently small, โ๐ข(๐โ1)(๐ฅ0 + โ) โ ๐ข(๐โ1)(๐ฅ0),
โ๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 โ โ), and the limits (in the metric ๐)
๐๐๐โโ0
๐ข(๐โ1)(๐ฅ0 + โ) โ ๐ข(๐โ1)(๐ฅ0)
โ= ๐๐๐
โโ0
๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 โ โ)
โ
= ๐ข(๐)(๐ฅ0)
(๐๐) โ โ > 0 Sufficiently small, โ๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 + โ),
โ๐ข(๐โ1)(๐ฅ0 โ โ) โ ๐ข(๐โ1)(๐ฅ0) and the limits (in the metric d)
๐๐๐โโ0
๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 + โ)
โ= ๐๐๐
โโ0
๐ข(๐โ1)(๐ฅ0 โ โ) โ ๐ข(๐โ1)(๐ฅ0)
โ
= ๐ข(๐โ1)(๐ฅ0),
(๐๐๐) โ โ > 0 Sufficiently small, โ๐ข(๐โ1)(๐ฅ0)(๐ฅ0 + โ) โ ๐ข(๐โ1)(๐ฅ0),
โ๐ข(๐โ1)(๐ฅ0)(๐ฅ0 โ โ) โ ๐ข(๐โ1)(๐ฅ0) and the limits (in the metric d)
51
๐๐๐โโ0
๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 + โ)
โโ= ๐๐๐
โโ0
๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 โ โ)
โโ
= ๐ข(๐)(๐ฅ0),
(๐๐ฃ) โ โ > 0 Sufficiently small, โ๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 + โ),
โ๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 โ โ), and the limits (in the metric d)
๐๐๐โโโ0
๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 + โ)
โโ= ๐๐๐
โโ0
๐ข(๐โ1)(๐ฅ0) โ ๐ข(๐โ1)(๐ฅ0 โ โ)
โ
= ๐ข(๐)(๐ฅ0).
Theorem(๐. ๐)[๐๐]:
Let ๐ข: ๐ โ ๐ธ be a fuzzy-valued function denoted by:
๐ข(๐ฅ) = (๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)) , ๐๐๐ ๐๐๐โ ๐ผ โ [0,1], ๐กโ๐๐:
(1) If ๐ข(๐ฅ) is (๐) โ differentiable, then ๐ข(๐ฅ, ๐ผ) ๐๐๐ ๐ข(๐ฅ, ๐ผ) are
differentiable functions, and ๐ขโฒ(๐ฅ) = (๐ขโฒ(๐ฅ, ๐ผ), ๐ขโฒ(๐ฅ, ๐ผ)).
(2) If ๐ข(๐ฅ) is (๐๐) โ differentiable, then ๐ข(๐ฅ, ๐ผ) ๐๐๐ ๐ข(๐ฅ, ๐ผ) are
differentiable functions, and ๐ขโฒ(๐ฅ) = ( ๐ขโฒ(๐ฅ, ๐ผ) , ๐ขโฒ(๐ฅ, ๐ผ) ).
Definition (๐. ๐) [๐๐]:
Let ๐ข(๐ฅ, ๐ผ) be strongly generalized differentiable of order ๐ in time
domain ๐, then if ๐ข(๐ฅ, ๐ผ) is differentiable in first form (๐)
๐ (๐ฅ, ๐, ๐ผ) = ๐๐ (๐ข (๐ฅ, ๐ผ))
๐ ๐ฅ๐ , ๐ (๐ฅ, ๐, ๐ผ) =
๐๐(๐ข (๐ฅ, ๐ผ))
๐ ๐ฅ๐ , โ๐ฅ โ ๐
52
then
๐๐(๐, ๐ผ) =
{
๐๐(๐, ๐ผ) = ๐ (๐ฅ๐ , ๐, ๐ผ) =
๐๐ (๐ข (๐ฅ, ๐ผ))
๐ ๐ฅ๐| ๐ฅ=๐ฅ๐
๐๐(๐, ๐ผ) = ๐(๐ฅ๐ , ๐, ๐ผ) = ๐๐(๐ข (๐ฅ, ๐ผ))
๐ ๐ฅ๐|๐ฅ=๐ฅ๐
, โ๐ โ ๐
If ๐ข(๐ฅ, ๐ผ) is differentiable in second form (๐๐)
๐ (๐ฅ, ๐, ๐ผ) = ๐๐(๐ข (๐ฅ, ๐ผ))
๐ ๐ฅ๐ , ๐ (๐ฅ, ๐, ๐ผ) =
๐๐ (๐ข (๐ฅ, ๐ผ))
๐ ๐ฅ๐ , โ๐ฅ โ ๐
then
๐๐(๐, ๐ผ) =
{
๐๐(๐, ๐ผ) = ๐ (๐ฅ๐ , ๐, ๐ผ) =
๐๐(๐ข(๐ฅ,๐ผ))
๐ ๐ฅ๐ |๐ฅ=๐ฅ๐
๐๐(๐, ๐ผ) = ๐ (๐ฅ๐ , ๐, ๐ผ) = ๐๐(๐ข (๐ฅ,๐ผ))
๐ ๐ฅ๐|๐ฅ=๐ฅ๐
, ๐ ๐๐ ๐๐ฃ๐๐
๐๐(๐, ๐ผ) =
{
๐๐(๐, ๐ผ) = ๐ (๐ฅ๐ , ๐, ๐ผ) =
๐๐(๐ข (๐ฅ,๐ผ))
๐ ๐ฅ๐|๐ฅ=๐ฅ๐
๐๐(๐, ๐ผ) = ๐ (๐ฅ๐ , ๐, ๐ผ) = ๐๐(๐ข(๐ฅ,๐ผ))
๐ ๐ฅ๐ |๐ฅ=๐ฅ๐
, ๐ ๐๐ ๐๐๐
where ๐๐(๐, ๐ผ) and ๐๐(๐, ๐ผ) are the lower and the upper spectrum of
๐ข(๐ฅ) at ๐ฅ = ๐ฅ๐ in the domain ๐, respectively.
If ๐ข(๐ฅ, ๐ผ) is (๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐, then ๐ข(๐ฅ, ๐ผ) can be represented as
follows:
53
{
๐ข(๐ฅ, ๐ผ) = โ
(๐ฅ โ ๐ฅ๐)๐
๐!๐(๐, ๐ผ)
โ
๐=0
๐ข(๐ฅ, ๐ผ) = โ(๐ฅ โ ๐ฅ๐)
๐
๐!๐(๐, ๐ผ)
โ
๐=0
, ๐ โ ๐
and if ๐ข(๐ฅ, ๐ผ) is (๐๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐, then ๐ข(๐ฅ, ๐ผ) can be represented
as follows:
{
๐ข(๐ฅ, ๐ผ) = โ
(๐ฅ โ ๐ฅ๐)๐
๐!๐(๐, ๐ผ) + โ
(๐ฅ โ ๐ฅ๐)๐
๐!๐(๐, ๐ผ)
โ
๐=1,๐๐๐
โ
๐=0,๐๐ฃ๐๐
๐ข(๐ฅ, ๐ผ) = โ(๐ฅ โ ๐ฅ๐)
๐
๐!๐(๐, ๐ผ) + โ
(๐ฅ โ ๐ฅ๐)๐
๐!๐(๐, ๐ผ)
โ
๐=1,๐๐๐
โ
๐=0,๐๐ฃ๐๐
(2.37)
equation (2.37) is known as the inverse transformation of ๐(๐, ๐ผ).
If ๐ข(๐ฅ, ๐ผ) is (๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐ then ๐(๐, ๐ผ) is defined as:
๐(๐, ๐ผ) = ๐(๐) [๐๐ (๐(๐ฅ)๐ข(๐ฅ, ๐ผ))
๐๐ฅ๐]
๐ฅ=๐ฅ0
,
๐(๐, ๐ผ) = ๐(๐) ๐๐(๐(๐ฅ)๐ข(๐ฅ, ๐ผ))
๐๐ฅ๐ ๐ฅ=๐ฅ0
.
if ๐ข(๐ฅ, ๐ผ) is (๐๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐ then ๐(๐, ๐ผ) can be described as:
๐(๐, ๐ผ) =
{
๐(๐, ๐ผ) = ๐(๐) [
๐๐ (๐(๐ฅ)๐ข(๐ฅ, ๐ผ))
๐๐ฅ๐]
๐ฅ=๐ฅ0
๐(๐, ๐ผ) = ๐(๐) ๐๐(๐(๐ฅ)๐ข(๐ฅ, ๐ผ))
๐๐ฅ๐ ๐ฅ=๐ฅ0
, ๐ ๐๐ ๐๐ฃ๐๐
54
(2.38)
๐(๐, ๐ผ) =
{
๐(๐, ๐ผ) = ๐(๐)
๐๐(๐(๐ฅ)๐ข(๐ฅ, ๐ผ))
๐๐ฅ๐ ๐ฅ=๐ฅ0
๐(๐, ๐ผ) = ๐(๐) ๐๐(๐(๐ฅ)๐ข(๐ฅ, ๐ผ))
๐๐ฅ๐ ๐ฅ=๐ฅ0
, ๐ ๐๐ ๐๐๐
now, if ๐ข(๐ฅ, ๐ผ)๐๐ (๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐ then ๐ข(๐ฅ, ๐ผ) can be represented as
follows:
๐ข(๐ฅ, ๐ผ) =
{
๐ข(๐ฅ, ๐ผ) =
1
๐(๐ฅ)โ
(๐ฅ โ ๐ฅ0)๐
๐!
โ
๐=0
๐(๐, ๐ผ)
๐(๐)
๐ข(๐ฅ, ๐ผ) =1
๐(๐ฅ)โ
(๐ฅ โ ๐ฅ0)๐
๐!
โ
๐=0
๐(๐, ๐ผ)
๐(๐)
and if ๐ข(๐ฅ, ๐ผ)๐๐ (๐๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐ then ๐ข(๐ฅ, ๐ผ) defined as:
{
1
๐(๐ฅ)( โ
(๐ฅ โ ๐ฅ0)๐
๐!
โ
๐=0,๐๐ฃ๐๐
๐(๐, ๐ผ)
๐(๐)+ โ
(๐ฅ โ ๐ฅ0)๐
๐!
โ
๐=1,๐๐๐
๐(๐, ๐ผ)
๐(๐))
1
๐(๐ฅ)( โ
(๐ฅ โ ๐ฅ0)๐
๐!
โ
๐=0,๐๐ฃ๐๐
๐(๐, ๐ผ)
๐(๐)+ โ
(๐ฅ โ ๐ฅ0)๐
๐!
โ
๐=1,๐๐๐
๐(๐, ๐ผ)
๐(๐))
(2.39)
where ๐(๐) > 0 and it is the weighting factor and ๐(๐ฅ) > 0 is regard as a
kernel corresponding to ๐ข(๐ฅ, ๐ผ). If ๐(๐) = 1 ๐๐๐ ๐(๐ฅ) = 1, then (2.37)
is a special case of (2.39). in this section the transformation
๐(๐) =๐ป๐
๐! ๐๐๐ ๐(๐ฅ) = 1 will be applied, where ๐ป is the time horizon of
interest. So, the equations (2.38) become
If ๐ข(๐ฅ, ๐ผ) is (๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐ then ๐(๐, ๐ผ) is defined as:
๐(๐, ๐ผ) =๐ป๐
๐!
๐๐๐ข(๐ฅ, ๐ผ)
๐๐ฅ๐,
55
and
๐(๐, ๐ผ) =๐ป๐
๐!
๐๐๐ข(๐ฅ, ๐ผ)
๐๐ฅ๐
if ๐ข(๐ฅ, ๐ผ) is (๐๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐ then ๐(๐, ๐ผ) can be described as:
๐(๐, ๐ผ) = ๐(๐, ๐ผ) =
๐ป๐
๐!
๐๐๐ข(๐ฅ, ๐ผ)
๐๐ฅ๐
๐(๐, ๐ผ) = ๐ป๐
๐!
๐๐๐ข(๐ฅ, ๐ผ)
๐๐ฅ๐
, ๐ ๐๐ ๐๐ฃ๐๐
(2.40)
๐(๐, ๐ผ) = ๐(๐, ๐ผ) =
๐๐๐ข(๐ฅ, ๐ผ)
๐๐ฅ๐
๐(๐, ๐ผ) =๐๐๐ข(๐ฅ, ๐ผ)
๐๐ฅ๐
, ๐ ๐๐ ๐๐๐
Using the fuzzy differential transformation, a fuzzy differential equation in
the domain of interest can be transformed to an algebraic equation in the
domain T, and ๐ข(๐ฅ) can be obtained as the finite- term Taylor series plus
reminder, as:
if ๐ข(๐ฅ, ๐ผ)๐๐ (๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐ then ๐ข(๐ฅ, ๐ผ) becomes:
๐ข(๐ฅ, ๐ผ) =
{
๐ข(๐ฅ, ๐ผ) =
1
๐(๐ฅ)โ
(๐ฅ โ ๐ฅ0)๐
๐!
โ
๐=0
๐(๐, ๐ผ)
๐(๐)+ ๐ ๐+1(๐ฅ)
๐ข(๐ฅ, ๐ผ) =1
๐(๐ฅ)โ
(๐ฅ โ ๐ฅ0)๐
๐!
โ
๐=0
๐(๐, ๐ผ)
๐(๐)+ ๐ ๐+1(๐ฅ)
hence
56
๐ข(๐ฅ, ๐ผ) =
{
๐ข(๐ฅ, ๐ผ) =
1
๐(๐ฅ)โ (
๐ฅ โ ๐ฅ0
๐ป)๐
๐(๐, ๐ผ)
โ
๐=0
+ ๐ ๐+1(๐ฅ)
๐ข(๐ฅ, ๐ผ) =1
๐(๐ฅ)โ (
๐ฅ โ ๐ฅ0
๐ป)๐
โ
๐=0
๐(๐, ๐ผ) + ๐ ๐+1(๐ฅ)
(2.41)
and, if ๐ข(๐ฅ, ๐ผ)๐๐ (๐๐) โ ๐๐๐๐๐๐๐๐๐ก๐๐๐๐๐ then ๐ข(๐ฅ, ๐ผ) becomes:
๐ข(๐ฅ, ๐ผ) =1
๐(๐ฅ)
(
โ(๐ฅ โ ๐ฅ0)
๐
๐!
โ
๐=0,๐๐ฃ๐๐
๐(๐, ๐ผ)
๐(๐)
+ โ(๐ฅ โ ๐ฅ0)
๐
๐!
โ
๐=1,๐๐๐
๐(๐, ๐ผ)
๐(๐))
+ ๐ ๐+1(๐ฅ),
๐ข(๐ฅ, ๐ผ) =1
๐(๐ฅ)
(
โ(๐ฅ โ ๐ฅ0)
๐
๐!
โ
๐=0,๐๐ฃ๐๐
๐(๐, ๐ผ)
๐(๐)
+ โ(๐ฅ โ ๐ฅ0)
๐
๐!
โ
๐=1,๐๐๐
๐(๐, ๐ผ)
๐(๐))
+ ๐ ๐+1(๐ฅ)
hence,
๐ข(๐ฅ, ๐ผ) =1
๐(๐ฅ)
(
โ (๐ฅ โ ๐ฅ0
๐ป)๐
๐(๐, ๐ผ)
โ
๐=0,๐๐ฃ๐๐
+ โ (๐ฅ โ ๐ฅ0
๐ป)๐
โ
๐=1,๐๐๐
๐(๐, ๐ผ))
+ ๐ ๐+1(๐ฅ),
(2.42)
๐ข(๐ฅ, ๐ผ) =1
๐(๐ฅ)
(
โ (๐ฅ โ ๐ฅ0
๐ป)๐
๐(๐, ๐ผ)
โ
๐=0,๐๐ฃ๐๐
+ โ (๐ฅ โ ๐ฅ0
๐ป)๐
โ
๐=1,๐๐๐
๐(๐, ๐ผ))
+ ๐ ๐+1(๐ฅ)
57
Our aim is finding the solution of the Volterra integral equation (1.17) at
the equally spaced grid points {๐ฅ0, ๐ฅ1, โฆ , ๐ฅ๐} where ๐ฅ0 = ๐, ๐ฅ๐ = ๐,
๐ฅ๐ = ๐ + ๐โ ๐๐๐ ๐๐๐โ ๐ = 0,1,2,โฆ , ๐, ๐๐๐ โ =๐โ๐
๐. This means, the
domain of interest is divided to ๐ sub-domain, and the fuzzy
approximation functions in each sub-domain are ๐ข๐(๐ฅ, ๐ผ) for ๐ =
0,1โฆ , ๐ โ 1, respectively.
Definition (๐. ๐๐)[๐]: (Fuzzy differential transformation)
Let ๐ข(๐ฅ, ๐ผ) be differentiable fuzzy-valued function, we define the one-
dimensional differential transform by:
๐(๐, ๐ผ) =
{
๐(๐, ๐ผ) =
1
๐! ๐๐๐ข(๐ฅ, ๐ผ)
๐๐ฅ๐ ๐ฅ=0
๐(๐, ๐ผ) = 1
๐! ๐๐๐ข(๐ฅ, ๐ผ)
๐๐ฅ๐ ๐ฅ=0
where ๐ข(๐ฅ) is the original function and ๐(๐) is the transformed function.
Definition (๐. ๐๐)[๐]: (Fuzzy differential inverse transformation) The
differential inverse transform of ๐(๐, ๐ผ) on the girds ๐ฅ๐+1 is defined by:
๐ข(๐ฅ๐+1, ๐ผ) โ ๐ข๐(๐ฅ๐+1, ๐ผ)
= ๐๐(0, ๐ผ) + ๐๐(1, ๐ผ)(๐ก1 โ ๐ก0) + ๐๐(2, ๐ผ)(๐ก2 โ ๐ก1)2 + โฏ
+ ๐๐(๐, ๐ผ)(๐ก๐+1 โ ๐ก๐)๐ = โ ๐๐(๐, ๐ผ)โ๐
๐
๐=0
,
and
58
๐ข(๐ฅ๐+1, ๐ผ) โ ๐ข๐(๐ฅ๐+1, ๐ผ)
= ๐๐(0, ๐ผ) + ๐๐(1, ๐ผ)(๐ก1 โ ๐ก0) + ๐๐(2, ๐ผ)(๐ก2 โ ๐ก1)2 + โฏ
+ ๐๐(๐, ๐ผ)(๐ก๐+1 โ ๐ก๐)๐ = โ ๐๐(๐, ๐ผ)โ๐
๐
๐=0
.
