Beruflich Dokumente
Kultur Dokumente
ABSTRACT
system .This work is a review of four papers [1], [2], [3] and [10].
OVERVIEW
This work is a review of the papers [1], [2], [3] and [10].
Definitions and statement of results theorems are deal with in this chapter.
periodic solution.
periodic solution.
CHAPTER V deals with the illustration of some examples extension to
previous chapter.
CONTA
NT
CONTENTS
I REVIEW OF LITERATURE 1
System
Solution 6
1.3 Definitions 8
Solution 15
III Non-Linear Differential System in Periodic Solution 34
Conclusion 78
REFERENCES 79
REVIEW
REVIEW OF LITERATURE
In “Solving the periodic boundary value problem with the initial value
PRELIMINARIES
DIFFERENTIAL SYSTEM
the unknown element is a function, rather than a number, and in which the
it is called homogeneous if c ( x) ≡ 0,
b c
With constant coefficient functions p= and f =
a a
variables
dx
= p (t ) dt
x
In | x |= ∫ p (t ) dt +c
Exponentiating both sides, we obtain
x = ke ∫ , k = e c ------------------- (1.3)
p ( t ) dt
The linear differential equation (1.1) can be reduced to the normal form
variables
dy
= q (t ) dt
y
In | y |= ∫q (t ) dt +c
The linear differential equation (1.1) can be reduced to the normal form
we separate variables
dz
= [ r (t ) dt + s (t )dt + ξ (t )] + c
z
z = ke ∫ +ke ∫
r ( t ) dt s ( t ) dt
+ξ(t )
z =ke ∫
r ( t ) +s ( t ) dt
+ξ(t )
PERIODIC SOLUTION
x11 = f1 (t , x1 , x2 ..........
xn )
x 1 = f (t , x , x .......... xn )
2 2 1 2
..........
.......... ..........
.......... -------------------- (1.8)
..........
.......... ..........
..........
x1n = f n (t , x1, x2 ..........
..xn )
periodic in t,
6
∂f k
, k =1,2.... n, j =1,2,..... n , are continuous and bounded
∂x j
If f k (t , x1 , x2 ,..... x n ) = x k +1 , k=1, 2… n
……….., xn = xn' −1
∂f
Where f is continuous and 2π - periodic in t, ∂x , k =1,2.... n is continuous
k
and bounded.
x = x1, x2 = p (t ) x1' ,
( p (t ) x1' ) ' = x2' = f 2 (t , x1 , x2 ) = − f (t , x1 )
f1 (t , x1 , x 2 ) = p −1 (t ) x2 ,
f 2 (t , x1 , x2 ) = − f (t , x1 ).
1.3 DEFINITIONS
if for all x, y ∈H ,
We have
〈T ( x ), y 〉 = 〈x, T ( y )〉
such that F(x) =x. Fixed points arise naturally while solving operator
equations.
such that
Tx = λx , λ ≠ 0
diagonal are equal to one and all the other element are equal to zero is called
1.3.9 BOUNDED
for every x, y ∈A
is bounded in X.
10
transpose of X).
11
1.3.15 NON-SINGULAR
AA −1 = A−1 A = I
12
1 2 1 −2 1 0
A =
0 A−1 = I =
1
0
1
0
1
1.3.16 ORTHOGONALITY
12
y ' ( x ) = A( x) y ( x ) +b( x) ,
Then n × n matrix
13
CHAPTER II
CHAPTER II
Lemma 2.1
Statement
Then,
w=0.
Proof
If V is finite dimensional, there exists a basis for V which is
orthogonal to X with respect to H.