Theorem (2.7 )[๐]:
Let ๐ข(๐ฅ, ๐ผ), ๐ฃ(๐ฅ, ๐ผ)๐๐๐ ๐ค(๐ฅ, ๐ผ) are fuzzy valued functions, and let
๐(๐ฅ, ๐ผ), ๐(๐ฅ, ๐ผ), ๐๐๐ ๐(๐ฅ, ๐ผ) are their differential transformations
respectively, then we get the following:
i. If ๐ข(๐ฅ, ๐ผ) = ๐ฃ(๐ฅ, ๐ผ) ยฑ ๐ค(๐ฅ, ๐ผ), ๐กโ๐๐ ๐(๐, ๐ผ) = ๐(๐, ๐ผ) ยฑ ๐(๐, ๐ผ).
ii. If ๐ข(๐ฅ, ๐ผ) = ๐ ๐ฃ(๐ฅ, ๐ผ), ๐กโ๐๐ ๐(๐, ๐ผ) = ๐ ๐(๐, ๐ผ),where a is a constant.
iii. if
๐ข(๐ฅ, ๐ผ) =๐๐๐ฃ(๐ฅ, ๐ผ)
๐๐ฅ๐ , ๐กโ๐๐ ๐(๐, ๐ผ) =
(๐+๐)!
๐!๐(๐ + ๐, ๐ผ)
iv. If ๐ข(๐ฅ, ๐ผ) = ๐ฅ๐, ๐กโ๐๐ ๐(๐, ๐ผ) = ๐ฟ(๐ โ ๐, ๐ผ), ๐คโ๐๐๐
๐ฟ(๐ โ ๐) = {1 ๐๐ ๐ = ๐0 ๐๐ ๐ โ ๐
v. If
๐ข(๐ฅ, ๐ผ) = ๐ฃ(๐ฅ, ๐ผ). ๐ค(๐ฅ, ๐ผ), ๐กโ๐๐ ๐(๐, ๐ผ) = โ๐(๐).๐(๐ โ ๐, ๐ผ)
๐
๐=0
.
vi. If ๐ข(๐ฅ, ๐ผ) = exp(๐ ๐ฅ, ๐ผ) , ๐กโ๐๐ ๐(๐, ๐ผ) = (๐๐
๐!โ , ๐ผ) , ๐คโ๐๐๐
๐ ๐๐ a constant.
59
vii. If
๐ข(๐ฅ, ๐ผ) = (๐ฅ + 1)๐, ๐กโ๐๐ ๐(๐, ๐ผ) = (๐(๐โ1)โฆ.(๐โ๐โ1)
๐!, ๐ผ)
viii. If
๐ข(๐ฅ, ๐ผ) = (sin(๐ ๐ฅ + ๐), ๐ผ), ๐กโ๐๐ ๐(๐, ๐ผ) = (๐๐
๐!sin ( ๐๐
2! + ๐),๐ผ)
๐คโ๐๐๐ ๐, ๐ ๐๐๐ ๐๐๐๐ ๐ก๐๐๐ก๐ .
ix. If
๐ข(๐ฅ, ๐ผ) = (cos(๐ ๐ฅ + ๐), ๐ผ), ๐กโ๐๐ ๐(๐, ๐ผ) = (๐๐
๐!cos( ๐๐
2! + ๐),๐ผ)
where ๐, ๐ are constants.
Proof: Using definition(2.9).
Theorem (2.8 )[๐]:
Suppose that ๐(๐, ๐ผ), ๐บ(๐, ๐ผ)๐๐๐ ๐(๐) are differential transformations of
the fuzzy- valued functions ๐ข(๐ฅ, ๐ผ) ๐๐๐ ๐(๐ฅ, ๐ผ), and positive real valued
function ๐ฃ(๐ฅ), respectively then the following are satisfied:
(1) If ๐(๐ฅ, ๐ผ) = โซ ๐ข(๐ก, ๐ผ)๐๐ก๐ฅ
๐ฅ0, 0 โค ๐ผ โค 1, ๐กโ๐๐
๐บ(๐, ๐ผ) =
{
๐บ(๐, ๐ผ) =๐(๐ โ 1, ๐ผ)
๐
๐บ(๐, ๐ผ) =๐(๐ โ 1, ๐ผ)
๐
,๐คโ๐๐๐ ๐ โฅ 1.
(2) If ๐(๐ฅ, ๐ผ) = โซ ๐ฃ(๐ก)๐ข(๐ก, ๐ผ)๐๐ก๐ฅ
๐ฅ0, 0 โค ๐ผ โค 1, ๐กโ๐๐
๐บ(๐, ๐ผ) =
{
๐บ(๐, ๐ผ) =
1
๐โ ๐(๐)๐(๐ โ ๐ โ 1, ๐ผ),
๐โ1
๐=0
๐บ(๐, ๐ผ) =1
๐โ ๐(๐)๐(๐ โ ๐ โ 1, ๐ผ),
๐โ1
๐=0
, ๐บ(0) = 0, ๐ โฅ 1.
(3) If ๐(๐ฅ, ๐ผ) = ๐ฃ(๐ฅ) โซ ๐ข(๐ก, ๐ผ)๐๐ก๐ฅ
๐ฅ0, 0 โค ๐ผ โค 1, ๐กโ๐๐
60
๐บ(๐, ๐ผ) =
{
๐บ(๐, ๐ผ) = โ
1
๐๐(๐ โ ๐)๐(๐ โ 1, ๐ผ),
๐
๐=1
๐บ(๐, ๐ผ) = โ1
๐๐(๐ โ ๐)๐(๐ โ 1, ๐ผ),
๐
๐=1
, ๐บ(0) = 0, ๐ โฅ 1.
(4) If ๐(๐ฅ, ๐ผ) = ๐ฃ(๐ฅ)๐ข(๐ฅ, ๐ผ), ๐กโ๐๐
๐บ(๐, ๐ผ) = โ๐(๐)๐(๐ โ ๐, ๐ผ),
๐
๐=0
Proof: see [6].
now, taking the fuzzy differential transformation for both sides of equation
(1.17), we get:
๐(๐, ๐ผ) =
{
๐(๐, ๐ผ) = ๐น(๐, ๐ผ) +
๐
๐โ ๐(๐)๐(๐ โ ๐ โ 1, ๐ผ)
๐โ1
๐=0
๐(๐, ๐ผ) = ๐น(๐, ๐ผ) +๐
๐โ ๐(๐)๐(๐ โ ๐ โ 1, ๐ผ)
๐โ1
๐=0
where ๐ > 0, ๐(๐, ๐ผ) ๐๐๐ ๐น(๐, ๐ผ)are fuzzy differential transformations
of fuzzy valued functions ๐ข(๐ฅ, ๐ผ), ๐๐๐ ๐(๐ฅ, ๐ผ) respectively, and ๐พ(๐) is a
fuzzy differential transformation of a positive real- valued function
๐(๐ฅ). ๐๐๐ ๐๐๐ ๐ ๐(0, ๐ผ) = ๐น(0, ๐ผ).
Example (๐. ๐):
Consider the fuzzy Volterra integral equation:
๐ข(๐ฅ) = (๐ผ + 1,3 โ ๐ผ) + โซ (๐ฅ โ ๐ก)
๐ฅ
0
๐ข(๐ก)๐๐ก, ๐ฅ โ [0, ๐], ๐ < โ
61
Solution:
๐ข(๐ฅ, ๐ผ) = (๐ผ + 1) + โซ(๐ฅ โ ๐ก)๐ข
๐ฅ
0
(๐ก, ๐ผ)๐๐ก,
๐ข(๐ฅ, ๐ผ) = (3 โ ๐ผ) + โซ(๐ฅ โ ๐ก)๐ข
๐ฅ
0
(๐ก, ๐ผ)๐๐ก.
taking fuzzy differential transformation, we get:
๐(๐, ๐ผ) = (๐ผ + 1)๐ฟ(๐ โ 0) + โ1
๐๐ฟ(๐ โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
๐
๐=1
โ1
๐โ ๐ฟ(๐ โ 1)๐(๐ โ ๐ โ 1, ๐ผ)
๐โ1
๐=0
,
and
๐(๐, ๐ผ) = (3 โ ๐ผ)๐ฟ(๐ โ 0) + โ1
๐๐ฟ(๐ โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
๐โ1
๐=0
โ1
๐โ ๐ฟ(๐ โ 1)๐(๐ โ ๐ โ 1, ๐ผ)
๐โ1
๐=0
.
First terms of differential transformation series are:
๐(0, ๐ผ) = (๐ผ + 1)
๐(1, ๐ผ) = (๐ผ + 1)๐ฟ(1 โ 0) + โ1
๐๐ฟ(1 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
1
๐=1
โ1
1โ ๐ฟ(๐ โ 1)๐(1 โ ๐ โ 1, ๐ผ)
๐โ1
๐=0
= 0
62
๐(2, ๐ผ) = (๐ผ + 1)๐ฟ(2 โ 0)
+ โ1
๐๐ฟ(2 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
2
๐=1
โ1
2โ๐ฟ(๐ โ 1)๐(2 โ ๐ โ 1, ๐ผ)
1
๐=0
=1
2๐(0, ๐ผ) =
1
2(๐ผ + 1)
๐(3, ๐ผ) = (๐ผ + 1)๐ฟ(3 โ 0)
+ โ1
๐๐ฟ(3 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
3
๐=1
โ1
3โ๐ฟ(๐ โ 1)๐(3 โ ๐ โ 1, ๐ผ)
2
๐=0
= 0
๐(4, ๐ผ) = (๐ผ + 1)๐ฟ(4 โ 0)
+ โ1
๐๐ฟ(4 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
4
๐=1
โ1
4โ๐ฟ(๐ โ 1)๐(4 โ ๐ โ 1, ๐ผ)
3
๐=0
=1
12๐(2, ๐ผ) =
1
24(๐ผ + 1)
๐(5, ๐ผ) = (๐ผ + 1)๐ฟ(5 โ 0)
+ โ1
๐๐ฟ(5 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
5
๐=1
โ1
5โ๐ฟ(๐ โ 1)๐(5 โ ๐ โ 1, ๐ผ)
4
๐=0
= 0
๐(6, ๐ผ) = (๐ผ + 1)๐ฟ(5 โ 0)
+ โ1
๐๐ฟ(6 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
6
๐=1
โ1
6โ๐ฟ(๐ โ 1)๐(6 โ ๐ โ 1, ๐ผ)
5
๐=0
=1
30๐(4, ๐ผ) =
1
720(๐ผ + 1)
63
and
๐(0, ๐ผ) = (3 โ ๐ผ)
๐(1, ๐ผ) = (3 โ ๐ผ)๐ฟ(1 โ 0) + โ1
๐๐ฟ(1 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
1
๐=1
โ1
1โ ๐ฟ(๐ โ 1)๐(1 โ ๐ โ 1, ๐ผ)
๐โ1
๐=0
= 0
๐(2, ๐ผ) = (3 โ ๐ผ)๐ฟ(2 โ 0)
+ โ1
๐๐ฟ(2 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
2
๐=1
โ1
2โ๐ฟ(๐ โ 1)๐(2 โ ๐ โ 1, ๐ผ)
1
๐=0
=1
2๐(0, ๐ผ) =
1
2(3 โ ๐ผ)
๐(3, ๐ผ) = (3 โ ๐ผ)๐ฟ(3 โ 0)
+ โ1
๐๐ฟ(3 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
3
๐=1
โ1
3โ๐ฟ(๐ โ 1)๐(3 โ ๐ โ 1, ๐ผ)
2
๐=0
= 0
๐(4, ๐ผ) = (3 โ ๐ผ)๐ฟ(4 โ 0)
+ โ1
๐๐ฟ(4 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
4
๐=1
โ1
4โ๐ฟ(๐ โ 1)๐(4 โ ๐ โ 1, ๐ผ)
3
๐=0
=1
12๐(2, ๐ผ) =
1
24(3 โ ๐ผ)
64
๐(5, ๐ผ) = (3 โ ๐ผ)๐ฟ(5 โ 0)
+ โ1
๐๐ฟ(5 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
5
๐=1
โ1
5โ๐ฟ(๐ โ 1)๐(5 โ ๐ โ 1, ๐ผ)
4
๐=0
= 0
๐(6, ๐ผ) = (3 โ ๐ผ)๐ฟ(5 โ 0)
+ โ1
๐๐ฟ(6 โ ๐ โ 1)๐(๐ โ 1, ๐ผ)
6
๐=1
โ1
6โ๐ฟ(๐ โ 1)๐(6 โ ๐ โ 1, ๐ผ)
5
๐=0
=1
30๐(4, ๐ผ) =
1
720( 3 โ ๐ผ)
therefore, the solution of fuzzy Volterra integral equation will be as
following:
๐ข(๐ฅ, ๐ผ) = โ ๐(๐, ๐ผ). ๐ฅ๐
โ
๐=0
= (๐ผ + 1) {1 +1
2!๐ฅ2 +
1
4!๐ฅ4 +
1
6!๐ฅ6 +
1
8!๐ฅ8 +
1
10!๐ฅ10 + โฏ},
and
๐ข(๐ฅ, ๐ผ) = โ ๐(๐, ๐ผ). ๐ฅ๐
โ
๐=0
= (3 โ ๐ผ) {1 +1
2!๐ฅ2 +
1
4!๐ฅ4 +
1
6!๐ฅ6 +
1
8!๐ฅ8 +
1
10!๐ฅ10 + โฏ}.
this converges to the closed form solution:
๐ข(๐ฅ, ๐ผ) = {๐ข(๐ฅ, ๐ผ) = (๐ผ + 1) cosh(๐ฅ)
๐ข(๐ฅ, ๐ผ) = (3 โ ๐ผ) cosh(๐ฅ)
65
2.6 Fuzzy successive approximation method
Consider the following fuzzy Volterra integral equation:
{
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ) ๐๐ก
๐ฅ
๐
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
(2.43)
where ๐ โค ๐ก โค ๐ฅ, ๐ฅ โ [๐, ๐], 0 โค ๐ผ โค 1, ๐๐๐
๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ) = {๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ), ๐(๐ฅ, ๐ก) โฅ 0
๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ), ๐(๐ฅ, ๐ก) < 0
and
๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ) = {๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ), ๐(๐ฅ, ๐ก) โฅ 0๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ), ๐(๐ฅ, ๐ก) < 0
for each 0 โค ๐ผ โค 1, ๐๐๐ ๐ โค ๐ฅ โค ๐.
The successive approximation method introduces the recurrence relation:
{
๐ข๐+1(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก , ๐ โฅ 0
๐ฅ
๐
๐ข๐+1(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก , ๐ โฅ 0
๐ฅ
๐
(2.44)
We can choose the ๐ง๐๐๐๐กโ โ ๐๐๐๐๐๐๐๐๐ก either [ 0, 0 ], [ ๐(๐ฅ, ๐ผ), ๐(๐ฅ, ๐ผ)]
The exact solution ๐ข(๐ฅ, ๐ผ)[ ๐ข(๐ฅ, ๐ผ), ๐ข(๐ฅ, ๐ผ)].
where
66
lim๐โโ
๐ข๐+1 = ๐ข(๐ฅ, ๐ผ) ๐๐๐ lim๐โโ
๐ข๐+1 = ๐ข(๐ฅ, ๐ผ)
Consider the fuzzy Volterra integral equation:
๐ข(๐ฅ) = (๐ผ + 1,3 โ ๐ผ) + โซ (๐ฅ โ ๐ก)
๐ฅ
0
๐ข(๐ก)๐๐ก, ๐ฅ โ [0, ๐], ๐ < โ
Solution:
In this example ๐(๐ฅ, ๐ก) โฅ 0, ๐๐๐ ๐๐๐โ 0 โค ๐ก โค ๐ฅ.
let ๐ข0(๐ฅ, ๐ผ) = (๐ผ + 1)
the first terms of successive approximation method series are:
๐ข1(๐ฅ, ๐ผ) = (๐ผ + 1) + โซ(๐ฅ โ ๐ก)๐ข0(๐ก, ๐ผ)๐๐ก
๐ฅ
0
= (๐ผ + 1) + โซ(๐ฅ โ ๐ก)(๐ผ + 1)๐๐ก
๐ฅ
0
=(๐ผ + 1) (1 +๐ฅ2
2)
๐ข2(๐ฅ, ๐ผ) = (๐ผ + 1) + โซ(๐ฅ โ ๐ก)๐ข1(๐ก, ๐ผ)๐๐ก
๐ฅ
0
= (๐ผ + 1) + โซ(๐ฅ โ ๐ก) (1 +๐ก2
2) (๐ผ + 1)๐๐ก
๐ฅ
0
=(๐ผ + 1) (1 +๐ฅ2
2+
๐ฅ4
24)
67
๐ข3(๐ฅ, ๐ผ) = (๐ผ + 1) + โซ(๐ฅ โ ๐ก)๐ข2(๐ก, ๐ผ)๐๐ก
๐ฅ
0
= (๐ผ + 1) + โซ(๐ฅ โ ๐ก) (1 +๐ก2
2+
๐ก4
24) (๐ผ + 1)๐๐ก
๐ฅ
0
=(๐ผ + 1) (1 +๐ฅ2
2+
๐ฅ4
24+
๐ฅ6
720)
and
๐ข0(๐ฅ, ๐ผ) = (3 โ ๐ผ)
๐ข1(๐ฅ, ๐ผ) = (3 โ ๐ผ) + โซ(๐ฅ โ ๐ก)๐ข0(๐ก, ๐ผ)๐๐ก
๐ฅ
0
= (3 โ ๐ผ) + โซ(๐ฅ โ ๐ก)(3 โ ๐ผ)๐๐ก
๐ฅ
0
=(3 โ ๐ผ)) (1 +๐ฅ2
2)
๐ข2(๐ฅ, ๐ผ) = (3 โ ๐ผ) + โซ(๐ฅ โ ๐ก)๐ข1(๐ก, ๐ผ)๐๐ก
๐ฅ
0
= (3 โ ๐ผ) + โซ(๐ฅ โ ๐ก) (1 +๐ก2
2) (3 โ ๐ผ)๐๐ก
๐ฅ
0
=(3 โ ๐ผ) (1 +๐ฅ2
2+
๐ฅ4
24)
๐ข3(๐ฅ, ๐ผ) = (3 โ ๐ผ) + โซ(๐ฅ โ ๐ก)๐ข2(๐ก, ๐ผ)๐๐ก
๐ฅ
0
= (3 โ ๐ผ) + โซ(๐ฅ โ ๐ก) (1 +๐ก2
2+
๐ก4
24) (3 โ ๐ผ)๐๐ก
๐ฅ
0
=(3 โ ๐ผ) (1 +๐ฅ2
2+
๐ฅ4
24+
๐ฅ6
720)
68
hence, the exact solution is:
๐ข(๐ฅ, ๐ผ) = (๐ผ + 1) lim๐โโ
(1 +๐ฅ2
2!+
๐ฅ4
4!+
๐ฅ6
6!+ โฏ)
= (๐ผ + 1) cosh(๐ฅ),
and
๐ข(๐ฅ, ๐ผ) = (3 โ ๐ผ) lim๐โโ
(1 +๐ฅ2
2!+
๐ฅ4
4!+
๐ฅ6
6!+ โฏ)
= (3 โ ๐ผ) cosh(๐ฅ)
69
Chapter Three
Numerical Methods for Solving Linear Fuzzy Volterra
Integral Equation of the Second Kind
Introduction
Several numerical methods for approximating linear Volterra integral
equation is known. In this chapter we introduce three numerical methods
to solve the Volterra fuzzy integral equation
๐ข(๐ฅ) = ๐(๐ฅ) + ๐ โซ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐ก
๐ฅ
๐
These methods are Taylor expansion method, trapezoidal method, and the
variational iteration method.