Suppose ω ∈V and H (u , ω) = 0 for all u ∈ PC 2 ( R )
14
0=H(x, x)-H(y, y)
So, w=0
Lemma 2.2
Statement
Assume
(A1) f ∈C 1 (RXR) is 2π - periodic with respect to the first variable, and
15
∂f
a≤ (t , x ) ≤ b On Rx R --------------- (2.1)
∂x
∂f
( t , x ) = f x , R 2 = RxR
∂x
m 2 M 2 < a ≤ b < ( m + 1) M 1
2
----------------- (2.2)
For h ∈ pc 2 ( R )
Proof
Define a bilinear form
2π
For u , v ∈ pc 2 ( R )
16
Put
∞
x = {X ∈ P (R C:) x = ∑ (cj c j o+ td j s j i) t n}
2
j = m+ 1
m
y = {Y ∈ P 2(R C:) y = C0 + ∑ (cj c j o+ td j s s j i) t n}
J= 1
Where cJ and dj are some constants, By Parseval’s formula
2π ' 2π n 2t
∫ ∑ ∫ ∑ µk ∫ f k (t ) dt
1
x (t ) x '
(t ) − ( x '
(t ), Bx (t )) dt = f k (t ) 2
dt − 2
0 k −1 0 k =1 0
∞
= ∑ ∑π (r 2 − µk )( ckr2 + d kr2 ) ≥ 0
k =1 r =N k +1
We obtain
2π 2π
12 2
H ( x, x ) = ∫ p(t ) x (t ) − h(t ) x (t )) dt ) ≥ ∫M x1 (t ) − bx 2 (t )) dt
2
1
0 0
(b ≥ M 1 (m + 1) 2 ;−b ≤ M 1 ( m + 1) 2 )
17
∞
x= ∑(c
j =m +1
j cos jt + d j sin jt )
∞
x1 = ∑(− jc
j =m +1
j sin jt + d j cos it ) 2
2 1 π
∞
2
H ( x, x ) = ∑ (j
j =m +1
2
c j . .4 + d j π − ∑c 2j (m + 1) 2 − d 2j ( m + 1) 2
2 2
∞
≥ ∑M
j =m +1
1 jc j (π / 4.4) + d j (π ) − c j ( m + 1) 2 + d j ( m + 1) 2 )
2 2 2 2
H(x, x) ≥ πM 1 ∑( j
J =m +1
2 2
− (m +1) 2 )( c j + d j )
2
2π
∫
2
H ( y, y ) = p (t ) y 1
(t ) − h (t ) y 2
(t ))dt
0
2π
∫ (M
2
≤ 2 y1 (t ) −by 2 (t )) dt
0
(b ≥ M 2 ;−b ≤ M 2 m 2 )
18
m
∴H ( y , y ) ≤ πM 2 ∑( j − m 2 )( c j + d j ) − 2πc0
2 2 2 2
j =1
Y.
Lemma (2.1) it follows that if bilinear form H (u, w) =0, for all u,
Then, w=0.
∴w=0
Theorem 2.3
Statement
Assume
(A1) f ∈C 1 (RXR) is 2π - periodic with respect to the first variable, and
∂f
a≤ (t, x) ≤ b On Rx R --------------- (2.7)
∂x
∂f
( t , x ) = f x , R 2 = RxR
∂x
( p (t ) x1 )1 + f (t , x ) = 0 ----------------- (2.9)
Proof
To Prove
( p(t ) x )
1 1
+ f (t , x) = 0 has a unique 2π - periodic solution.
To prove
Existence part
1
Let β = 2 ( M 2 m 2 + M 1 (m + 1) 2 )
( p(t ) x )
1 1
+ βx + λ(h(t , x) x + f (t ,0) − βx ) = 0
( p(t ) x )
1 1
+ (1 − λ) βx + λ( h(t , x ) x + f (t ,0)) = 0 λ∈[0,1] ------- (2.11)
21
Equation (2.11) becomes
dx
For the differential equation + βx = 0 , x = ce −βx
dt
Then,
xλ (t ) 1 ≤ M 0 ------------------- (2.12)
Such that
xj
1
→∞( j →∞) and xj is a solution of (2.11) for λ = λj Write
xj
yj =
| x j |1
22
Then
λj f ( t ,0 ) x j
( p(t ) y )j
1 1
+ (1 −λj ) βy j + λj h(t , x j ) y j +
xj
= 0 ---- (2.13)
1
By assumption ( A2 ), we have
M 2 m 2 < a ≤ h(t , x j (t )) ≤ b < M 1 (m +1) 2 -------- (2.14)
in C(s)].
λi → λ0 , y j → y 0 , y j → z 0 and
xj
1
→∞ as j →∞
23
λj f (t ,0)
There fore, →0
| x j |1
Hence,
( p (t ) y1 )1 + ((1 − λ0 ) β + λ0 h0 (t )) y = 0 has only a trivial 2π – periodic solution
24
To prove
Uniqueness part
Let x1 (t ) and x 2 (t ) be any two 2π - periodic solution of (2.9),
1
( p (t ) x ) + ∫ f x (t , x 2 (t ) + θx(t ) dθx = 0 ------------ (A)
1 1
0
1
a ≤ ∫ f x (t , x2 ct ) + θx(t )) dθ ≤ b
0
Assume that p (t) satisfies ( A1 ) and p (t) and h (t) satisfy ( A2 ) by replacing
fx by h, for h ∈PC (R )
25
I.e., x(t ) ≡ 0 on R
⇒ x1 (t ) = x2 (t )
Theorem 2.4
Statement
Assume
(A2) There exists two positive constant A and B and a nonnegative integer
m such that
n −1
∂f
0 < M 2 m 2 n + ∑ (−1) j −nα j m 2 j < A ≤ (t , x) ≤ B
j =1 ∂x
n −1
<M 1(m + 1) 2 n + ∑(−1) j −1α j (m + 1) 2 n
j =1
26
( p(t ) x )
n −1
( n) ( n)
+ ∑α j x ( 2 j ) + (−1) n +1 f (t , x ) = 0 ------------ (2.18)
j −1
has a unique 2π - periodic solution.(where (n) stands for the n-th derivative).
Proof
Let the assumption (A1) and (A2) hold.
To prove
( p(t ) x ) ∑α x
n −1
(n) (n)
j
(2 j)
+ (−1) n +1 f (t , x) + h(t , x, x1..... x ( 2 n −1) = 0 Let us
j =1
( p (t ) x ( n ) ) ( n ) + h(t , x) x + f ( t , x ) = 0
27
1
Let β = 2 ( M 2 m 2 n + M 1 (m +1) 2 n
dx
For, the differential equation dt + βx = 0 , x = ce −βx
28
Such that
xj
1
→∞( j →∞) and xj is a solution of (2.19) for λ = λj
xj
Write yj =
xj
1
Then
λj f (t ,0) x j
( p (t ) y j ) ( n ) + (1 − λ j ) β y j + λ j
(n) h(t , x j ) y j + = 0 -----
xj (2.21)
1
λj f (t ,0)
There fore, →0
xj
1
29
t
p (t ) z 0 (t ) = ( p0 ) z 0 (0) − ∫ (1 − λ0 ) β + λ0 h0 (t )) dt
0
( p (t ) y ( n ) ) ( n ) + (1 − λ0 ) β + λ0 h0 (t ) y = 0 --------------- (2.23)
To Prove
Uniqueness Part
Let x1 (t ) and x 2 (t ) any two 2π - periodic solutions for
2n-th order differential equation.
n −1
( p (t ) x ( n ) ) ( n ) + ∑α j x ( 2 j ) + ( −1) n +1 f ( t , x ) = 0 , x ∈ R →( A)
j =1
Where αj is constant
Then x (t) = x1(t) - x2(t) is a 2π - periodic solution of the following equation
30
n −1 1
( p (t ) x ( n ) ) ( n ) + ∑α j x ( 2 j ) + ( −1) n+1 ∫ f x (t , x2 +θx )dθx = 0
j =1 0
Assume that p (t) satisfies (A1) & p (t) & h (t), satisfy (A2) by replacing fx by
h, for h ∈PC 2 ( R )
2π - periodic solution
i.e., x(t ) ≡ 0 on R
n −1 n −1
0 < M 2 m 2 n + ∑(−1) j −n α j m 2 j < a ≤ f x (t , x) ≤ b < M 1 (m +1) 2 n + ∑( −1) j −nα J (m +1) 2 j
j =1 j =1
HAPTER III
CHAPTER III
Lemma 3.1
Statement
with respect to u.