3.1 Taylor Expansion Method
if ๐(๐ฅ) is continuous and infinity differentiable function on some
open interval ๐ผ, and if ๐ฅ = ๐ง โ ๐ผ, then the taylor series of ๐(๐ฅ)
about ๐ฅ = ๐ง is given as:
๐(๐ฅ) = ๐(๐ง) +1
2!๐โฒ(๐ง)(๐ฅ โ ๐ง) + โฏ+
1
๐!๐(๐)(๐ง)(๐ฅ โ ๐ง)๐ + โฏ
Here we use the Taylor expansion method to solve linear fuzzy Volterra
integral equation of the second kind. This method depends on
differentiating the fuzzy integral equation of the second kind ๐ ๐ก๐๐๐๐ ,
then substitute the Taylor series expansion for the unknown function. as a
70
result, we get a linear system for which the solution of this system yields
the unknown Taylor coefficient of the solution functions.
Definition(๐. ๐)[๐๐]:
The second kind fuzzy Volterra integral equations system is in the form:
๐ข๐(๐ฅ) = ๐๐(๐ฅ) + โ(๐๐,๐ โซ๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก)๐๐ก
๐ฅ
๐
)
๐
๐=1
(3.1)
where ๐ โค ๐ก โค ๐ฅ โค ๐, ๐๐๐ ๐๐,๐ โ 0 ๐๐๐ ๐, ๐ = 1,2,โฆ ,๐ are real
constants. Moreover, in system (3.1) the fuzzy function ๐๐(๐ฅ) and ๐๐,๐(๐ฅ, ๐ก)
are given, and assumed to be sufficiently differentiable with respect to all
their arguments on the interval ๐ โค ๐ฅ โค ๐ก โค ๐. also we assume that the
kernel function ๐๐,๐(๐ฅ, ๐ก) โ ๐ฟ2([๐, ๐] ร [๐, ๐]), and ๐ข(๐ฅ) =
[๐ข1(๐ฅ),โฆ , ๐ข๐(๐ฅ)]๐is the solution to be determined.
Now, the parametric form of ๐๐(๐ฅ) ๐๐๐ ๐ข๐(๐ฅ) are
(๐๐(๐ฅ, ๐ผ), ๐๐(๐ฅ, ๐ผ)) ๐๐๐ (๐ข๐(๐ฅ, ๐ผ), ๐ข๐(๐ฅ, ๐ผ)), respectively. ( 0 โค ๐ผ โค 1,
๐ โค ๐ฅ โค ๐). To simplify we assume ๐๐,๐ > 0, (๐, ๐ = 1,2,โฆ ,๐).
In order to design a numerical scheme for solving (3.1), we write the
parametric form of the given fuzzy integral equations system as:
๐ข๐(๐ฅ, ๐ผ) = ๐๐(๐ฅ, ๐ผ) + โ(๐๐,๐ โซ๐ข๐,๐(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
)
๐
๐=1
๐ข๐(๐ฅ, ๐ผ) = ๐๐(๐ฅ, ๐ผ) + โ(๐๐,๐ โซ๐ข๐,๐(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
)
๐
๐=1
(3.2)
71
where
๐ข๐,๐(๐ก, ๐ผ) = {๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ) ๐๐,๐ โฅ 0
๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ) ๐๐,๐ < 0
(3.3)
๐ข๐,๐(๐ก, ๐ผ) = {๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ) ๐๐,๐ โฅ 0
๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ) ๐๐,๐ < 0
now, we assume that
{
๐๐,๐๐๐,๐(๐ฅ, ๐ก) โฅ 0, ๐ โค ๐ก โค ๐๐,๐
๐๐,๐๐๐,๐(๐ฅ, ๐ก) < 0, ๐๐,๐ โค ๐ก โค ๐ฅ (3.4)
then the system (3.2) becomes:
๐ข๐(๐ฅ, ๐ผ) = ๐๐(๐ฅ, ๐ผ) + โ๐๐,๐
(
โซ ๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐๐,๐
๐
+ โซ๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ฅ
๐๐,๐ )
๐
๐=1
๐ข๐(๐ฅ, ๐ผ) = ๐๐(๐ฅ, ๐ผ) + โ๐๐,๐
(
โซ ๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐๐,๐
๐
+ โซ๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ฅ
๐๐,๐ )
๐
๐=1
(3.5)
We seek the solution of system (3. 5) in the form of
๐ข๐,๐ (๐ก, ๐ผ) = โ (1
๐! ๐(๐)๐ข๐๐ (๐ฅ,๐ผ)
๐|๐ฅ=๐ง
. (๐ก โ ๐ง)๐)๐ ๐=0
๐ข๐,๐ (๐ก, ๐ผ) = โ (1
๐! ๐(๐)๐ข๐๐ (๐ฅ,๐ผ)
๐๐ฅ๐ |๐ฅ=๐ง
. (๐ก โ ๐ง)๐)๐ ๐=0
๐ โค ๐ฅ, ๐ง โค ๐ (3.6)
72
for ๐ = 1,โฆ ,๐ which are the Taylor expansion of degree ๐ at ๐ฅ = ๐ง
for the unknown functions ๐ข๐๐ (๐ฅ, ๐ผ), ๐ข๐๐ (๐ฅ, ๐ผ), respectively.
To obtain the solution in the form of expression (3.6) we find the ๐ ๐กโ
derivative of each equation in the system (3.5) with respect to ๐ฅ by using
the Leibnitz's rule, (๐ = 0,1,โฆ , ๐ ) and obtain by [ 46 ]:
๐(๐)๐ข๐๐ (๐ฅ, ๐ผ)
๐๐ฅ๐=
๐(๐)๐๐(๐ฅ, ๐ผ)
๐๐ฅ๐+ โ๐๐,๐ {โซ
๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐๐ข๐๐ (๐ก, ๐ผ)๐๐ก
๐๐,๐
๐
๐
๐=1
+ โ (๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐|๐ก=๐ฅ
. ๐ข๐๐ (๐ฅ, ๐ผ))
(๐โ๐โ1)๐โ1
๐=0
+ โซ๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐๐ข๐๐ (๐ก, ๐ผ)๐๐ก
๐ฅ
๐๐,๐
}
๐(๐)๐ข๐๐ (๐ฅ, ๐ผ)
๐๐ฅ๐=
๐(๐)๐๐(๐ฅ, ๐ผ)
๐๐ฅ๐+ โ๐๐,๐ {โซ
๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐๐ข๐๐ (๐ก, ๐ผ)๐๐ก
๐๐,๐
๐
๐
๐=1
+ โ (๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐|๐ก=๐ฅ
. ๐ข๐๐ (๐ฅ, ๐ผ))
(๐โ๐โ1)๐โ1
๐=0
+ โซ๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐๐ข๐๐ (๐ก, ๐ผ)๐๐ก
๐ฅ
๐๐,๐
}
(3.7)
for ๐ = 0,1,โฆ , ๐ , ๐๐๐ ๐ = 1,2,โฆ ,๐.
From the Leibniz's rule, the system (3.7) becomes:
73
๐(๐)๐ข๐๐ (๐ฅ, ๐ผ)
๐๐ฅ๐=
๐(๐)๐๐(๐ฅ, ๐ผ)
๐๐ฅ๐+ โ๐๐,๐ {โซ
๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐๐ข๐๐ (๐ก, ๐ผ)๐๐ก
๐๐,๐
๐
๐
๐=1
+ โ โ (๐ โ ๐ โ 1
๐)
๐โ๐โ1
๐=0
(๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐|๐ก=๐ฅ
)
(๐โ๐โ๐โ1)
(๐ข๐๐ (๐ฅ, ๐ผ))(๐)
๐โ1
๐=0
+ โซ๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐๐ข๐๐ (๐ก, ๐ผ)๐๐ก
๐ฅ
๐๐,๐
}
๐(๐)๐ข๐๐ (๐ฅ, ๐ผ)
๐๐ฅ๐=
๐(๐)๐๐(๐ฅ, ๐ผ)
๐๐ฅ๐+ โ๐๐,๐ {โซ
๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐๐ข๐๐ (๐ก, ๐ผ)๐๐ก
๐๐,๐
๐
๐
๐=1
+ โ โ (๐ โ ๐ โ 1
๐)
๐โ๐โ1
๐=0
(๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐|๐ก=๐ฅ
)
(๐โ๐โ๐โ1)
(๐ข๐๐ (๐ฅ, ๐ผ))(๐)
๐โ1
๐=0
+ โซ๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐๐ข๐๐ (๐ก, ๐ผ)๐๐ก
๐ฅ
๐๐,๐
] (3.8)
our purpose is determining the coefficients ๐ข๐๐ (๐)(๐ง, ๐ผ) and
๐ข๐๐ (๐)
(๐ง, ๐ผ), ๐๐๐ ๐ = 0,โฆ , ๐ , ๐๐๐ ๐ = 1,โฆ ,๐ in the system (3.7)
so that, we expand ๐ข๐๐ (๐ก, ๐ผ) and ๐ข๐๐ (๐ก, ๐ผ) in Taylor's series at arbitrary
point ๐ง: ๐ โค ๐ง โค ๐.
{
๐ข๐,๐ (๐ก, ๐ผ) = โ(
1
๐! ๐(๐)๐ข๐,๐ (๐ฅ, ๐ผ)
๐๐ฅ๐|๐ฅ=๐ง
. (๐ก โ ๐ง)๐)
๐
๐=0
๐ข๐,๐ (๐ก, ๐ผ) = โ(1
๐! ๐(๐)๐ข๐,๐ (๐ฅ, ๐ผ)
๐๐ฅ๐|๐ฅ=๐ง
. (๐ก โ ๐ง)๐)
๐
๐=0
, ๐ โค ๐ฅ, ๐ง โค ๐
0 โค ๐ผ โค 1 , for ๐ = 1,โฆ ,๐ (3.9)
74
where
๐ข๐,๐ (๐)(๐ง, ๐ผ) = ๐๐
(๐)(๐ง, ๐ผ) + โ โ ๐ท๐,๐(๐,๐)
๐โ1
๐=0
. ๐ข๐๐(๐)
(๐ง, ๐ผ) + โ๐ธ๐,๐(๐,๐)
๐
๐=0
. ๐ข๐๐(๐)(๐ง, ๐ผ)
๐
๐=1
+โ๐ธโฒ๐,๐(๐,๐)
๐
๐=0
. ๐ข๐๐(๐)
(๐ง, ๐ผ)
๐ข๐,๐ (๐)
(๐ง, ๐ผ) = ๐๐
(๐)(๐ง, ๐ผ) + โ โ ๐ท๐,๐
(๐,๐)
๐โ1
๐=0
. ๐ข๐๐(๐)(๐ง, ๐ผ) + โ๐ธ๐,๐
(๐,๐)
๐
๐=0
. ๐ข๐๐(๐)
(๐ง, ๐ผ)
๐
๐=1
+โ๐ธโฒ๐,๐(๐,๐)
๐
๐=0
. ๐ข๐๐(๐)(๐ง, ๐ผ) (3.10)
then substitute the ๐ ๐กโ truncation in (3.8), we get
๐ธ๐,๐(๐,๐)
=๐๐,๐
๐!โซ
๐(๐)๐๐,๐
๐๐ฅ๐
๐๐,๐
๐
(๐ฅ, ๐ก)|
๐ฅ=๐ง
. (๐ก โ ๐ง)๐๐๐ก
๐ธโฒ๐,๐(๐,๐)
=๐๐,๐
๐!โซ
๐(๐)๐๐,๐
๐๐ฅ๐
๐
๐๐,๐
(๐ฅ, ๐ก)|
๐ฅ=๐ง
. (๐ก โ ๐ง)๐๐๐ก
๐ท๐,๐(๐,๐)
= ๐๐,๐ โ(๐ โ ๐ โ 1
๐) (
๐(๐)๐๐,๐
๐๐ฅ๐(๐ฅ, ๐ก)|
๐ฅ=๐ง
)
(๐โ๐โ๐โ1)๐โ๐
๐=0
๐, ๐ = 1,2,โฆ ,๐
๐๐๐ ๐ = 0 ,โ โ ๐ท๐,๐(๐,๐)
๐โ1
๐=0
๐ข๐๐(๐)
(๐ง, ๐ผ) = 0
๐
๐=1
๐๐๐ ๐ โค ๐,๐ค๐ โ๐๐ฃ๐ ๐ท๐,๐(๐,๐)
= 0,๐ค๐ ๐ก๐๐๐ ๐, ๐ = 0,1โฆ ๐
75
Consequently, the equation (3.8) can be written in the matrix form:
(๐ท + ๐ธ)๐ = โฑ (3.11)
where
๐น = [
โ๐1(๐ง, ๐ผ),โฆ ,โ๐1(๐)(๐ฅ, ๐ผ)|
๐ฅ=๐ง, โ๐
1(๐ง, ๐ผ), โฆ ,โ๐
1
(๐)(๐ฅ, ๐ผ)|
๐ฅ=๐ง, โฆ ,
โ๐๐(๐ง, ๐ผ),โฆ , โ๐๐(๐)(๐ฅ, ๐ผ)|
๐ฅ=๐ง, โ๐
๐(๐ง, ๐ผ),โฆ , โ๐
๐
(๐)(๐ฅ, ๐ผ)|
๐ฅ=๐ง ]
๐ก
๐ข = [๐ข1๐(๐ง, ๐ผ),โฆ , ๐ข1๐
(๐)(๐ฅ, ๐ผ)|๐ฅ=๐ง
, ๐ข1๐(๐ง, ๐ผ),โฆ , ๐ข1๐(๐)
(๐ฅ, ๐ผ)|๐ฅ=๐ง
, โฆ ,
๐ข๐๐(๐ง, ๐ผ), โฆ , ๐ข๐๐(๐)(๐ฅ, ๐ผ)|
๐ฅ=๐ง, ๐ข๐๐(๐ง, ๐ผ), โฆ , ๐ข๐๐
(๐)(๐ฅ, ๐ผ)|
๐ฅ=๐ง ]
๐ก
D= ๐ท(1,1) โฏ ๐ท(1,๐)
โฎ โฑ โฎ๐ท(๐,1) โฏ ๐ท(๐,๐)
, E= ๐ธ(1,1) โฏ ๐ธ(1,๐)
โฎ โฑ โฎ๐ธ(๐,1) โฏ ๐ธ(๐,๐)
(3.12)
Parochial matrices ๐ท(๐,๐) ( ๐๐๐ ๐, ๐ = 1,โฆ๐) are defined with the
following elements:
๐ท(๐,๐) = ๐ท1,1
(๐,๐)๐ท1,2
(๐,๐)
๐ท2,1(๐,๐)
๐ท2,2(๐,๐)
, ๐ธ(๐,๐) = ๐ธ1,1
(๐,๐)๐ธ1,2
(๐,๐)
๐ธ2,1(๐,๐)
๐ธ2,2(๐,๐)
, ๐, ๐ = 1,โฆ ,๐
(3.13)
where
๐ธ1,1(๐,๐)
= ๐ธ2,2(๐,๐)
=
[ ๐0,0
(๐,๐)โ 1 ๐0,1
(๐,๐)โฏ ๐0,๐ โ1
(๐,๐)๐0,๐
(๐,๐)
๐1,0(๐,๐)
๐1,1(๐,๐)
โ 1 โฏ ๐1,๐ โ1(๐,๐)
๐1,๐ (๐,๐)
โฎ โฎ โฑ โฎ โฎ
๐๐ โ1,1(๐,๐)
๐๐ โ1,1(๐,๐)
โฏ ๐๐ โ1,๐ โ1(๐,๐)
โ 1 ๐๐ โ1,๐ (๐,๐)
๐๐ ,0(๐,๐)
๐๐ ,1(๐,๐)
โฏ ๐๐ ,๐ โ1(๐,๐)
๐๐ ,๐ (๐,๐)
โ 1]
,
76
๐ธ1,2(๐,๐)
= ๐ธ2,1(๐,๐)
=
[ ๐โฒ0,0
(๐,๐)โ 1 ๐โฒ0,1
(๐,๐)โฏ ๐โฒ0,๐ โ1
(๐,๐)๐โฒ0,๐
(๐,๐)
๐โฒ1,0(๐,๐)
๐โฒ1,1(๐,๐)
โ 1 โฏ ๐โฒ1,๐ โ1(๐,๐)
๐โฒ1,๐ (๐,๐)
โฎ โฎ โฑ โฎ โฎ
๐โฒ๐ โ1,0(๐,๐)
๐โฒ๐ โ1,1(๐,๐)
โฏ ๐โฒ๐ โ1,๐ โ1(๐,๐)
โ 1 ๐โฒ๐ โ1,๐ (๐,๐)
๐โฒ๐ ,0(๐,๐)
๐โฒ๐ ,1(๐,๐)
โฏ ๐โฒ๐ ,๐ โ1(๐,๐)
๐โฒ๐ ,๐ (๐,๐)
โ 1]
,
(3.14)
๐ท1,2(๐,๐)
= ๐ท2,1(๐,๐)
=
[
0 0 โฏ 0 0
๐โฒ1,0(๐,๐)
0 โฏ 0 0
โฎ โฎ โฑ โฎ โฎ
๐โฒ๐ โ1,0(๐,๐)
๐โฒ๐ โ1,1(๐,๐)
โฏ 0 0
๐โฒ๐ ,0(๐,๐)
๐โฒ๐ ,1(๐,๐)
โฏ ๐โฒ๐ ,๐ โ1(๐,๐)
0]
๐ท1,1(๐,๐)
= ๐ท2,2(๐,๐)
=
[ 0 0 โฏ 0 00 0 โฏ 0 0โฎ โฎ โฑ โฎ โฎ0 0 โฏ 0 00 0 โฏ 0 0]
(๐ +1)ร(๐ +1)
3.2 Convergence Analysis
In this section, we will prove that the approximation solution from Taylor
expansion method converges to the exact solution of fuzzy system (3.2).
Theorem (๐. ๐) [๐๐]: Let the kernel be bounded and belongs to ๐ฟ2 and
๐ข๐,๐ (๐ฅ, ๐ผ)๐๐๐ ๐ข๐,๐
(๐ฅ, ๐ผ) ( ๐๐๐ ๐ = 1,โฆ ,๐) be Taylor polynomials of
degree ๐ , and their coefficients are computed by solving the linear system
(3.12), then these polynomials converges to the exact solution of fuzzy
system (3.2), ๐คโ๐๐ ๐ โ +โ.