Proof
Consider the system of ordinary differential equations
u (t ) + g r a d(uG(t ) )= p(t )
" in the form of equation
Lu + Nu = Fu .Where L is linear differential expression and N is a
32
H 1 ( y ) = [ L + N ' (u )] −1 , for u ∈ Rk
= M 1 || N (u + k ) − Nu − N ' (u ) k ||Y
→0 as || h ||Y → 0
[ L + N ' (u )] −1 : R K → R K
Hence H 1 ( y ) =[ L + N ' (u )] −1
Let y1 , y2 ∈Y , Then
1
H ( y1 ) − H ( y 2 ) = ∫ ( H ' ( y 2 + t ( y1 − y2 ))( y1 − y 2 ) dt
0
Where u (t ) = H ( y2 + t ( y1 − y2 ))
Where
||| H ( y1 ) − H ( y2 ) ||| ≤ M 1 || y1 − y2 ||
I.e., H is homeomorphism,
Lemma 3.2
Statement
If A is an nxn matrix and there exists δ > 0 such that | λ |≥ δ for all
eigenvalues λ of A.
Then || A −1 ||≤ δ −n || A || n −1
−−−
[Where ||A|| =max ( A* A) A = (a ji ) here A* = (aij ) if A= (aij)
34
Proof
Since || A * Ae i || =|| µi ei ||
We have
35
µ k 0 || A * A ||n −1 ≥| λ1 |2 | λ 2 |2 .... | λ n |2 ≥ δ 2 n
δ −n || A || ( n−1) ≥ ( µk0 ) −1
δ −n || A || ( n −1) ≥|| A || −1
(i.e.) || A −1 ||≤ δ −n || A || n −1
Theorem 3.3
Statement
∂f k
, k =1,2,...... n, j =1,2,..... n are continuous and bounded, For any
∂z j
36
2π
∂f i
Where aij ( x ) = ∫ (t , x1 (t ), x 2 (t )....... x n (t )) dt , where i, j=1, 2…., n.
0
∂x j
Such that | RP (λ) |≥ δ where RP (Z ) is the real part of Z. and for those {λk }
x11 = f1 (t , x1 , x2 ..........
xn )
x 1 = f (t , x , x .......... xn )
2 2 1 2
..........
.......... ..........
.......... --------------------- (3.3)
..........
.......... ..........
..........
x1n = f n (t , x1, x2 ..........
..xn )
has a unique 2π - periodic solution.
37
Proof
Define x = ( x1 , x 2 ,....... x n ) ∈ R n
defined on R × R n
equation of u 1 = F (t , u ) , u ( 0) = u 0
∂F (t , x )
y1 = y ----------------------- (3.4)
∂x
38
∂F (t , x ) ∂f
y1 = = k (t , x ) ---------------------- (3.5)
∂x
∂x j
∂F
derivatives ∂u
(t , u ).
Let the solution u(t)=u(t,u0) of u1=F(t,u), u(0)=u0 exists for t ∈[0,2π] and let
∂F ∂u (t , u 0 )
H (t , u 0 ) = (t , u (t ,u 0 ) function ϕ(t , u 0 ) = exists and is the
∂u ∂u 0
∂x
solution of w1 = H (t , u ) , w matrix function w(t)= ∂u
(t , u ).
∂x1 (t , u ) ∂x1 (t , u ) ∂x1 (t , u )
∂α1 ∂α 2 ∂α n
∂x (t , u ) ∂x2 (t , u ) ∂x2 (t , u )
= 2 ---------------------- (3.6)
∂α1 ∂α 2 ∂α n
∂xn (t , u ) ∂xn (t , u ) ∂xn (t , u )
∂α ∂α 2 ∂α n
1
39
Then, we have
H 1 (u ) = I n − h1 (u )
∂x
= In − ( 2π , u )
∂u
T
2π
∂f (t , x), (t , u ) T
= I n − exp ∫0
∂x dt
40
λ (u )
The eigenvalues µk (u ) of H 1 (u ) are µ k (u ) = 1 − e , k = 1,2,...n
k
arg λ
Since λk (u ) =| λk (u ) | e k (u )i
, k=1,2,…,n.
λk (u ) =| λk (u ) | {cos θk (u ) + i sin θk (u )}
We have
|λ k(u) |
µ k u)( 1−= e { θ k u)( c+ is θ ko(u)i n}s
|λ k ( u )|c θ k u | o s
µ (u) (1−= e ) θ ({)+ iu s θ (u)c+ i |s }λ n(u)o|is θ (un) is n
k k k k k
Squaring on both sides, we get,
| µ k (u ) |2 = (1 − e|λk ( u )| cos θ k (u )) 2
[1 − cos | λk (u ) | sin θk (u )]
41
1λ ( y ) | sin θ k (u )
sin θ
2
2
2
| λ (k ) sin θ k (u )
= (1 − e1λk (u )|cos θk (u ) ) 2 + 4e |λk u|cos θk ( u ) ) sin 2 k
2
----------------- (3.9)
| λk u | sin θ k (u )
| µk (u ) | 2 = (1 − 1) 2 + 4e |λk u|0 sin 2 ( ) (Where k=1, 2.., n)
2
| λ (u ) |
| µ k (u ) | 2 = 4 sin 2 k
2
a b
≥ 4 minsin 2 k sin 2 k =: δ k > 0
2 2
42
| µ j (u ) |2 ≥ min (1 − e|λ k ( u )|cosθ k ( u ) ) 2 =: ε j > 0
|λ j ( u )|≥ δ
Let δ0 = min{ δk ,ε j }
∂f k
Since (t , x ) are bounded
∂x j
We have
Then if follows from lemma (3.2) that the mapping H 1 (u ) is invertible and
| µk (u ) |≥ δ −n || H 1 (u ) || C 0n −1
ie , || [ H 1 (u )] −1 || ≤ M
( M = C 0N −1 / δ N )
Which means that there exists a unique u*= (α1* , α2* ,......... ....., αn* ) ∈ R n
43
H (u * ) = u * − h(u * )
0 = u * − h(u * )
u * = h(u * )
⇒ h(u * ) = u *
x(t , u * ) = ( x1 (t , α 1 ), x 2 (t , α 2 ),..... x n (t , α n ),
* * *
1
x1 = f 1 (t , x1 , x 2 ,..... x n )
1
x 2 = f 2 (t , x1 , x 2 ,..... x n )
……..
1
x n = f n (t , x1 , x 2 ,..... x n )
periodic in t,
∂f k
, k =1,2.... n, j =1,2,..... n are continuous and bounded
∂x j
44
Lemma 3.4
Statement
∂f
partial derivatives ∂u (t , u ) .
Assume the solution u (t ) = u (t , u0 ) of (3.10) exists for t ∈[0,2π] and
∂F
Let H (t , u 0 ) = ∂u
(t , u (t , u 0 ))
∂u (t , u0 )
ϕ(t , u0 ) = ------------------ (3.11) exists and is the solution of
∂u0
w1 = H (t , u 0 )W − − − − − − − − − − − −(3.12 )
Such that
Proof
solution is defined on R × R n
45
equation of u 1 = F (t , u ) , u ( 0) = u 0
∂F (t , x )
y1 = y ----------------------- (3.13)
∂x
∂F (t , x ) ∂f
y1 = = k (t , x ) ---------------------- (3.14)
∂x ∂x
j
∂F
derivatives ∂u
(t , u ).