77
Proof:
Consider the fuzzy system (3.2). Since the series (3.10) converges to
๐ข๐(๐ฅ, ๐ผ)๐๐๐ ๐ข๐(๐ฅ, ๐ผ)( ๐๐๐ ๐ = 1,โฆ ,๐), respectively, ๐. ๐
๐ข๐(๐ฅ, ๐ผ) = lim๐ โโ
๐ข๐๐ (๐ฅ, ๐ผ) , ๐๐๐ ๐ข๐(๐ฅ, ๐ผ) = lim๐ โโ
๐ข๐๐ (๐ฅ, ๐ผ)
then, we conclude that
๐ข๐๐ (๐ฅ, ๐ผ) = ๐๐(๐ฅ, ๐ผ) + โ๐๐,๐
(
โซ๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐
๐
+ โซ๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ฅ
๐๐,๐ )
๐
๐=1
,
๐ข๐๐ (๐ฅ, ๐ผ) = ๐๐(๐ฅ, ๐ผ) + โ๐๐,๐
(
โซ ๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐๐,๐
๐
+ โซ๐๐,๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ฅ
๐๐,๐ )
๐
๐=1
๐, ๐ = 1,โฆ ,๐
(3.15)
now, we define the error function ๐๐ (๐ฅ, ๐ผ)as a difference between the two
systems (3.5) ๐๐๐ (3.15), we obtain:
๐๐ (๐ฅ, ๐ผ) = โ๐๐,๐ (๐ฅ, ๐ผ),
๐
๐=1
(3.16)
where, ๐๐,๐ (๐ฅ, ๐ผ) = ๐๐,๐ (๐ฅ, ๐ผ) + ๐๐,๐ (๐ฅ, ๐ผ),
78
๐๐,๐ (๐ฅ, ๐ผ) = (๐๐(๐ฅ, ๐ผ) โ ๐๐,๐ (๐ฅ, ๐ผ))
+โ๐๐,๐ (โซ ๐๐,๐(๐ฅ, ๐ก) (๐๐(๐ก, ๐ผ)๐๐ก โ ๐๐,๐ (๐ก, ๐ผ)) ๐๐ก
๐๐,๐
๐
)
๐
๐=1
+โ๐๐,๐
๐
๐=1
( โซ๐๐,๐(๐ฅ, ๐ก) (๐๐(๐ก, ๐ผ) โ ๐๐,๐ (๐ก, ๐ผ))๐๐ก
๐ฅ
๐๐,๐
)
(3.17)
we must show when ๐ โ +โ, ๐กโ๐๐ ๐๐ (๐ฅ, ๐ผ) โ 0
taking the โ โof system (3.17 ) we proceed as:
โ๐๐ โ = โโ๐๐๐
๐
๐=1
โ โค โโ๐๐๐ โ
๐
๐=1
= โโ๐๐๐ + ๐๐๐ โ โค โ(โ๐๐๐ โ + โ๐๐๐ โ)
๐
๐=1
๐
๐=1
= โ{(๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ))
๐
๐=1
+ โ๐๐,๐ (โซ ๐๐,๐(๐ฅ, ๐ก) (๐ข๐(๐ก, ๐ผ)๐๐ก โ ๐ข๐,๐ (๐ก, ๐ผ)) ๐๐ก
๐๐,๐
๐
)
๐
๐=1
+ โ๐๐,๐
๐
๐=1
( โซ๐๐,๐(๐ฅ, ๐ก) (๐ข๐(๐ก, ๐ผ) โ ๐ข๐,๐ (๐ก, ๐ผ)) ๐๐ก
๐ฅ
๐๐,๐
)}
+โ(๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ))
+ โ๐๐,๐ (โซ ๐๐,๐(๐ฅ, ๐ก) (๐ข๐(๐ก, ๐ผ) โ ๐ข๐,๐ (๐ก, ๐ผ))๐๐ก
๐๐,๐
๐
)
๐
๐=1
+ โ๐๐,๐
๐
๐=1
( โซ๐๐,๐(๐ฅ, ๐ก) (๐ข๐(๐ก, ๐ผ) โ ๐ข๐,๐ (๐ก, ๐ผ)) ๐๐ก
๐ฅ
๐๐,๐
)โ
79
โค โโ(๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ)) + (๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ))โ
๐
๐=1
+โโโ๐๐,๐ โซ ๐๐,๐(๐ฅ, ๐ก)(๐ข๐(๐ก, ๐ผ)๐๐ก โ ๐ข๐,๐ (๐ก, ๐ผ) + ๐ข๐(๐ก, ๐ผ)๐๐ก โ ๐ข๐,๐ (๐ก, ๐ผ))๐๐ก
๐๐,๐
๐
๐
๐=1
๐
๐=1
+ โ๐๐,๐ โซ๐๐,๐(๐ฅ, ๐ก)(๐ข๐(๐ก, ๐ผ) โ ๐ข๐,๐ (๐ก, ๐ผ) + ๐ข๐(๐ก, ๐ผ) โ ๐ข๐,๐ (๐ก, ๐ผ))๐๐ก
๐ฅ
๐๐,๐
๐
๐=1
โ
โค โโ(๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ)) + (๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ))โ
๐
๐=1
+ โโ ๐๐,๐
๐
๐=1
๐
๐=1
โโซ๐๐,๐(๐ฅ, ๐ก) ( ๐ข๐(๐ก, ๐ผ) โ ๐ข๐,๐ (๐ก, ๐ผ)
๐ฅ
๐
+ ๐ข๐(๐ก, ๐ผ)๐๐ก โ ๐ข๐,๐ (๐ก, ๐ผ))๐๐กโ
โค โโ(๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ)) + (๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ))โ
๐
๐=1
+ โ{โ ๐๐,๐
๐
๐=1
โซโ๐๐,๐(๐ฅ, ๐ก)โ( โ๐ข๐(๐ก, ๐ผ) โ ๐ข๐,๐ (๐ก, ๐ผ)โ +
๐ฅ
๐
โ๐ข๐(๐ก, ๐ผ)
๐
๐=1
โ ๐ข๐,๐ (๐ก, ๐ผ)โ )๐๐ก}
since โ๐๐,๐(๐ฅ, ๐ก)โ is continuous on [๐, ๐] ๐ โค ๐ฅ โค ๐, therefore,
โ๐๐,๐(๐ฅ, ๐ก)โ is bounded, this implies that โ๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ)โ โ 0,
and, โ๐ข๐(๐ฅ, ๐ผ) โ ๐ข๐,๐ (๐ฅ, ๐ผ)โ โ 0 ๐๐ ๐ โ +โ, โ๐๐๐๐ โ๐๐ โ โ 0.
80
3.3 Trapezoidal Method
We consider fuzzy Volterra integral equation of the second kind (1.18) and
subdivide the interval of integration [๐, ๐ฅ] into ๐ equal subintervals
[๐ฅ๐โ1, ๐ฅ๐] of width ๐ฟ =๐โ๐
๐ , ๐ โฅ 1, where b is the end point we choose
for ๐ฅ. Now let ๐ฅ๐ = ๐ + ๐๐ฟ = ๐ก๐. 0 โค ๐ โค ๐
the Trapezoidal rule is:
โซ๐(๐ฅ)๐๐ฅ = ๐ฟ ๐(๐) + ๐(๐)
2+ โ๐(๐ฅ๐)
๐ โ1
๐=1
๐
๐
using the trapezoidal rule with ๐ subintervals, we approximate the fuzzy
Volterra integral equation by [20]:
โซ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐ก โ ๐ฟ ๐(๐ฅ, ๐ก0)๐ข(๐ก0) + ๐(๐ฅ, ๐ก๐)๐ข(๐ก๐)
2+ โ๐(๐ฅ, ๐ก๐)๐ข(๐ก๐)
๐โ1
๐=1
๐ฅ
๐
(3.18)
๐ฟ =๐ก๐ โ ๐
๐=
๐ฅ๐ โ ๐
๐ , ๐ก๐ โค ๐ฅ, ๐ โฅ 1, ๐ = ๐ฅ๐ = ๐ก๐ , ๐ = 0,1,โฆ , ๐
let us define:
{
๐บ๐ (๐ฅ, ๐ผ) = ๐ฟ
๐(๐ฅ, ๐)๐ข(๐) + ๐(๐ฅ, ๐)๐ข(๐)
2+ โ ๐(๐ฅ, ๐ก๐)๐ข(๐ก๐)
๐โ1
๐=1
๐บ๐ (๐ฅ, ๐ผ) = ๐ฟ ๐(๐ฅ, ๐)๐ข(๐) + ๐(๐ฅ, ๐)๐ข(๐)
2+ โ ๐(๐ฅ, ๐ก๐)๐ข(๐ก๐)
๐โ1
๐=1
(3.19)
81
hence, the fuzzy Volterra integral equation (1.18) is approximated by:
๐ข(๐ฅ, ๐ผ) = { ๐(๐ฅ, ๐ผ) + ๐บ๐ (๐ผ)
๐(๐ฅ, ๐ผ) + ๐บ๐ (๐ผ) (3.20)
we substitute ๐ฅ = ๐ฅ๐ in the equation (3.20), we get:
๐ข(๐ฅ๐ , ๐ผ) = ๐(๐ฅ๐ , ๐ผ) + ๐ฟ [
๐(๐ฅ๐,๐)๐ข(๐)+๐(๐ฅ๐,๐ฅ๐)๐ข(๐ฅ๐)
2+ โ ๐(๐ฅ๐ , ๐ฅ๐)๐ข(๐ฅ๐)๐โ1
๐=1 ]
๐(๐ฅ๐ , ๐ผ) + ๐ฟ ๐(๐ฅ๐,๐)๐ข(๐)+๐(๐ฅ๐,๐ฅ๐)๐ข(๐ฅ๐)
2+ โ ๐(๐ฅ๐ , ๐ฅ๐)๐ข(๐ฅ๐)๐โ1
๐=1
(3.21)
now, we assume that
{๐(๐ฅ๐ , ๐ฅ๐) โฅ 0, 0 โค ๐ โค ๐
๐(๐ฅ๐ , ๐ฅ๐) โค 0, ๐ + 1 โค ๐ โค ๐ (3.22)
then, the equation (3.21) becomes:
๐ข(๐ฅ๐ , ๐ผ) = ๐(๐ฅ๐ , ๐ผ) +๐ฟ
2๐(๐ฅ๐ , ๐)๐ข(๐, ๐ผ) + +๐ฟ โ ๐(๐ฅ๐ , ๐ฅ๐)๐ข(๐ฅ๐
๐
๐=1
)
+ ๐ฟ โ ๐(๐ฅ๐ , ๐ฅ๐)๐ข(๐ฅ๐
๐
๐=๐+1
) +๐ฟ
2๐(๐ฅ๐ , ๐ฅ๐)๐ข(๐ฅ๐, ๐ฅ๐),
and (3.23)
๐ข(๐ฅ๐ , ๐ผ) = ๐(๐ฅ๐ , ๐ผ) +๐ฟ
2๐(๐ฅ๐ , ๐)๐ข(๐, ๐ผ) + +๐ฟ โ ๐(๐ฅ๐ , ๐ฅ๐)๐ข(๐ฅ๐
๐
๐=1
)
+ ๐ฟ โ ๐(๐ฅ๐ , ๐ฅ๐)๐ข(๐ฅ๐
๐
๐=๐+1
) +๐ฟ
2๐(๐ฅ๐ , ๐ฅ๐)๐ข(๐ฅ๐ , ๐ฅ๐)
82
๐ = 0,1,2, โฆ , ๐
from the equation (3.234) , the following system was obtained:
(๐ + ๐) ๐ = ๐น (3.24)
where
๐ข = [๐ข(๐ฅ0, ๐ผ), ๐ข(๐ฅ1, ๐ผ), โฆ , ๐ข(๐ฅ๐ , ๐ผ), ๐ข(๐ฅ0, ๐ผ), ๐ข(๐ฅ1, ๐ผ),โฆ , ๐ข(๐ฅ๐ , ๐ผ)]๐ก
๐น = [๐(๐ฅ0, ๐ผ), ๐(๐ฅ1, ๐ผ),โฆ , ๐(๐ฅ๐ , ๐ผ), ๐(๐ฅ0, ๐ผ), ๐(๐ฅ1, ๐ผ), โฆ , ๐(๐ฅ๐ , ๐ผ)]๐ก
๐ = ๐1,1 ๐1,2
๐2,1 ๐2,2
where ๐1,1 = (๐ค)๐,๐ ,0 โค ๐, ๐ โค ๐ , 1โค ๐ โค ๐ , with:
(๐ค)๐,๐ =
{
โ๐ฟ ๐(๐ฅ๐ , ๐ฅ๐) ,1 โค ๐ โค ๐ โค ๐
1 ,0 โค ๐ = ๐ โค ๐ โโ๐ฟ
2 ๐(๐ฅ๐ , ๐ฅ๐) ,1 โค ๐ โค ๐, ๐ = 0
0 , ๐๐๐ ๐ ๐คโ๐๐๐
๐1,1 = ๐2,2 , ๐1,2 = ๐2,1 = ๐ง๐๐๐ ๐๐๐ก๐๐๐ฅ .
๐ = ๐1,1 ๐1,2
๐2,1 ๐2,2
where ๐1,1 = (๐ฃ)๐,๐ ,0 โค ๐, ๐ โค ๐ , 1โค ๐ โค ๐ ,with:
83
(๐ฃ)๐,๐ =
{
โ๐ฟ ๐(๐ฅ๐ , ๐ฅ๐) , ๐ + 1 โค ๐ โค ๐ โค ๐
โ๐ฟ
2 ๐(๐ฅ๐ , ๐ฅ๐) ,1 โค ๐ = ๐ โค ๐
โ๐ฟ
2 ๐(๐ฅ๐ , ๐ฅ๐) , ๐ + 1 โค ๐ โค ๐ , ๐ = 0
0 , ๐๐๐ ๐ ๐คโ๐๐๐
๐1,1 = ๐2,2 , ๐1,2 = ๐2,1 = ๐ง๐๐๐ ๐๐๐ก๐๐๐ฅ .
for arbitrary fixed ๐ผ we have
๐ข(๐ฅ, ๐ผ) = lim๐โโ
๐บ๐(๐ผ) = โซ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐ก
๐ฅ
๐
,
and (3.25)
๐ข(๐ฅ, ๐ผ) = lim๐โโ
๐บ๐(๐ผ) = โซ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐ก
๐ฅ
๐
.
Theorem (๐. ๐) [๐๐]:
If ๐(๐ฅ, ๐ผ) is continuous in the metric ๐ท, then ๐บ๐(๐ผ) ๐๐๐ ๐บ๐(๐ผ)
converges uniformly in ๐ผ ๐ก๐ ๐ข(๐ฅ, ๐ผ) ๐๐๐ ๐ข(๐ฅ, ๐ผ), respectively.
proof: see [46]
3.4 Variation Iteration Method
In this section, we will use the variation iteration method (VIM) to
approximate the solution of Volterra integral equations of the second kind.
The method constructs a convergent sequence of functions, which
approximates the exact solution with some iteration.
84
Consider the following general nonlinear system by [29]:
๐ฟ[๐ข(๐ฅ)] + ๐[๐ข(๐ก)] = โ(๐ฅ), (3.26)
where ๐ฟ is a linear operator, ๐ is a nonlinear operator and โ(๐ฅ) is a given
continuous function. from the Variational iteration method, we can
construct the following correction functional:
๐ข๐+1(๐ฅ) = ๐ข๐(๐ฅ) + โซ๐(๐ง)[๐ฟ[๐ข๐(๐ง)] + ๐[๏ฟฝฬ๏ฟฝ๐(๐ง)] โ โ(๐ง)]
๐ฅ
0
๐๐ง
(3.27)
where ๐ = 0,1,3, โฆ, and ๐ is a langrage multiplier which can identified
optimally via variational theory, ๐ข๐ is the ๐ ๐กโ approximate solution,
and ๏ฟฝฬ๏ฟฝ๐ is a restricted variation (๐. ๐, ๐ฟ๏ฟฝฬ๏ฟฝ๐ = 0).
now, let us differentiate both sides of the Volterra integral equation of the
second kind with respect to ๐ฅ , we get:
๐ขโฒ(๐ฅ) = ๐โฒ(๐ฅ) +๐
๐๐ฅโซ ๐(๐ฅ, ๐ก)๐ข(๐ก)๐๐ก
๐ฅ
๐ (3.28)
We use the variation iteration method in direction ๐ฅThen we get the
following iteration sequence:
๐ข๐+1(๐ฅ) =
๐ข๐(๐ฅ) + โซ๐(๐ง) ๐ข๐โฒ (๐ง) โ ๐โฒ(๐ง) โ
๐
๐๐งโซ๐(๐ง, ๐ก)๐ข๐(๐ก)๐๐ก
๐ง
๐
๐ฅ
0
๐๐ง
(3.29)
85
then we calculate variation with respect to ๐ข๐, notice that ๐ฟ๐ข๐(0)=0,
yields:
๐ฟ๐ข๐+1 = ๐ฟ๐ข๐ + ๐๐ฟ๐ข๐ ๐ง=๐ฅโ โซ ๐โฒ(๐ง)๐ฟ๐ข๐๐๐ง = 0
๐ฅ
0 (3.30)
therefore, we get the following stationary conditions:
๐โฒ(๐ง)|๐ง=๐ฅ = 0, 1 + ๐(๐ง)|๐ง=๐ฅ = 0
hence, the lagrange multiplier, can be identified:
๐(๐ง) = โ1,
If we substitute the value of the the Lagrange multiplier into the equation
(3.29), we obtain the following iteration formula:
๐ข๐+1(๐ฅ) =
๐ข๐(๐ฅ) โ โซ ๐ข๐โฒ (๐ง) โ ๐โฒ(๐ง) โ
๐
๐๐งโซ๐(๐ง, ๐ก)๐ข๐(๐ก)๐๐ก
๐ง
๐
๐ฅ
0
๐๐ง
(3.31)
we apply variational iteration method on fuzzy Volterra integral equation
of the second kind:
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐ โซ ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐ โซ ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก๐ฅ
๐
(3.32)
we suppose that the kernel ๐(๐ฅ, ๐ก) > 0 ๐๐๐ ๐ โค ๐ก โค ๐ , ๐๐๐
86
๐(๐ฅ, ๐ก) < 0 ๐๐๐ ๐ โค ๐ก โค ๐ฅ, ๐กโ๐๐ ๐กโ๐ ๐ ๐ฆ๐ ๐ก๐๐ (3.32) becomes by [1]:
{
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก + ๐โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
๐
๐
๐ข(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + ๐โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐
๐
+ ๐โซ๐(๐ฅ, ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
๐
(3.33)
for each 0 โค ๐ผ โค 1 ๐๐๐ ๐ โค ๐ฅ โค ๐,
now, using variation iteration method and equation (3.31), we get the
following iteration formulas:
๐ข๐+1(๐ฅ, ๐ผ) = ๐ข๐(๐ฅ, ๐ผ) โ
โซ[๐ข๐โฒ (๐ง, ๐ผ) โ ๐โฒ(๐ง, ๐ผ) โ โซ
๐๐(๐ง, ๐ก)
๐๐ง๐ข๐(๐ก, ๐ผ)๐๐ก โ
๐ถ
๐
๐(๐ง, ๐ง)๐ข๐(๐ง, ๐ผ)
๐ฅ
0
โ โซ๐๐(๐ง, ๐ก)
๐๐ง(๐ก, ๐ผ)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ง
๐
] ๐๐ง
(3.34)
and
๐ข๐+1(๐ฅ, ๐ผ) = ๐ข๐(๐ฅ, ๐ผ) โ
โซ [๐ข๐โฒ(๐ง, ๐ผ) โ ๐
โฒ(๐ง, ๐ผ) โ โซ
๐๐(๐ง, ๐ก)
๐๐ง๐ข๐(๐ก, ๐ผ)๐๐ก โ
๐ถ
๐
๐(๐ง, ๐ง)๐ข(๐ง, ๐ผ)
๐ฅ
0
โ โซ๐๐(๐ง, ๐ก)
๐๐ง(๐ก, ๐ผ)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ง
๐
] ๐๐ง
87
where ๐ = 0,1,2, โฆ
Using the formulas in (3.34), we can find a solution of equation (3.33) and
hence obtain a fuzzy solution of the linear fuzzy Volterra integral equation
of the second kind.