Thus the conditons are unique solution has proved on theorem (3.3)
∂u (t , u 0 )
Then the function ϕ(t , u0 ) is the unit matrix, ϕ(t , u0 ) = ∂u 0
, t ∈(0,2π )
∂u (0, u 0 )
ϕ(t , u 0 ) = exists and is the solution of (3.12) and satisfies the
∂u 0
∂u (t , u 0 )
relation = −ϕ(t , u 0 ) f (t , u 0 ) , t ∈(0,2π)
∂u 0
46
C
HAPTER IV
CHAPTER IV
COROLLARY 4.1
STATEMENT
2π
∂f
Where d k ( x) = (2π) ∫ ∂x
n −k
(t , x(t ), x ' (t ),......, x ( n −1) (t )) dt ,
0 k
{ a k }, { bk } such that,
π -periodic solution.
47
Proof
1
x 2 = f 2 (t , x1 , x 2 ,..... x n )
……..
1
xn = f n (t , x1 , x2 ,..... xn ) for
x = x1, x2 = p (t ) x1' ,
( p (t ) x1' ) ' = x2' = f 2 (t , x1 , x2 ) = − f (t , x1 )
f1 (t , x1 , x 2 ) = p −1 (t ) x2 ,
f 2 (t , x1 , x2 ) = − f (t , x1 ).
In this case, Zeros λ of the polynomial f (λ)
2π 2π
dt ∂f
| λI 2 − A( x) |= λ2 + ∫ .∫ (t , x(t )) dt
0
p (t ) 0 ∂x
48
2π 2π
dt ∂f
are pure by imaginary with λ1, 2 = ± ∫
0
.∫
p (t ) 0 ∂x
(t , x (t )) dt
a b
⇒ 2πm < 2π ≤| λ1, 2 |≤ 2π < 2(m + 1) are satisfied
m2 m2
equation
n −1
( p (t ) x ( n ) ) ( n ) + ∑α j x ( 2 j ) + (−1) n +1 f (t , x) = 0 ----------------- (4.4)
j =1
x n +2 = ( p (t ) x n +1 ) ' ,
x n +3 = x n' +2 ,
……………….
x 2 n = x 2' n −1 .
49
x1' = x 2 , x 2' = x3 ,
……………
xn = xn' +1
x n' +2 = p −1 (t ) x n +2 − p −1 (t ) p ' (t ) x n +1 ,
xn' +2 = xn +3
……………
x 2' n −1 = x 2 n ,
n −1
x2' n = −∑α j x ( 2 j ) + ( −1) n f (t , x1 )
j +1
If n is even, and
x ( 2 j ) = x 2 j +1 , for j ≠k,
50
If n=2k-1 is odd
2π
∂f
∫ ∂x (t , x(t )) dt
0
2π
∫ p (t ) / p(t )dt =0
'
Since
0
In this case
2π 2π
dt ∂f
| λI 2 − A( x ) |= λ + ( −1) ∫ .∫
2n n +1
(t , x(t )) dt
0
p (t ) 0 ∂x
If the values of α1 , α2 ,......., αn −1 are not all equal to zero, the eigenvalues
51
Then, we have
H 1 (u ) = I n − h1 (u ) ---------------------- (4.6)
∂f
= In − (t , x(t ))
∂xk
T
2π
∂f
= I n − exp (2π ) ∫
n −k
(t , x(t ), x ' (t )... x n−1 (t ) dt
T
∂xk
0
where k=1,2,…,n-1
2π
∂f
Where d k ( x) = ( 2π ) ∫ ∂x (t , x(t ), x (t )... x (t )) dt
n −1 ' n −1
k= 1…n-1
0 k
d k ( x ).
λ
are d k ( x) = 1 − e , k = 1,2,...n
(u )
The eigenvalues d k ( x ) of H 1 (u ) k
arg λk ( u ) i
Since λk (u ) =| λk (u ) | e , k=1, 2… n.
λk (u ) =| λk (u ) | {cos θk (u ) + i sin θk (u )}
52
We have
| λ k ( u )|
dk x)( 1−= e { θ k u)( c+ is θ ko(ui) n}s
| λ k ( u )|c θ k u | o s
k(xd ) (1−= e ) θ ( {)+ iu s θ (u) c+ i |s }λ n(uo)|si θ (u)n is n
k k k k
Squaring on both sides, we get,
| d k ( x) |2 = (1 − e|λk ( u )| cos θk (u )) 2
[1 − cos | λk (u ) | sin θk (u )]
1λ ( y ) | sin θ k (u )
sin θ
2
2
2
| λ (k ) sin θ k (u )
= (1 − e1λk (u )|cos θk (u ) ) 2 + 4e |λk u|cos θk ( u ) ) sin 2 k
2
----------------- (4.7)
53
| λk u | sin θ k (u )
| d k ( x) |2 = (1 − 1) 2 + 4e|λk u|0 sin 2 ( ) (Where k=1, 2.., n)
2
| λ (u ) |
| d k ( x) |2 = 4 sin 2 k
2
a b
≥ 4 minsin 2 k sin 2 k =: δ k > 0
2 2
Let δ0 = min{ δk ,ε j }
∂f
Since ∂xk
(t , x (t )) are bounded
54
We have
Then if follows from lemma (3.2) that the mapping H 1 (u ) is invertible and
| d k ( x) |≥ δ −n || H 1 (u ) || C0n−1
ie , || [ H 1 (u )] −1 || ≤ M
( M = C 0N −1 / δ N )
H (u * ) = u * − h(u * )
0 = u * − h(u * )
u * = h(u * )
⇒ h(u * ) = u *
55
of
1
x + f1 (t , x, x1 ,..., x n −1 ) = 0
2
x + f1 (t , x, x1 ,..., x n−1 ) = 0
……………………….......
n
x + f1 (t , x, x1 ,..., x n −1 ) = 0
∂f
, k =1,2.... n are continuous and bounded.
∂xk
COROLLARY 4.2
STATEMENT
56
∂f
f is continuous and 2π - periodic in t, ∂t
(t , x ) is continuous and bounded
for all (t , x ) ∈R ×R
n −1 2π
p0 = π ∫p
−1
Denote k =1
k (t ) dt
0
Then, equation (1) has a unique 2π -periodic solution provided one of the
2π
∂f
| ∫ ∂x (t , x(t )dt ) | p
0
0 ≥ δ1 ;
( A2 ) n=2p, p ∈N and for any continuous function x (t), there exists
2π
∂f
(2m1π ) < a ≤ (−1) ∫ ∂x (t , x(t )) dt . p ≤ b < (2(m1 +1)π ) n
n p −1
0
0
57
2π
∂f
constant δ 2 > 0 such that (−1) ∫ ∂x (t , x(t )) dt . p ≥ δ2
q
0
0
Proof
Let x1 (t ) = x,
x 2 (t ) = p1 (t ) x1' (t )
--------------------------
--------------------------
x n (t ) = p n −1 (t ) x n' −1 (t ),
1
x1 = f 1 (t , x1 , x 2 ,..... x n )
1
x 2 = f 2 (t , x1 , x 2 ,..... x n )
---------------------------
1
x n = f n (t , x1 , x 2 ,..... x n )
58
In this case,
λ k (u ) =| G (u ) |1 / n eθ k ( u )i , where k=1, 2… n
( 2k −1)π
Where θk (u ) = θk = , ifG (u ) > 0.
n
( 2k −1)π
, ifG (u ) < 0. Since n is odd
n
( A1 ) Let us verify that cos θk ≠0, k=1, 2… n and it follows from the
We have
1/ n
cosθ k 2
minn | µ K (u )) |2 ≥ min (1 − e|G (u )| ) =: δ k > 0
u∈R |G ( u )|≥δ 1
59
( A2 ) We prove the case that p is an odd number only.