88
Chapter Four
Numerical Examples and Results
In this chapter, we will use algorithms and MAPLE software to solve
numerical examples, then draw a comparison between approximate
solutions and exact ones, to demonstrate the efficiency of these
numerical schemes in chapter three. These include: Taylor expansion
method, Trapezoidal method, and variation iteration method.
Numerical example (๐. ๐): (๐ป๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐ )
Consider the linear fuzzy Volterra integral equations:
{
๐ข(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2 + โซ(๐ฅ โ ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
0
๐ข(๐ฅ, ๐ผ) = (3 โ ๐ผ)๐ฅ2 + โซ(๐ฅ โ ๐ก)๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
0
(4.1)
0 โค ๐ฅ โค 1, 0 โค ๐ผ โค 1
the analytical solution of the above problem is given by:
{๐ข(๐ฅ, ๐ผ) = (๐ผ + 1) (๐๐ฅ + ๐โ๐ฅ โ 2)
๐ข(๐ฅ, ๐ผ) = (3 โ ๐ผ)(๐๐ฅ + ๐โ๐ฅ โ 2) 0 โค ๐ผ โค 1 (4.2)
we expand the unknown functions in the Taylor series at ๐ง =1
4
to solve the previous example by using Taylor expansion method using
MAPLE software, we will apply the following algorithm:
89
Algorithm (๐. ๐)
1. Input ๐, ๐, ๐ง,๐, ๐ , ๐๐,๐(๐ฅ, ๐ก), ๐๐(๐ฅ, ๐ผ), ๐๐(๐ฅ, ๐ผ)
2. Input the Taylor expansion degree ( ๐ )
3. Calculate ๐(๐)๐๐,๐(๐ฅ, ๐ผ)
๐๐ฅ๐,๐(๐)๐๐(๐ฅ, ๐ผ)
๐๐ฅ๐,๐(๐)๐
๐(๐ฅ, ๐ผ)
๐๐ฅ๐ , ๐ = 0,1,โฆ , ๐
4. Calculate
๐๐+1,๐+1(๐,๐)
=๐๐,๐
๐! โซ
๐๐๐๐,๐
๐๐ฅ๐(๐ฅ, ๐ก).
๐๐,๐
๐
(๐ก โ ๐ง)๐ ๐๐ก ๐, ๐ = 1,โฆ ,๐
5. Calculate
๐โฒ๐+1,๐+1(๐,๐)
=๐๐,๐
๐! โซ
๐๐๐๐,๐
๐๐ฅ๐(๐ฅ, ๐ก).
๐๐,๐
๐
(๐ก โ ๐ง)๐ ๐๐ก ๐, ๐ = 1,โฆ ,๐
6. Put ๐ธ1,1
(๐,๐)= ๐ธ2,2
(๐,๐)
=
[ ๐1,1
(๐,๐)โ 1 ๐1,2
(๐,๐)โฏ ๐1,๐
(๐,๐)๐1,๐ +1
(๐,๐)
๐2,1(๐,๐)
๐2,2(๐,๐)
โ 1 โฏ ๐2,๐ (๐,๐)
๐2,๐ +1(๐,๐)
โฎ โฎ โฑ โฎ โฎ
๐๐ ,1(๐,๐)
๐๐ ,2(๐,๐)
โฏ ๐๐ ,๐ (๐,๐)
โ 1 ๐๐ ,๐ +1(๐,๐)
๐๐ +1,1(๐,๐)
๐๐ +1,2(๐,๐)
โฏ ๐๐ +1,๐ (๐,๐)
๐๐ +1,๐ +1(๐,๐)
โ 1]
(๐ +1)ร(๐ +1)
7. ๐ธโฒ
1,2(๐,๐)
= ๐ธโฒ2,1(๐,๐)
=
[ ๐
โฒ1,1(๐,๐)
โ 1 ๐โฒ1,2(๐,๐)
โฏ ๐โฒ1,๐ (๐,๐)
๐โฒ1,๐ +1(๐,๐)
๐โฒ2,1(๐,๐)
๐โฒ2,2(๐,๐)
โ 1 โฏ ๐โฒ2,๐ (๐,๐)
๐โฒ2,๐ +1(๐,๐)
โฎ โฎ โฑ โฎ โฎ
๐โฒ๐ ,1(๐,๐)
๐โฒ๐ ,2(๐,๐)
โฏ ๐โฒ๐ ,๐ (๐,๐)
โ 1 ๐โฒ๐ ,๐ +1(๐,๐)
๐โฒ๐ +1,1(๐,๐)
๐โฒ๐ +1,2(๐,๐)
โฏ ๐โฒ๐ +1,๐ (๐,๐)
๐โฒ๐ +1,๐ +1(๐,๐)
โ 1]
90
where ๐, ๐ = 1,2,โฆ ,๐
8. Calculate
๐๐+1,๐+1โฒ = โ(
๐ โ ๐ โ 1
๐ โ ๐ โ ๐ โ 1)(
๐(๐)๐๐,๐(๐ฅ, ๐ก)
๐๐ฅ๐|๐ก=๐ฅ
)
(๐โ๐โ๐โ1)
|
๐ฅ=๐ง
๐โ๐โ1
๐=0
9. Put
๐ท1,1(๐,๐)
= ๐ท2,2(๐,๐)
=
[ 0 0 โฏ 0 00 0 โฏ 0 0โฎ โฎ โฑ โฎ โฎ0 0 โฏ 0 00 0 โฏ 0 0]
(๐ +1)ร(๐ +1)
10. Put
๐ทโฒ1,2(๐,๐)
= ๐ทโฒ2,1(๐,๐)
=
[
0 0 โฏ 0 0
๐โฒ2,1(๐,๐)
0 โฏ 0 0
โฎ โฎ โฑ โฎ โฎ
๐โฒ๐,1(๐,๐)
๐โฒ๐,2(๐,๐)
โฏ 0 0
๐โฒ๐+1,1(๐,๐)
๐โฒ๐+1,2(๐,๐)
โฏ ๐โฒ๐+1,๐(๐,๐)
0]
(๐ +1)ร(๐ +1)
11. denote ๐ท(๐,๐) = ๐ท1,1
(๐,๐)๐ท1,2
(๐,๐)
๐ท2,1(๐,๐)
๐ท2,2(๐,๐)
,
12. put
โฑ = [โ๐1(๐ง, ๐ผ), โฆ ,โ๐1
(๐)(๐ง, ๐ผ),โ๐1(๐ง, ๐ผ),โฆ ,โ๐
1
(๐ )(๐ง, ๐ผ),โฆ ,
โ๐๐(๐ง, ๐ผ),โฆ , โ๐๐(๐ )(๐ง, ๐ผ), โ๐
๐(๐ง, ๐ผ), โฆ ,โ๐
๐
(๐ )(๐ง, ๐ผ)
]
๐ก
91
13. put
๐ข = ๐ข1๐ (๐ง, ๐ผ),โฆ , ๐ข1๐
(๐ )(๐ง, ๐ผ), ๐ข1๐ (๐ง, ๐ผ), โฆ , ๐ข1๐ (๐ )
(๐ง, ๐ผ), โฆ ,
๐ข๐๐ (๐ง, ๐ผ),โฆ , ๐ข๐๐ (๐ )(๐ง, ๐ผ), ๐ข๐๐ (๐ง, ๐ผ),โฆ , ๐ข๐๐
(๐ )(๐ง, ๐ผ)
๐ก
14. solve the following linear system (๐ธ + ๐ท)๐ข = โฑ
15. estimate ๐ข(๐ง, ๐ผ), ๐ข(๐ง, ๐ผ) by computing Taylor expansion for ๐ข
{
๐ข๐,๐ = โ(
1
๐! ๐(๐)๐ข๐(๐ฅ, ๐ผ)
๐๐ฅ๐|๐ฅ=๐ง
(๐ฅ โ ๐ง)๐)
๐
๐=0
๐ข๐,๐ = โ(1
๐! ๐(๐)๐ข๐(๐ฅ, ๐ผ)
๐๐ฅ๐|๐ฅ=๐ง
(๐ฅ โ ๐ง)๐)
๐
๐=0
, ๐ โค ๐ฅ, ๐ง โค ๐,
0 โค ๐ผ โค 1, ๐ = 1,2,โฆ ,๐
We get the following results
๐ธ1,1(1,1)
= ๐ธ2,2(1,1)
= [
โ0.96875 โ0.00520833 0.00048828 โ0.000032550.25000 โ1.03125000 0.00260416 โ0.00016276
0 0 โ1 00 0 0 โ1
]
๐ธ1,2(1,1)
= ๐ธ2,1(1,1)
= [
0 0 0 00 0 0 00 0 0 00 0 0 0
]
โ๐๐๐๐,
๐ธ = ๐ธ1,1
(1,1)๐ธ1,2
(1,1)
๐ธ2,1(1,1)
๐ธ2,2(1,1)
๐ท1,1(1,1)
= ๐ท2,2(1,1)
= [
0 0 0 00 0 0 01 0 0 00 1 0 0
]
๐ท1,2(1,1)
= ๐ท2,1(1,1)
= [
0 0 0 00 0 0 00 0 0 00 0 0 0
]
92
๐ท = ๐ท1,1
(1,1)๐ท1,2
(1,1)
๐ท2,1(1,1)
๐ท2,2(1,1)
=
[ 0 0 0 0 0 0 0 00 0 0 0 0 0 0 01 0 0 0 0 0 0 00 1 0 0 0 0 0 00 0 0 0 0 0 0 00 0 0 0 0 0 0 00 0 0 0 1 0 0 00 0 0 0 0 1 0 0]
โฑ(๐ง, ๐ผ) = [
โ๐(๐ง, ๐ผ),โ๐โฒ(๐ง, ๐ผ),โ๐โฒโฒ(๐ง, ๐ผ),โ๐โฒโฒโฒ(๐ง, ๐ผ),
โ๐(๐ง, ๐ผ),โ๐โฒ(๐ง, ๐ผ),โ๐
โฒโฒ(๐ง, ๐ผ), โ๐
โฒโฒโฒ(๐ง, ๐ผ)
]
๐ก
โฑ(๐ง, ๐ผ) =
โ1
16(๐ผ + 1),
โ1
2(๐ผ + 1),โ2(๐ผ + 1), 0,
โ1
16(3 โ ๐ผ),
โ1
2(3 โ ๐ผ),โ2(3 โ ๐ผ),0
๐ก
Solving the linear system
(๐ท + ๐ธ)๐ข = โฑ
we get the following:
๐ข(๐ง, ๐ผ) =
[ 0.0628227120219095(๐ผ + 1)
0.505207645425615(๐ผ + 1)
2.06282271202191(๐ผ + 1)
0.505207645425615(๐ผ + 1)
0.0628227120219095(3 โ ๐ผ)
0.505207645425615(3 โ ๐ผ)
2.06282271202191(3 โ ๐ผ)
0.505207645425615(3 โ ๐ผ) ]
{
๐ข(๐ฅ, ๐ผ) = โ(
1
๐! ๐(๐)๐ข๐(๐ฅ, ๐ผ)
๐๐ฅ๐|๐ฅ=๐ง
(๐ฅ โ1
4)๐
)
๐
๐=0
๐ข(๐ฅ, ๐ผ) = โ(1
๐! ๐(๐)๐ข๐(๐ฅ, ๐ผ)
๐๐ฅ๐|๐ฅ=๐ง
(๐ฅ โ1
4)๐
)
๐
๐=0
, 0 โค ๐ฅ โค 1,0 โค ๐ผ โค 1
hence, the approximated solution is:
93
๐ข(๐ฅ, ๐ผ) = 2.000407015 ๐ฅ + 1.738708618 ๐ผ + .3524026447๐ฅ3
+ 0.1975243281 ๐ฅ4 + 0.1905848255 ๐ฅ5 โ 0.73872694+ 0.04282267662 ๐ผ ๐ฅ4 โ 0.01020620726 ๐ผ ๐ฅ5
โ 1.000007782 ๐ผ ๐ฅ + 0.9961625156 ๐ฅ2 โ 0.4999246903 ๐ผ ๐ฅ2 + 0.1662755673 ๐ผ ๐ฅ3,
and
๐ข(๐ฅ, ๐ผ) = 0.18846813606572846 โ 0.06282271202190949๐ผ
+(3.09423406803286 โ 1.0314113560109548 ๐ผ) (๐ฅ โ1
4)2
+(1.5156229362768463 โ 0.5052076454256154 ๐ผ) (๐ฅ โ1
4)
+(0.2526038227128076 โ 0.08420127423760254๐ผ) (๐ฅ โ1
4)3
Figure (4.1) (a) compares between the exact and numerical solutions for
๐ข(๐ฅ, ๐ผ) in Example (4.1) at ๐ฅ =1
2
Figure (4.1) (a) shows the absolute error between the exact and numerical solutions for
๐ข(๐ฅ, ๐ผ) in Example (4.1) at ๐ฅ =1
2
94
Table (๐. ๐)(๐) shows the comparison between the exact solution and the approximate
solution, and the resulted error.
95
Table 4.1: The exact and numerical solutions for ๐(๐, ๐ถ), and the resulted error by algorithm (๐. ๐) ๐๐ ๐ = ๐
๐
๐ถ Numerical solution ๐(๐, ๐ถ) Exact solution ๐(๐, ๐ถ) Numerical solution ๐(๐, ๐ถ) Exact solution ๐(๐, ๐ถ) Error =D
(๐๐๐๐๐๐, ๐๐๐๐)
0 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 1.045 ร๐๐โ๐
0.1 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 1.0105 ร๐๐โ๐
0.2 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 9.7566 ร๐๐โ๐
0.3 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 9.4082 ร๐๐โ๐
0.4 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 9.0598ร๐๐โ๐
0.5 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 8.7113 ร๐๐โ๐
0.6 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 8.3628 ร๐๐โ๐
0.7 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 8.0144 ร๐๐โ๐
0.8 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 7.6659 ร๐๐โ๐
0.9 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 7.3175 ร๐๐โ๐
1.0 ๐. ๐๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ ๐. ๐๐๐๐๐๐๐๐๐ 6.9690 ร๐๐โ๐
96
These results show the accuracy of the Taylor expansion method to solve
the equation (4.1) with maximum error
= 1.045 ร 10โ3 .
Numerical example (๐. ๐): (๐ป๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐ )
Consider the linear fuzzy Volterra integral equations:
{
๐ข(๐ฅ, ๐ผ) = (๐ผ) cot(๐ฅ) + โซ๐๐ฅโ๐ก๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
0
๐ข(๐ฅ, ๐ผ) = (2 โ ๐ผ) cot(๐ฅ) + โซ๐๐ฅโ๐ก๐ข(๐ก, ๐ผ)๐๐ก
๐ฅ
0
0 โค ๐ผ โค 1 (4.3)
๐ฅ โ [0 , ๐
4]
The analytical solution of the above problem is given by,
๐ข(๐ฅ, ๐ผ) = (๐ผ) (
3
5cos(๐ฅ) +
1
5sin(๐ฅ) +
2
5๐2๐ฅ)
๐ข(๐ฅ, ๐ผ) = (2 โ ๐ผ) (3
5cos(๐ฅ) +
1
5sin(๐ฅ) +
2
5๐2๐ฅ)
(4.4)
0 โค ๐ผ โค 1
Expand the unknown functions in Taylor series at ๐ง =๐
12
To solve the previous example by using the Taylor expansion method using
MAPLE software, we will apply algorithm(4.1) :
๐ธ1,1(1,1)
= ๐ธ2,2(1,1)
=
[ โ.70073 โ.04088 .00364 โ.00024 1.275 ร 10โ5 โ5.598 ร 10โ7
. 29927 โ1.0409 . 00364 โ.00024 1.275 ร 10โ5 โ5.598 ร 10โ7
. 29927 โ.04088 โ.99636 โ.00024 1.275 ร 10โ5 โ5.598 ร 10โ7
. 29927 โ.04088 . 00364 โ1.0002 1.275 ร 10โ5 โ5.598 ร 10โ7
. 29927 โ.04088 . 00364 โ.00024 โ0.99999 โ5.598 ร 10โ7
. 29927 โ.04088 . 00364 โ.00024 1.275 ร 10โ5 โ1.000 ]
97
๐ธ1,2(1,1)
= ๐ธ2,1(1,1)
=
[ 0 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 0]
โ๐๐๐๐,
๐ธ = ๐ธ1,1
(1,1)๐ธ1,2
(1,1)
๐ธ2,1(1,1)
๐ธ2,2(1,1)
๐ท1,1(1,1)
= ๐ท2,2(1,1)
=
[ 0 0 0 0 0 01 0 0 0 0 01 1 0 0 0 01 1 1 0 0 01 1 1 1 0 01 1 1 1 1 0]
๐ท1,2(1,1)
= ๐ท2,1(1,1)
=
[ 0 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 00 0 0 0 0 0]
๐ท = ๐ท1,1
(1,1)๐ท1,2
(1,1)
๐ท2,1(1,1)
๐ท2,2(1,1)
โฑ(๐ง, ๐ผ)
= [
โ๐(๐ง, ๐ผ), โ๐โฒ(๐ง, ๐ผ),โ๐โฒโฒ(๐ง, ๐ผ),โ๐โฒโฒโฒ(๐ง, ๐ผ),โ๐(4)(๐ง, ๐ผ),โ๐(5)(๐ง, ๐ผ)
โ๐(๐ง, ๐ผ),โ๐โฒ(๐ง, ๐ผ), โ๐
โฒโฒ(๐ง, ๐ผ),โ๐
โฒโฒโฒ(๐ง, ๐ผ),โ๐
(4)(๐ง, ๐ผ), โ๐
(5)(๐ง, ๐ผ)
]
๐ก
98
๐ =
[ โ๐ผ cos (
๐
12)
๐ผ ๐ ๐๐ (๐
12)
๐ผ cos (๐
12)
โ๐ผ ๐ ๐๐ (๐
12)
โ๐ผ cos (๐
12)
๐ผ ๐ ๐๐ (๐
12)
โ(2 โ ๐ผ) cos (๐
12)
(2 โ ๐ผ) ๐ ๐๐ (๐
12)
(2 โ ๐ผ) cos (๐
12)
โ(2 โ ๐ผ) ๐ ๐๐ (๐
12)
โ(2 โ ๐ผ) cos (๐
12)
(2 โ ๐ผ) ๐ ๐๐ (๐
12) ]
Solving the linear system
(๐ท + ๐ธ)๐ข = โฑ
we get the following:
99
๐ข(๐ง, ๐ผ) =
[
1.30655494406977๐ผ
1.38836501660070๐ผ
2.06962325214024๐ผ
5.36399137514409๐ผ
11.4350895320246๐ผ
21.4350895320246๐ผ
1.30655494406977(2 โ ๐ผ)
1.38836501660070(2 โ ๐ผ)
2.06962325214024(2 โ ๐ผ)
5.36399137514409(2 โ ๐ผ)
11.4350895320246(2 โ ๐ผ)
21.4350895320246(2 โ ๐ผ)]
{
๐ข(๐ฅ, ๐ผ) = โ(
1
๐! ๐(๐)๐ข๐(๐ฅ, ๐ผ)
๐๐ฅ๐|๐ฅ=๐ง
(๐ฅ โ๐
12)๐
)
๐
๐=0
๐ข(๐ฅ, ๐ผ) = โ(1
๐! ๐(๐)๐ข๐(๐ฅ, ๐ผ)
๐๐ฅ๐|๐ฅ=๐ง
(๐ฅ โ๐
12)๐
)
๐
๐=0
0 โค ๐ฅ โค 1, 0 โค ๐ผ โค 1
the approximated solution is:
๐ข(๐ฅ, ๐ผ) = 0.9999816737 ๐ผ + 1.000399233 ๐ผ ๐ฅ
+ 0.4962378250 ๐ผ ๐ฅ2 + 0.5186782119๐ผ ๐ฅ3
+ 0.2403470047 ๐ผ ๐ฅ4 + 0.1803786182 ๐ผ ๐ฅ5
100
๐ข(๐ฅ, ๐ผ) = 2.000407015 ๐ฅ + 1.738708618 ๐ผ + 0.3524026447 ๐ฅ3 +
0.1975243281 ๐ฅ4 + 0.1905848255 ๐ฅ5 โ 0.7387269443 +
0.04282267662 ๐ผ ๐ฅ 4 โ 0.01020620726 ๐ผ ๐ฅ5 โ 1.000007782 ๐ผ ๐ฅ +
0.9961625156 ๐ฅ2 โ 0.4999246903 ๐ผ ๐ฅ2 + 0.1662755673 ๐ผ ๐ฅ3
Figure (4.2) (a) compares between the exact and numerical solutions for ๐ข(๐ฅ, ๐ผ) in Example
(4.2) at ๐ฅ =๐
8
101
Figure (4.2) (b) shows the absolute error between the exact and numerical solutions for
๐(๐ฅ, ๐ผ) in Example (4.2) at ๐ฅ =๐
8
Table (4.2) shows a comparison between the exact solution and the
approximate solution, and the resulted error.
102
Table 4.2: The exact and numerical solutions for ๐(๐, ๐ถ), and the resulted error by algorithm (4.1) ๐๐ ๐ = ๐
๐
๐ถ Numerical solution ๐(๐,๐ถ) Exact solution ๐(๐,๐ถ)
Numerical solution ๐(๐,๐ถ) Exact solution ๐(๐,๐ถ) Error =D
(๐๐๐๐๐๐, ๐๐๐๐๐๐๐)
0 ๐ 0 3.016349432 3.01635285 3.4210 ร ๐๐โ๐
0.1 0.1508174716
0.15081764 2.865531960 2.86553521 3.2500ร ๐๐โ๐
0.2 0.3016349432 0.30163528 2.714714489 2.71471756 ๐. ๐๐๐๐ ร ๐๐โ๐
0.3 0.4524524148 0.45245292 2.563897017 2.56389992 ๐. ๐๐๐๐ ร ๐๐โ๐
0.4 0.6032698864 0.60327057 2.413079546 2.41308228 ๐. ๐๐๐๐ ร ๐๐โ๐
0.5 0.7540873580 0.75408821 2.262262074 2.26226464 ๐. ๐๐๐๐ ร ๐๐โ๐
0.6 0.9049048296 0.90490585 2.111444602 2.11144699 ๐. ๐๐๐๐ ร ๐๐โ๐
0.7 1.055722301 1.05572349 1.960627131 1.96062935 ๐. ๐๐๐๐ ร ๐๐โ๐
0.8 1.206539773 1.20654114 1.809809659 1.80981171 ๐. ๐๐๐๐ ร ๐๐โ๐
0.9 1.357357244 1.35735878 1.658992188 1.65899406 ๐. ๐๐๐๐ ร ๐๐โ๐
1.0 1.508174716 1.50817642 1.508174716 1.50817642 ๐. ๐๐๐๐ ร ๐๐โ๐
103
These results show the accuracy of the Taylor expansion method to solve
the equation (4.2) with a maximum error
= 3.4210 ร 10โ6.
Numerical example (๐. ๐): (๐ป๐๐๐๐๐๐๐๐ ๐๐ ๐๐๐๐๐๐ )
The fuzzy Volterra integral equations (4.1) have the analytical
solution (4.2)
The following algorithm implements the trapezoidal rule using the MAPLE
software.
Algorithm (๐. ๐)
1. Input ๐ฅ0 = ๐, ๐ฅ๐ = ๐, ๐ , ๐๐,๐(๐ฅ, ๐ก), ๐๐(๐ฅ, ๐ผ), ๐๐(๐ฅ, ๐ผ)
2. Input ๐ก๐ = ๐ฅ๐ , ๐๐๐ ๐ = 0,1, , โฆ , ๐
3. Input ( ๐ ) the number of subdivisions of [๐, ๐]
4. Calculate ๐ฟ =๐โ๐
๐ , s= 0,1,โฆ , ๐
5. Calculate ๐ฅ๐= ๐. ๐ฟ ๐ = 0,โฆ , ๐
6. Put ๐1,1 = ๐๐,๐ =
{
โ๐ฟ. ๐(๐ฅ๐ , ๐ฅ๐) 1 โค ๐ < ๐ โค ๐
1 โ๐ฟ
2. ๐(๐ฅ๐ , ๐ฅ๐) 1 โค ๐ โค ๐
1 ๐ = ๐ = 0
โ๐ฟ
2. ๐(๐ฅ๐ , ๐ฅ0) 1 โค ๐ โค ๐
0 ๐๐๐ ๐ ๐คโ๐๐๐
7. Put ๐1,1 = ๐2,2
104
8. Put
๐1,2 = ๐2,1 =
[ 0 0 โฏ 0 00 0 โฏ 0 0โฎ โฎ โฑ โฎ โฎ0 0 โฏ 0 00 0 โฏ 0 0]
(๐ +1)ร(๐ +1)
9. denote ๐ = ๐1,1 ๐1,2
๐2,1 ๐2,2 ,
10. put
โฑ = [โ๐1(๐ง, ๐ผ), โฆ ,โ๐1
(๐)(๐ง, ๐ผ),โ๐1(๐ง, ๐ผ),โฆ ,โ๐
1
(๐ )(๐ง, ๐ผ),โฆ ,
โ๐๐(๐ง, ๐ผ),โฆ , โ๐๐(๐ )(๐ง, ๐ผ), โ๐
๐(๐ง, ๐ผ), โฆ ,โ๐
๐
(๐ )(๐ง, ๐ผ)
]
๐ก
11. put
๐ข = ๐ข1๐ (๐ง, ๐ผ),โฆ , ๐ข1๐
(๐ )(๐ง, ๐ผ), ๐ข1๐ (๐ง, ๐ผ), โฆ , ๐ข1๐ (๐ )
(๐ง, ๐ผ), โฆ ,
๐ข๐๐ (๐ง, ๐ผ),โฆ , ๐ข๐๐ (๐ )(๐ง, ๐ผ), ๐ข๐๐ (๐ง, ๐ผ),โฆ , ๐ข๐๐
(๐ )(๐ง, ๐ผ)
๐ก
12. solve the following linear system ๐๐ข = โฑ
13. estimate ๐ข(๐ง, ๐ผ), ๐ข(๐ง, ๐ผ)
If we take ๐ฅ = 21. ๐ฟ , we get :
๐ข(๐ฅ, ๐ผ) = 0.7479719827489786. ๐ผ + 0.7479719827489785 ,
and
๐ข(๐ฅ, ๐ผ) = โ0.7479719827489785. ๐ผ + 2.2439159482469355.
105
Figure (4.3) (a) compares the exact and numerical solutions for ๐ข(๐ฅ, ๐ผ) in Example (4.3) at
๐ฅ = 0.84 with ๐ = 25
Figure (4.3) (b) shows the absolute error between the exact and numerical solutions in Example
(4.3) at ๐ฅ = 0.84 ๐ค๐๐กโ ๐ = 25
106
Table 4.3: The exact and numerical solutions for ๐(๐, ๐ถ), and the resulted error by algorithm (4.2) ๐๐ ๐ = ๐. ๐๐
๐ถ
Numerical
solution
๐(๐,๐ถ)
Exact solution ๐(๐,๐ถ) Numerical
solution ๐(๐,๐ถ)
Exact solution
๐(๐,๐ถ)
Error =D
(๐๐๐๐๐๐, ๐๐๐๐๐๐๐)
0 0.747971982 0.7480775000 2.243915948246 2.244232500 3.165517๐ ร ๐๐โ๐
0.1 0.822769181 0.8228852500 2.169118749972 2.169424750 3.0600002ร ๐๐โ๐
0.2 0.897566379 0.8976930000 2.094321551697 2.094617000 2.9544830ร ๐๐โ๐
0.3 0.972363577 0.9725007500 2.019524353422 2.019809250 2.8489657ร ๐๐โ๐
0.4 1.047160775 1.047308500 1.944727155147 1.945001500 2.7434485ร ๐๐โ๐
0.5 1.121957974 1.122116250 1.869929956872 1.870193750 2.6379312ร ๐๐โ๐
0.6 1.196755172 1.196924000 1.795132758597 1.870193750 2.5324140ร ๐๐โ๐
0.7 1.271552370 1.271731750 1.720335560322 1.720578250 2.4268967ร ๐๐โ๐
0.8 1.346349568 1.346539500 1.645538362047 1.645770500 2.3213795ร ๐๐โ๐
0.9
1.421146767
1.421347250 1.570741163772 1.570962750 2.2158622ร ๐๐โ๐
1.0 1.495943965 1.496155000 1.495943965497 1.496155000 2.1103450ร ๐๐โ๐
107
These results show the efficiency of the Trapezoidal method to solve the
equation (4.1) with a maximum error
= 3.1655179 ร 10โ4
Numerical example (๐. ๐): (๐ป๐๐๐๐๐๐๐๐ ๐๐ ๐๐๐๐๐๐ )
The fuzzy Volterra integral equations (4.3) have the analytical solution
(4.4)
Algorithm (4.2) implements the trapezoidal method using the MAPLE
software, on the interval [0,๐
4] with s=25
If we take ๐ฅ = 23. ๐ฟ =0.23๐ , we get:
๐ข(๐ฅ, ๐ผ) = 2.275392420 . ๐ผ, where , 0โค ๐ผ โค 1
๐ข(๐ฅ, ๐ผ) = โ2.275392420. ๐ผ + 4.550784838.
108
Figure (4.4) (a) compares the exact and numerical solutions for ๐ข(๐ฅ, ๐ผ) in Example (4.4) at
๐ฅ = 0.23๐ with ๐ = 25
Figure (4.4) (b) shows the absolute error between the exact and numerical solutions for๐ข(๐ฅ, ๐ผ)
in Example (4.4) at ๐ฅ = 0.23๐ with ๐ = 25
109
Table 4.4: The exact and numerical solutions for ๐(๐, ๐ถ), and the resulted error by algorithm (4.2) at ๐ = ๐. ๐๐๐ with ๐ = ๐๐
๐ถ Numerical
solution ๐(๐,๐ถ) Exact solution ๐(๐,๐ถ) Numerical solution ๐(๐,๐ถ)
Exact solution ๐(๐,๐ถ)
Error =D (๐๐๐๐๐๐, ๐ผ๐๐๐๐)
0 0 0 4.550784838 4.55858982 7.804982ร ๐๐โ๐
0.1 0.2275392420 0.2279294911 4.323245596 4.33066033 7.414733ร ๐๐โ๐
0.2 0.4550784840 0.4558589821 4.095706354 4.10273083 7.024484ร ๐๐โ๐
0.3 0.6826177260 0.6837884731 3.868167112 3.874801348 6.634235ร ๐๐โ๐
0.4 0.9101569680 0.9117179641 3.640627870 3.64687185 6.243986ร ๐๐โ๐
0.5 1.137696210 1.139647455 3.413088628 3.41894236 5.853736ร ๐๐โ๐
0.6 1.365235452 1.367576946 3.185549386 3.19101287 5.463487ร ๐๐โ๐
0.7 1.592774694 1.595506437 2.958010144 2.96308338 5.073238ร ๐๐โ๐
0.8 1.820313936
1.823435928
2.730470902 2.73515389 4.682989ร ๐๐โ๐
0.9
2.047853178
2.051365419 2.502931660 2.50722440 4.292740ร ๐๐โ๐
1.0 2.275392420 2.279294911 2.275392418 2.279294911 3.902491ร ๐๐โ๐
110
These results show the efficiency of the Trapezoidal method to solve the
equation (4.3) with a maximum error
= 7.804982 ร 10โ3.
We notice from our numerical test cases that the Taylor expansion method
is more accurate than the trapezoidal method.
Numerical example (๐. ๐): (๐๐๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐ )
The fuzzy Volterra integral equation(4.1) have the exact solution(4.2).
In the view of the variational iteration method, we construct a correction
functional in the following form:
๐ข๐+1(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + โซ๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ฅ
0
,
and (4.5)
๐ข๐+1(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + โซ๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ฅ
0
.
where ๐ = 0,1,2, โฆ , ๐
Starting with the initial approximation:
๐ข0(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2,
and
๐ข0(๐ฅ, ๐ผ) = (3 โ ๐ผ)๐ฅ2.
111
in equation (4.1) successive approximations ๐ข๐(๐ฅ, ๐ผ)โฒ๐ will be achieved.
In this example we calculate the 8th order of approximate solution using
the variational iteration method by MAPLE software.
Algorithm (๐. ๐)
1. Input ๐, ๐, ๐, ๐(๐ฅ, ๐ก), ๐ , ๐(๐ฅ, ๐ผ), ๐(๐ฅ, ๐ผ)
2. Input ๐ข0(๐ฅ, ๐ผ), ๐ข0(๐ฅ, ๐ผ)
3. for ๐ = 1 ๐ก๐ ๐ + 1, compute
๐ข๐+1(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + โซ ๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก๐ฅ
0,
and
๐ข๐+1(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + โซ๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ฅ
0
We obtain the following results:
๐ข1(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2 โ1
4(๐ผ + 1)๐ฅ4 +
1
3๐ฅ4(๐ผ + 1)
๐ข2(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2 โ
1
6(
1
12๐ผ +
1
12) ๐ฅ6 +
1
5๐ฅ6 (
1
12๐ผ +
1
12) โ
1
4๐ฅ4(๐ผ + 1) +
1
3๐ฅ4(๐ผ + 1)
๐ข3(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2 โ1
8(
1
360๐ผ +
1
360) ๐ฅ8 +
1
7๐ฅ8 (
1
360๐ผ +
1
360) โ
1
6๐ฅ6 (
1
12๐ผ +
1
12) +
1
5๐ฅ6 (
1
12๐ผ +
1
12) โ
1
4๐ฅ4(๐ผ + 1) +
1
3๐ฅ4(๐ผ + 1)
๐ข4(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2 โ1
10(
1
20160๐ผ +
1
20160) ๐ฅ10 +
1
9๐ฅ10 (
1
20160๐ผ +
1
20160) โ
1
8(
1
360๐ผ +
1
360) ๐ฅ8 +
1
7๐ฅ8 (
1
360๐ผ +
1
360) โ
1
6๐ฅ6 (
1
12๐ผ +
1
12) +
1
5๐ฅ6 (
1
12๐ผ +
1
12) โ
1
4๐ฅ4(๐ผ + 1) +
1
3๐ฅ4(๐ผ + 1)
๐ข5(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2 โ1
12(
1
1814400๐ผ +
1
1814400) ๐ฅ12 +
1
11๐ฅ12 (
1
1814400๐ผ +
1
1814400) โ
1
10(
1
20160๐ผ +
1
20160) ๐ฅ10 +
112 1
9๐ฅ10 (
1
20160๐ผ +
1
20160) โ
1
8(
1
360๐ผ +
1
360) ๐ฅ8 +
1
7๐ฅ8 (
1
360๐ผ +
1
360) โ
1
6๐ฅ6 (
1
12๐ผ +
1
12) +
1
5๐ฅ6 (
1
12๐ผ +
1
12) โ
1
4๐ฅ4(๐ผ + 1) +
1
3๐ฅ4(๐ผ + 1)
๐ข6(๐ฅ, ๐ผ) =
1
13(
1
239500800๐ผ +
1
239500800) ๐ฅ14 โ
1
12(
1
1814400๐ผ +
1
1814400) ๐ฅ12 +
1
11๐ฅ12 (
1
1814400๐ผ +
1
1814400) โ
1
10(
1
20160๐ผ +
1
20160) ๐ฅ10 +
1
9๐ฅ10 (
1
20160๐ผ +
1
20160) โ
1
8(
1
360๐ผ +
1
360) ๐ฅ8 +
1
7๐ฅ8 (
1
360๐ผ +
1
360) โ
1
6๐ฅ6 (
1
12๐ผ +
1
12) +
1
5๐ฅ6 (
1
12๐ผ +
1
12) โ
1
4๐ฅ4(๐ผ + 1) +
1
3๐ฅ4(๐ผ + 1)
๐ข7(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2 โ
1
16(
1
43589145600๐ผ +
1
43589145600) ๐ฅ16 +
1
15(
1
43589145600๐ผ +
1
43589145600) ๐ฅ16 โ
1
14(
1
239500800๐ผ +
1
239500800) ๐ฅ14 +
1
13(
1
239500800๐ผ +
1
239500800) ๐ฅ14 โ
1
12(
1
1814400๐ผ +
1
1814400) ๐ฅ12 +
1
11๐ฅ12 (
1
1814400๐ผ +
1
1814400) โ
1
10(
1
20160๐ผ +
1
20160) ๐ฅ10 +
1
9๐ฅ10 (
1
20160๐ผ +
1
20160) โ
1
8(
1
360๐ผ +
1
360) ๐ฅ8 +
1
7๐ฅ8 (
1
360๐ผ +
1
360) โ
1
6๐ฅ6 (
1
12๐ผ +
1
12) +
1
5๐ฅ6 (
1
12๐ผ +
1
12) โ
1
4๐ฅ4(๐ผ + 1) +
1
3๐ฅ4(๐ผ + 1)
๐ข8(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2 โ1
18(
1
10461394944000๐ผ +
1
10461394944000) ๐ฅ18 +
1
17(
1
10461394944000๐ผ +
1
10461394944000) ๐ฅ18 +
1
16(
1
43589145600๐ผ +
1
43589145600) ๐ฅ16 +
1
15(
1
43589145600๐ผ +
1
43589145600) ๐ฅ16 โ
1
14(
1
239500800๐ผ +
1
239500800) ๐ฅ14 +
1
13(
1
239500800๐ผ +
1
239500800) ๐ฅ14 โ
1
12(
1
1814400๐ผ +
1
1814400) ๐ฅ12 +
1
11๐ฅ12 (
1
1814400๐ผ +
1
1814400) โ
1
10(
1
20160๐ผ +
1
20160) ๐ฅ10 +
1
9๐ฅ10 (
1
20160๐ผ +
1
20160) โ
1
8(
1
360๐ผ +
1
360) ๐ฅ8 +
1
7๐ฅ8 (
1
360๐ผ +
1
360) โ
1
6๐ฅ6 (
1
12๐ผ +
1
12) +
1
5๐ฅ6 (
1
12๐ผ +
1
12) โ
1
4๐ฅ4(๐ผ + 1) +
1
3๐ฅ4(๐ผ + 1)
113
and
๐ข1(๐ฅ, ๐ผ) = (๐ผ + 1)๐ฅ2 โ1
4(3 โ ๐ผ)๐ฅ4 +
1
3๐ฅ4(3 โ ๐ผ)
๐ข2(๐ฅ, ๐ผ) = (3 โ ๐ผ)๐ฅ2 โ
1
6(1
4โ
1
12๐ผ)๐ฅ6 +
1
5๐ฅ6 (
1
4โ
1
12๐ผ) โ
1
4๐ฅ4(3 โ ๐ผ) +
1
3๐ฅ4(3 โ ๐ผ)
๐ข3(๐ฅ, ๐ผ) = (3 โ ๐ผ)๐ฅ2 โ1
8(
1
120โ
1
360๐ผ) ๐ฅ8 +
1
7๐ฅ8 (
1
120โ
1
360๐ผ) โ
1
6๐ฅ6 (
1
4โ
1
12๐ผ) +
1
5๐ฅ6 (
1
4โ
1
12๐ผ) โ
1
4๐ฅ4(3 โ ๐ผ) +
1
3๐ฅ4(3 โ ๐ผ)
๐ข4(๐ฅ, ๐ผ) = (3 โ ๐ผ)๐ฅ2 โ
1
10(
1
6720โ
1
20160๐ผ)๐ฅ10 +
1
9๐ฅ10 (
1
6720โ
1
20160๐ผ) โ
1
8(
1
120โ
1
360๐ผ)๐ฅ8 +
1
7๐ฅ8 (
1
120โ
1
360๐ผ) โ
1
6๐ฅ6 (
1
4โ
1
12๐ผ) +
1
5๐ฅ6 (
1
4โ
1
12๐ผ) โ
1
4๐ฅ4(3 โ ๐ผ) +
1
3๐ฅ4(3 โ ๐ผ)
๐ข5(๐ฅ, ๐ผ) = (3 โ ๐ผ)๐ฅ2 โ1
12(
1
604800โ
1
1814400๐ผ)๐ฅ12 +
1
11๐ฅ12 (
1
604800โ
1
1814400๐ผ) โ
1
10(
1
6720โ
1
20160๐ผ)๐ฅ10 +
1
9๐ฅ10 (
1
6720โ
1
20160๐ผ) โ
1
8(
1
120โ
1
360๐ผ)๐ฅ8 +
1
7๐ฅ8 (
1
120โ
1
360๐ผ) โ
1
6๐ฅ6 (
1
4โ
1
12๐ผ) +
1
5๐ฅ6 (
1
4โ
1
12๐ผ) โ
1
4๐ฅ4(3 โ ๐ผ) +
1
3๐ฅ4(3 โ ๐ผ).
๐ข6(๐ฅ, ๐ผ) = (3 โ ๐ผ)๐ฅ2 โ
1
14(
1
79833600โ
1
239500800๐ผ) ๐ฅ14 +
1
13(
1
79833600โ
1
239500800๐ผ)๐ฅ14 โ
1
12(
1
604800โ
1
1814400๐ผ)๐ฅ12 +
1
11๐ฅ12 (
1
604800โ
1
1814400๐ผ) โ
1
10(
1
6720โ
1
20160๐ผ) ๐ฅ10 +
1
9๐ฅ10 (
1
6720โ
1
20160๐ผ) โ
1
8(
1
120โ
1
360๐ผ) ๐ฅ8 +
1
7๐ฅ8 (
1
120โ
1
360๐ผ) โ
1
6๐ฅ6 (
1
4โ
1
12๐ผ) +
1
5๐ฅ6 (
1
4โ
1
12๐ผ) โ
1
4๐ฅ4(3 โ ๐ผ) +
1
3๐ฅ4(3 โ ๐ผ).
๐ข7(๐ฅ, ๐ผ) = (3 โ ๐ผ)๐ฅ2 โ
1
16(
1
14529715200โ
1
43589145600๐ผ)๐ฅ16 +
1
15(
1
14529715200โ
1
43589145600๐ผ)๐ฅ16 โ
1
14(
1
79833600โ
1
239500800๐ผ)๐ฅ14 +
1
13(
1
79833600โ
1
239500800๐ผ) ๐ฅ14 โ
114 1
12(
1
604800โ
1
1814400๐ผ)๐ฅ12 +
1
11๐ฅ12 (
1
604800โ
1
1814400๐ผ) โ
1
10(
1
6720โ
1
20160๐ผ)๐ฅ10 +
1
9๐ฅ10 (
1
6720โ
1
20160๐ผ) โ
1
8(
1
120โ
1
360๐ผ)๐ฅ8 +
1
7๐ฅ8 (
1
120โ
1
360๐ผ) โ
1
6๐ฅ6 (
1
4โ
1
12๐ผ) +
1
5๐ฅ6 (
1
4โ
1
12๐ผ) โ
1
4๐ฅ4(3 โ ๐ผ) +
1
3๐ฅ4(3 โ ๐ผ).
๐ข8(๐ฅ, ๐ผ) = (3 โ ๐ผ)๐ฅ2 โ
1
18(
1
3487131648000โ
1
10461394944000๐ผ)๐ฅ18 +
1
17(
1
3487131648000โ
1
10461394944000๐ผ)๐ฅ18 โ
1
16(
1
14529715200โ
1
43589145600๐ผ) ๐ฅ16 +
1
15(
1
14529715200โ
1
43589145600๐ผ)๐ฅ16 โ
1
14(
1
79833600โ
1
239500800๐ผ)๐ฅ14 +
1
13(
1
79833600โ
1
239500๐ผ)๐ฅ14 โ
1
12(
1
604800โ
1
1814400๐ผ)๐ฅ12 +
1
11๐ฅ12 (
1
604800โ
1
1814400๐ผ) โ
1
10(
1
6720โ
1
20160๐ผ) ๐ฅ10 +
1
9๐ฅ10 (
1
6720โ
1
20160๐ผ) โ
1
8(
1
120โ
1
360๐ผ)๐ฅ8 +
1
7๐ฅ8 (
1
120โ
1
360๐ผ) โ
1
6๐ฅ6 (
1
4โ
1
12๐ผ) +
1
5๐ฅ6 (
1
4โ
1
12๐ผ) โ
1
4๐ฅ4(3 โ ๐ผ) +
1
3๐ฅ4(3 โ ๐ผ) .
Figure (4.5) (a) compares the exact and numerical solutions for ๐ข(๐ฅ, ๐ผ) in Example (4.5) at ๐ฅ = 0.5
115
Figure (4.5) (b) shows the absolute error between the exact and numerical solutions for ๐ข(๐ฅ, ๐ผ) in Example (4.5) at ๐ฅ = 0.5
Table (4.5)(๐) shows the exact and numerical results when applying
algorithm(4.3), and showing the absolute error between the exact and
numerical solutions of equation(4.1)
116
Table ๐. ๐: the exact and numerical results, the absolute resulted error of applying algorithm (๐. ๐) for
equation(๐. ๐) ๐๐ ๐ = ๐. ๐.
๐ถ Numerical solution ๐(๐,๐ถ) Exact solution ๐(๐,๐ถ) Numerical solution
๐(๐,๐ถ) Exact solution
๐(๐,๐ถ) Error =D
(๐๐๐๐๐๐, ๐๐๐๐๐๐๐)
0 0.2552519305 0.255251931 0.7657557912 0.765755793 1.8000ร ๐๐โ๐
0.1 0.2807771236 0.2807771241 0.7402305982 0.7402305999 1.7500ร ๐๐โ๐
0.2 0.3063023166 0.3063023172 0.7147054051 0.7147054068 1.7000ร ๐๐โ๐
0.3 0.3318275096 0.3318275103 0.6891802120 0.6891802137 1.6500ร ๐๐โ๐
0.4 0.3573527027 0.3573527034 0.6636550190 0.6636550206 1.6000ร ๐๐โ๐
0.5 0.3828778957 0.3828778965 0.6381298260 0.6381298275 1.5500ร ๐๐โ๐
0.6 0.4084030888 0.4084030896 0.6126046329 0.6126046344 1.5000ร ๐๐โ๐
0.7 0.4339282819 0.4339282827 0.5870794398 0.5870794413 1.4500ร ๐๐โ๐
0.8 0.4594534749 0.4594534758 0.5615542468 0.5615542482 1.4000ร ๐๐โ๐
0.9 0.4849786679 0.4849786689 0.5360290538 0.5360290551 ๐. ๐๐๐๐ ร ๐๐โ๐
1.0 0.5105038610 0.5105038620 0.5105038607 0.5105038620 ๐. ๐๐๐๐ ร ๐๐โ๐
117
The results showed that the convergence and accuracy of variational
iteration method for numerical solution the Volterra integral equations were
in a good agreement with the analytical solutions.
Error =๐ท (๐ข๐๐ฅ๐๐๐ก(. 5, ๐ผ), ๐ข๐๐๐๐๐๐ฅ๐๐๐๐ก๐(. 5, ๐ผ))
= sup0โค๐ผโค1
{๐๐๐ฅ ๐ข๐๐ฅ๐๐๐ก โ ๐ข๐๐๐๐๐๐ฅ๐๐๐๐ก๐ ,๐๐๐ฅ ๐ข๐๐ฅ๐๐๐ก โ ๐ข๐๐๐๐๐๐ฅ๐๐๐๐ก๐ }
=1. 8 ร 10โ9
Numerical example (๐. ๐): (๐๐๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐๐๐๐ ๐๐๐๐๐๐ )
The fuzzy Volterra integral equation(4.3) have the exact solution(4.4).
In the view of the variational iteration method, we construct a correction
functional in the following form:
๐ข๐+1(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + โซ๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ฅ
0
,
and
๐ข๐+1(๐ฅ, ๐ผ) = ๐(๐ฅ, ๐ผ) + โซ๐(๐ฅ, ๐ก)๐ข๐(๐ก, ๐ผ)๐๐ก
๐ฅ
0
.
where ๐ = 0,1,2, โฆ , ๐
Starting with the initial approximation:
๐ข0(๐ฅ, ๐ผ) = (๐ผ) cos ๐ฅ,
and
๐ข0(๐ฅ, ๐ผ) = (2 โ ๐ผ) cos ๐ฅ.
118
where , 0 โค ๐ผ โค 1
In equation (4.3) successive approximations ๐ข๐(๐ฅ, ๐ผ)โฒ๐ will be achieved.
In this example we calculate the 6th order of approximate solution using the
variational iteration method by MAPLE software.
applying algorithm(4.3), we obtain the following results:
๐ข1(๐ฅ, ๐ผ) =1
2๐ผ cos ๐ฅ +
1
2๐ผ๐๐ฅ +
1
2๐ผ sin ๐ฅ
๐ข2(๐ฅ, ๐ผ) =
1
2๐ผ cos ๐ฅ +
1
2๐ผ๐๐ฅ +
1
2๐ผ๐๐ฅ๐ฅ
๐ข3(๐ฅ, ๐ผ) =3
4๐ผ cos ๐ฅ +
1
4๐ผ๐๐ฅ +
1
4๐ผ sin ๐ฅ +
1
2๐ผ๐๐ฅ๐ฅ +
1
2๐ผ๐๐ฅ๐ฅ2
๐ข4(๐ฅ, ๐ผ) = 1
2๐ผ cos ๐ฅ +
1
2๐ผ๐๐ฅ +
1
4๐ผ sin ๐ฅ++
1
4๐ผ๐๐ฅ๐ฅ +
1
4๐ผ๐๐ฅ๐ฅ2 +
1
12๐ผ๐๐ฅ๐ฅ3
๐ข5(๐ฅ, ๐ผ) =
5
8๐ผ cos ๐ฅ +
3
8๐ผ๐๐ฅ +
1
8๐ผ sin ๐ฅ++
1
2๐ผ๐๐ฅ๐ฅ +
1
8๐ผ๐๐ฅ๐ฅ2 +
1
12๐ผ๐๐ฅ๐ฅ3 +
1
48๐ผ๐๐ฅ๐ฅ4
๐ข6(๐ฅ, ๐ผ) =
5
8๐ผ cos ๐ฅ +
3
8๐ผ๐๐ฅ +
1
4๐ผ sin ๐ฅ++
3
8๐ผ๐๐ฅ๐ฅ +
1
4๐ผ๐๐ฅ๐ฅ2 +
1
24๐ผ๐๐ฅ๐ฅ3 +
1
48๐ผ๐๐ฅ๐ฅ4 1
240๐ผ๐๐ฅ๐ฅ5
and ๐ข1(๐ฅ, ๐ผ) = (2 โ ๐ผ )cos ๐ฅ + ๐๐ฅ โ
1
2๐ผ๐๐ฅ โcos ๐ฅ + sin ๐ฅ +
1
2๐ผ cos ๐ฅ โ
1
2๐ผ sin ๐ฅ
119
๐ข2(๐ฅ, ๐ผ) = (2 โ ๐ผ )cos ๐ฅ + ๐๐ฅ โ 1
2๐ผ๐๐ฅ โcos ๐ฅ +
1
2๐ผ cos ๐ฅ + ๐๐ฅ๐ฅ -
1
2๐ผ๐๐ฅ๐ฅ.
๐ข3(๐ฅ, ๐ผ) = (2 โ ๐ผ )cos ๐ฅ +1
2๐๐ฅ โ
1
4๐ผ๐๐ฅ โ
1
2cos ๐ฅ +
1
2sin ๐ฅ +
1
4๐ผ cos ๐ฅ ๐ โ
1
4๐ผ sin ๐ฅ + ๐๐ฅ๐ฅ โ
1
2๐ผ๐๐ฅ๐ฅ +
1
2๐๐ฅ๐ฅ2 โ
1
4๐ผ๐๐ฅ๐ฅ2
๐ข4(๐ฅ, ๐ผ) = (2 โ ๐ผ )cos ๐ฅ + ๐๐ฅ โ
1
2๐ผ๐๐ฅ โ
1
12๐ผ๐๐ฅ๐ฅ3 +
1
2๐๐ฅ๐ฅ +
1
2๐๐ฅ๐ฅ2 +
1
6๐๐ฅ๐ฅ3 โ cos ๐ฅ +
1
2sin ๐ฅ โ
1
4๐ผ๐๐ฅ๐ฅ โ
1
4๐ผ๐๐ฅ๐ฅ2 +
1
2๐ผ cos ๐ฅ โ
1
4๐ผ sin ๐ฅ
๐ข5(๐ฅ, ๐ผ) = (2 โ ๐ผ )cos ๐ฅ +
3
4๐๐ฅ โ
3
8๐ผ๐๐ฅ +
1
4๐๐ฅ๐ฅ2 +
3
8๐ผ cos ๐ฅ +
1
6๐๐ฅ๐ฅ3 +
1
4sin ๐ฅ + ๐๐ฅ๐ฅ โ
3
4cos ๐ฅ โ
1
48๐ผ๐๐ฅ๐ฅ4 โ
1
8๐ผ sin ๐ฅ โ
1
12๐ผ๐๐ฅ๐ฅ3 โ
1
8๐ผ๐๐ฅ๐ฅ2 โ
1
2๐ผ๐๐ฅ๐ฅ +
1
24๐๐ฅ๐ฅ4
๐ข6(๐ฅ, ๐ผ) = (2 โ ๐ผ )cos ๐ฅ +
3
4๐๐ฅ โ
3
8๐ผ๐๐ฅ +
1
2๐๐ฅ๐ฅ2 +
3
8๐ผ cos ๐ฅ +
1
12๐๐ฅ๐ฅ3 +
1
2sin ๐ฅ +
3
4 ๐๐ฅ๐ฅ โ
3
4cos ๐ฅ +
1
120๐๐ฅ๐ฅ5 โ
1
240๐ผ๐๐ฅ๐ฅ5 โ
1
48๐ผ๐๐ฅ๐ฅ4 โ
1
4๐ผ sin ๐ฅ โ
1
24๐ผ๐๐ฅ๐ฅ3 โ
1
4๐ผ๐๐ฅ๐ฅ2 โ
3
8๐ผ๐๐ฅ๐ฅ +
1
24๐๐ฅ๐ฅ4
120
Figure (4.6) (a) compares the exact and numerical solutions for ๐ข(๐ฅ, ๐ผ) in Example (4.6) at
๐ฅ =๐
8
Figure (4.6) (b) shows absolute error between the exact and numerical solutions for ๐ข(๐ฅ, ๐ผ)
in Example (4.6) at ๐ฅ =๐
8
Table (4.6) shows the exact and numerical results when applying algorithm
(4.3), and showing the resulted error of equation(4.3).
121
Table ๐. ๐: the exact and numerical results, the resulted absolute error by applying algorithm (๐. ๐) for equation
(๐. ๐) ๐๐ ๐ =๐
๐.
๐ถ Numerical solution
๐(๐,๐ถ) Exact solution ๐(๐,๐ถ)
Numerical solution
๐(๐,๐ถ) Exact solution ๐(๐,๐ถ)
Error =D
(๐๐๐๐๐๐, ๐๐๐๐๐๐๐)
0 0 0 3.016352007 3.01635285 . ๐๐๐๐ ร ๐๐โ๐8
0.1 0.1508176002 0.1508176426 2.865534406 2.865535210 8.0400ร ๐๐โ๐
0.2 0.3016352007 0.3016352853 2.714716807 2.71471756 7.6100ร ๐๐โ๐
0.3 0.4524528010 0.4524529279 2.563899206
2.56389992
7.1900ร ๐๐โ๐
0.4 0.6032704013 0.6032705706 2.413081606 2.41308228 6.7700ร ๐๐โ๐
0.5 0.7540880017 0.7540882132 2.262264004 2.262264640 6.3600ร ๐๐โ๐
0.6 0.9049056021 0.9049058558 2.111446404 2.11144699 5.9300ร ๐๐โ๐
0.7 1.055723203 1.055723498 1.960628805 1.96062935 5.5000ร ๐๐โ๐
0.8 1.206540802 1.206541141 1.809811204 1.80981171 5.0700ร ๐๐โ๐
0.9 1.357358403 1.357358784 1.658993604 1.65899406 ๐. ๐๐๐๐ ร ๐๐โ๐
1.0 1.508176002 1.508176426 1.508176002 1.508176426 ๐. ๐๐๐๐ ร ๐๐โ๐
122
Table (4.6) shows the convergence and accuracy of variational iteration
method for numerical solution the Volterra integral equations were in a
good agreement with the analytical solutions , with a maximum error =
8.46 ร 10โ7
from our numerical test cases, we conclude that the variational iteration
method is more efficient than the trapezoidal and the Taylor methods .
123
Conclusion
In this thesis we have solved the linear fuzzy Volterra integral equation of
the second kind using various analytical and numerical methods.
The numerical methods methods were implemented in a form of
algorithms to solve some numerical tests cases using MAPLE software
We have obtained the following results
Numerical method Maximum error
Taylor expansion method 3.4210 ร 10โ6
Trapezoidal rule 7.804982 ร 10โ3
Variation iteration method 8.46 ร 10โ7
Numerical results have shown to be in a close agreement with the
analytical ones. Moreover, the variation iteration method is one of the
most powerful numerical technique for solving fuzzy Volterra integral
equation of the second kind in comparison with other numerical techniques.
124
References
[๐] S. Abbasbandy and M. Alvai, A method for Solving Fuzzy Linear
Systems, Iranian Journal of Fuzzy Systems, Vol. 2, No. 2, 2005, 37-
43.
[๐] T. Allahviranloo, M. Ghanbari, and R. Nuraei, An application of semi-
analytical method on linear fuzzy Volterra integral equations,
Journal of Fuzzy set valued Analysis, Vol.2014, 2014, 1-15.
[๐] T. Allahviranloo, M. Jahantigh, and M. Otidi, Numerical Solution
of Fuzzy Integral Equations, Applied Mathematical Science,
Vol. 2, No.1, 2008, 33-46.
[๐] T. Allahviranloo, N. Motamedi, and N. Kiana, Solving fuzzy
integral equations by differential transformation method,
Information Sciences, Vol. 197, Issue 7, 2009, 956-966.
[๐] T. Allahviranloo, S. Salahshour, M. Nejad, and D. Baleanu, General
Solutions of Fully Fuzzy Linear Systems, Abstract and Applied
Analysis, Vol. 2013, 2013, 1-9.
[๐] M. Amawi, Fuzzy Fredholm Integral Equation of the Second Kind,
M. Sc Thesis, An-Najah University, 2014.
[๐] M. Amawi and N. Qatanani, Numerical Methods for Solving Fuzzy
Fredholm Integral Equation of the Second Kind, International
Journal of Applied Mathematics, Vol.28, No.3, 2015, 177-195.
125
[๐] S. Amrahov and I. Askerzade, Strong Solution of Fuzzy Linear
Systems, Ankara university, Computer Modeling in Engineering and
Science,Vol.76, No. 4, 2011, 207-216.
[๐] E. Babolian, H. Goughery, and S. Abbasbandy, Numerical solution
of linear Fredholm fuzzy integral equations of the second kind by
Adomian method, Applied Mathematics and Computation, Vol.
161, Issue 3, 2005, 733-744.
[๐๐] S. Banerjee and T. Roy, Arithmatic Operations on Generalized
Trapezoidal Fuzzy Number and its Applications, An Official
Journal of Turkish Fuzzy systems Association, Vol. 3, No. 1, 2012,
16-44.
[๐๐] M. Barkhordari Ahmadi, and N. Kiani, Solving Fuzzy Partial
Differential Equation by Differential Transformation Method,
Journal of Applied Mathematics, Vol.7, No.4, 2011, 1-16.
[๐๐] B. Beats, D. Dubois, and E. Hullermeier, Fuzzy Sets and Systems,
An International Journal in Information Science and
Engineering, Vol. 293, 2016, 1-144.
[๐๐] B. Bede, I. Rudas, and L. Bencsik, First order linear fuzzy
differential equations under generalized differentiability, Fuzzy
Set Applictions in Industrial Engineering, Vol. 177, Issue 7, 2007,
1648- 1662.
126
[๐๐] B. Bede, Mathematics of Fuzzy Sets and Fuzzy Logic, Studies in
Fuzziness and soft computing, Vol. 295, 2013, 1-64.
[๐๐] S. Behiry, R. Abd-Elmonem, and A. Gomaa, Discrete Adomain
decomposition solution of nonlinear Fredholm integral equation,
Ainshams Engineering Journal, Vol. 1, Issue 1, 2010, 97-101.
[๐๐] Sh. Behzadi, Solving Fuzzy Nonlinaer Volterra Fredholm
Integral Equations by using Homotopy Analysis and Adomain
Decomposition Methods, Journal of Fuzzy Set Valued Analysis,
Vol. 2011, 2011, 1-13.
[17] S. Chang and L. Zadeh, On fuzzy mapping and control, IEEE
Transactions on Systems, Man, and Cybernetics, Vol. 2, Issue 1,
1972, 30-34.
[18] W. Cong-xin, and M. Ming, Embedding problem of fuzzy number
space: part 1, Fuzzy sets and systems, Vol. 44, Issue 1, 1991,
33-38.
[๐๐] M. Dehghan, and B. Hashemi, Iterative Solution of Fuzzy Linear
systems, Applied Mathematics and Computation, Vol. 175, 2006,
645-674.
[๐๐] P. Diamond, and P. Kloeden, Metric Spaces of fuzzy sets, Fuzzy
Sets and Systems, Vol. 100, Issue 1, 1999, 63-71.
127
[21 ] D. Dubois and H. Prade, Operations on fuzzy numbers,
International Journal of Systems Science, Vol. 9, Issue 6, 1978,
613- 626.
[๐๐] M. Farooq, S. Ullah, L. Ahmad, and S. Abdullah, Application of
fuzzy Laplace transforms for solving fuzzy partial Volterra
integro-differential equations, Cornell University Library, General
Mathematics, 2014, 1-11.
[23] M. Friedman, M. Ma and A. Kandel, Numerical solutions of
Fuzzy differential and integral equations, Fuzzy Sets and
Systems, Vol. 106, Issue 1, 1999, 35-48.
[๐๐] A. Gani, and S. Assarudeen, A new operation on triangular fuzzy
number for solving fuzzy linear programming problem, Applied
Mathematical Sciences, Vol. 6, No. 11, 2012, 525-532.
[๐๐] SH. Gao, Z. Zhang , and C. Cao, Multiplication Operation On
Fuzzy Number, Journal of Software, Vol. 4, No. 4, 2009, 331-
338.
[๐๐] M. Ghanbari, Numerical Solution of fuzzy linear Volterra
Integral Equations of The Second Kind, International Journal
of Industrial Mathematics, Vol.2, No.2, 2010, 73-87.
[๐๐] M. Ghanbari, Numerical solution of fuzzy linear Volterra
integral equations of the second kind by Homotopy Analysis
Method, Int. J. Industrial Mathematics, Vol.2, No.2, 2010, 73-87.
128
[๐๐] B. Ghazanfari, and R. Nazari, Two Steps Runge - kutta- like
Method for Numarical Solutions of Fuzzy Differential
Equations, International Research Journal of Applied and Basic
Sciences, Vol.4(10), 2013, 3209-3219.
[๐๐] M. Gholami, and B. Ghanbari, Variational Iteration Method for
Solving Voltarra and Fredholm Integral Equations of the Second
Kind, Vol.2, No. 1, 2011, 143-148.
[30] R. Goetschel and W. Voxman, Elementary calculus, Fuzzy Sets and
Systems, Vol. 18, Issue 1, 1986, 31-43.
[๐๐] E. Hetmaniok, T. Trawinski, B. Witula, and D. Slota, Usage of the
homotopy analysis method for solving the nonlinear and linear
equations of the second kind, Numerical Algorithms, Vol. 67,
Issue 1, 2014, 163- 185.
[๐๐] A. Jafarian, S. Measoomy Nia, and S. Tavan, A Numerical Scheme
to Solve Fuzzy Linear Volterra Integral Equations System, Journal
of Applied Mathematics, Vol. 2012, 2012, pages 1-17.
[33] O. Kaleva, Fuzzy differential equations, Fuzzy Set and Systems,
Fuzzy Numbers, Vol. 24, Issue 3, 1987, 301-317.
[34] S. Liao, Beyond Perturbation: introduction to the homotopy
analysis method, Modern Mechanics and Mathematics, Vol. 2,
2003, 1-336.
129
[๐๐] M. Matinfar, M. Saeidy, and J. Vahidi, Application of Homotopy
Analysis Method for Solving Systems of Volterra Integral
Equations, Advances in Applied Mathematics and Mechanics,
Vol.4, No.1, 2012, 36-45.
[๐๐] S. Measoomy Nia, A.Jafarian, and S. Tavan, A numerical scheme to
solve fuzzy Volterra integral equations system, Hindawi publishing
corporation journal of Applied Mathematics , vol. 2012, 2012,
1-17.
[๐๐] N. Mikaeilvand, S. Khakrangin, and T. Allahvirnaloo, Solving fuzzy
Volterra integro-differential equation by fuzzy differential
transform method, EUSFLAT, Advances in Intellingent Systems
Research, Vol. 2011, 2011, 891-896.
[๐๐] M. Mosleh and M. Otidi, Fuzzy Fredholm integro- differential
equations with artificial neural networks, Communications in
Numerical analysis, Vol. 2012 (2012), 1-13.
[39] S. Nanda, On integration of fuzzy mappings, Fuzzy Sets and
Systems, Vol. 32, 1989, 95-101.
[๐๐] S. Nasseri and G.Gholami, Linear System of Equations with
Trapizoidal Fuzzy Numbers, The Journal of Mathematics and
Computer Science, Vol.3, No.1, 2011, 71-79.
130
[๐๐] M. Paripour, T. Lotfi, and N. Najafi, Computing the fuzzy Laplace
transform via Homotopy perturbation Method, Journal of
Interpolation and Approximation in Scientific Computing,
Vol. 2014, 2014, 1-13.
[๐๐] I. Perfilieva, Fuzzy Functions: Theoretical and Practical Point of
View, University of Ostrava, Inst. for Research and Applications of
Fuzzy Modeling, 2011, 480-532.
[๐๐] A. Sadeghi, A. Ismail, and A. Jameel, Solving Systems of Fuzzy
Differential Equation, International Mathematical Forum, Vol.6,
No. 42, 2011, 2087-2100.
[๐๐] S. Salahshour, A. Ahmadian, N. Senu, D. Baleanu, and P. Agarwal,
On Analytical solutions of the Fractional Differential Equation
with uncertainty: Application to the Basset problem, Entropy,
Vol. 17, Issue 2, 2015, 885-902.
[๐๐] S. Salahshour, M.Khezerloo, S. Hajighasemi, and M. Khorasany,
Solving Fuzzy Integral Equations of the Second Kind by Fuzzy
Laplace Transform Method, Int. J. Industrial Mathematics, Vol. 4,
No. 1, 2012, 21-29.
[๐๐] H. Sorkun, and S. Yalcinbas, Approximate Solutions of linear
Volterra integral equation systems with variabl coefficients,
Applied Mathematical Modeling, Vol. 34, Issue 11, 2010, 3451-
3464.
131
[๐๐] Yu.Syau, L. Sugianto, E. Lee, Continuity and Semi continuity of
Fuzzy Mappings, Computers and Mathematics with Applications,
Vol. 61, Issue 4, 2011, 1122-1128.
[๐๐] L. Zadeh, Fuzzy Sets, Information and control, Vol. 8, 1965,
338-358.
[49] L. Zadeh, Linguistic variables, approximate reasoning and
disposition, Medical Informatics, Vol. 8, Issue 3, 1983, 173-186.
[๐๐] H. Zimmermann, Fuzzy set theory and applications, Springer
Science and Business Media, Vol.2, 2010, 317-332.
ุฌุงู ุนุฉ ุงููุฌุงุญ ุงููุทููุฉ
ูููุฉ ุงูุฏุฑุงุณุงุช ุงูุนููุง
ุญู ู ุนุงุฏูุฉ ูููุชูุฑุง ุงูุชูุงู ููุฉ ุงูุฎุทูุฉ ุงูุถุจุงุจูุฉ ู ู ุงูููุน ุงูุซุงูู ุจุงูุทุฑู ุงูุชุญููููุฉ ูุงูุนุฏุฏูุฉ
ุฅุนุฏุงุฏ ุฌููุงู ุชุญุณูู ุนุจุฏ ุงูุฑุญูู ุญู ุงูุฏู
ุฅุดุฑุงู ุฃ.ุฏ. ูุงุฌู ูุทูุงูู
ูุฏู ุช ูุฐู ุงุฃูุทุฑูุญุฉ ุงุณุชูู ุงุงู ูู ุชุทูุจุงุช ุงูุญุตูู ุนูู ุฏุฑุฌุฉ ุงูู ุงุฌุณุชูุฑ ูู ุงูุฑูุงุถูุงุช ุจูููุฉ .ููุณุทูู -ุงูุฏุฑุงุณุงุช ุงูุนููุง ูู ุฌุงู ุนุฉ ุงููุฌุงุญ ุงููุทููุฉุ ูุงุจูุณ
2016
ุจ
ุญู ู ุนุงุฏูุฉ ูููุชูุฑุง ุงูุชูุงู ููุฉ ุงูุฎุทูุฉ ุงูุถุจุงุจูุฉ ู ู ุงูููุน ุงูุซุงูู ุจุงูุทุฑู ุงูุชุญููููุฉ ูุงูุนุฏุฏูุฉ
ุฅุนุฏุงุฏ ุงูุฑุญูู ุญู ุงูุฏูุฌููุงู ุชุญุณูู ุนุจุฏ ุฅุดุฑุงู
ุฃ.ุฏ. ูุงุฌู ูุทูุงูู
ุงูู ูุฎุต
ู ู ุชุทุจููุงุช ุงูู ุนุงุฏุงูุช ุงูุชูุงู ููุฉ ุจุดูู ุนุงู ุชูุนุจ ุฏูุฑุง ูุงู ุง ุฌุฏุง ูู ุงูููุฏุณุฉ ูุงูุชูููููุฌูุง ูู ุง ููุงุงูุชุทุจููุงุช ู ุซู ุงูุชุญูู ูุงุณุนุฉ. ู ุนุงุฏุงูุช ูููุชูุฑุง ุงูุชูุงู ููุฉ ุงูุถุจุงุจูุฉ ุจุดูู ุฎุงุต ููุง ุงูุนุฏูุฏ ู ู
ุงูุถุจุงุจูุฉ. ุงูุชุญููู ูุงููุธู ุงุงููุชุตุงุฏูุฉุงูุถุจุงุจูุฉ ู
ุจุนุฏ ุฃู ุชูุงูููุง ุงูู ูุงููู ุงุฃูุณุงุณูุฉ ูู ุงูุฑูุงุถูุงุช ุงูุถุจุงุจูุฉุ ูู ูุง ุจุงูุชุฑููุฒ ุนูู ุงูุทุฑู ุงูุชุญููููุฉ ูุงูุนุฏุฏูุฉ ูุญู ู ุนุงุฏูุฉ ูููุชูุฑุง ุงูุชูุงู ููุฉ ุงูุถุจุงุจูุฉ ู ู ุงูููุน ุงูุซุงูู. ููุญู ู ุนุงุฏูุฉ ูููุชูุฑุง ุงูุชูุงู ููุฉ
ุงุจูุฉุ ุทุฑููุฉ ููู ูุชูุจู ุงูุชุญููููุฉ ุจุงูุทุฑู ุงูุชุญููููุฉ ูุฏู ูุง ุงูุทุฑู ุงูุชุงููุฉ: ุทุฑููุฉ ุชุญููู ุงูุจุงูุณ ุงูุถุจุงูุถุจุงุจูุฉุ ุทุฑููุฉ ุฃุฏูู ูุงู ุงูุชุญููููุฉ ุงูุถุจุงุจูุฉุ ุทุฑููุฉ ุงูุชุญููู ุงูุชูุงุถููุฉ ุงูุถุจุงุจูุฉุ ุทุฑููุฉ ุงูุชูุฑูุจ
ุงูู ุชุชุงูู ุงูุถุจุงุจูุฉ.
ููุญู ู ุนุงุฏูุฉ ูููุชูุฑุง ุงูุชูุงู ููุฉ ุจุงูุทุฑู ุงูุนุฏุฏูุฉ ูู ูุง ุจุชููุฐ ุทุฑู ู ุฎุชููุฉ ููู: ุทุฑููุฉ ุชุงููุฑ ุงูุชูุณุนูุฉุ .ุทุฑููุฉ ุชุจุงูู ุงูุชูุฑุงุฑุจู ุงูู ูุญุฑูุ ุทุฑููุฉ ุด
ุ ุญูุซ ุฃุธูุฑุช ุงููุชุงุฆุฌ ูุฉ ูู ูุง ุจุญู ุจุนุถ ุงุฃูู ุซูุฉ ุงูุนุฏุฏูุฉูููุชุญูู ู ู ููุงุกุฉ ูุฐู ุงูุทุฑู ุงูุนุฏุฏุงูุนุฏุฏูุฉ ุฏูุชูุง ููุฑุจูุง ู ู ุงููุชุงุฆุฌ ุงูุชุญููููุฉุ ููุงูุช ุทุฑููุฉ ุชุจุงูู ุงูุชูุฑุงุฑ ูู ุงุฃูููู ูุงุฃูุฏู ูู ุญู
ู ุน ุงูุทุฑู ุงูุนุฏุฏูุฉ ุงุฃูุฎุฑู.ู ุนุงุฏูุฉ ูููุชูุฑุง ุงูุชูุงู ููุฉ ุงูุถุจุงุจูุฉ ู ู ุงูููุน ุงูุซุงูู ุจุงูู ูุงุฑูุฉ
ุฌ