mK = m1
We have G (u ) ≥ a > 0
| λk u | sin θ k (u )
| µk (u ) | 2 = (1 − 1) 2 + 4e |λk u|0 sin 2 ( ) (Where k=1, 2.., n)
2
60
| λ (u ) |
| µ k (u ) | 2 = 4 sin 2 k
2
a b
≥ 4 minsin 2 k sin 2 k =: δ k > 0
2 2
On the other hand, for those j where RP λj (u ) ≠ 0
Let δ2 = min{ δ K }
∂f
f is continuous and 2π - periodic in t, ∂t
(t , x ) is continuous and bounded
61
CHAPTE
RV
CHAPTER V
ILLUSTRATION
Example (1)
Where
− (1 + 2 sin t )
p (t ) = , q (t ) = −1 +2 cos t , r (t ) = 4 + 2 sin 2 t +1,
2
1
s (t ) = 3 + 4 sin t , ξ(t ) = −5 cos t ,
2
Let
dx
= p (t ) y + f (t )
dt
dy
= q (t ) z + g (t ) --------------------- (5.2)
dt
dz
= r (t ) x + s (t ) y +ξ(t ) z + h(t )
dt
62
− (1 + 2 sin t )
Here f1 = 0 f 2 = p (t ) = f3 =0
2
2π 2π 2π
∂f ∂f 2 − (1 + 2 sin t )
a11 = ∫ 1 =0 , ∫ = ∫ dt
0
∂x 0
∂y 0
2
2π
− (1 + 2 sin t )
a12 = ∫ dt
0
2
1
= ( t + 2 cos t ) 02π
2
2π 2π
∂f ∂f
a12 = ∫ 2 = −π , a13 = ∫ ∂x3 = 0
0
∂y 0
2π
∂f i
Second row, we have where aij ( x ) = ∫ (t , x (t , u )) dt
0
∂x j
Here f1 = 0 f2 = 0 f 3 = q (t ) = −1 + 2 cos t
63
2π 2π
∂f ∂f
a 21 = ∫ 1 =0 , a 22 = ∫ ∂x2 = 0
0
∂x 0
2π 2π
∂f 3
a 23 = ∫ = ∫ ( −1 + 2 cos t ) dt
0
∂x 0
2π
a23 = ∫ (−1 + 2 cos t )dt
0
2π
∂f 3
a 23 = ∫ ∂x
0
= −2π
1
f 3 =ξ (t ) = − 5 cos t
2
2π 2π
∂f1
a31 = ∫ = ∫ ( 4 sin 2 t +1) dt
0
∂x 0
64
2π
69
2π
4t cos 2 t
=
+t
2 0
2π
∂f1
a31 = ∫ ∂x
0
= 6π
2π 2π
∂f 2
a32 = ∫0 ∂x = ∫0 (3 + sin t )dt
2π
= ( 3t + 4 cos t ) 0
2π
2π
∂f 2
a32 = ∫ ∂x
0
= 6π
2π
∂f 3 2π −1
a33 = ∫0 ∂x = ∫0 2 − 5 cos t dt
65
2π
−1
i.e. a33 = ∫ 2 − 5 cos t dt
0
2π
−t
= − 5 sin t
2 0
2π
∂f 3
a33 = ∫ ∂x
0
=π
The matrix aij ( x) = A
0 −π 0
⇒ A( x ) = A is a constant matrix A(x) =A= 0 0 − 2π
6π 6π π
λ −π λ
We have, | A − λI |= λ λ − 2π = 0
6π 6π π
66
i.e. λ1=π λ2 ,3 = ± 12 πi
substituting
2π
∂f i
Where aij ( x ) = ∫ (t , x1 (t ), x 2 (t )....... x n (t )) dt , where i, j=1, 2…., n.
0
∂x j
∴By theorem the linear system of equations must have has unique 2π -
periodic solution.
67
Example (2)
2π
dθ
(By using the result I = ∫ a + b sin 2 θ = π ) and
0
2π 2π
∂f
4π < p0 ∫
2
(t , x)dt = π ∫ (4 − 9 sin t )dt = 8π 2 < 9π 2
0
∂x 0
such that
2π
∂f
( 2m1π ) n < a ≤ ( −1) p −1 ∫ (t , x (t )) dt . p0 ≤ b < ( 2(m1 +1)π ) n
0
∂x
-------------------- (5.6)
68
2π
∂f
4π < a ≤ ( −1) ∫ ∂x (t , x(t )) dt . p ≤ b < 9π 2
2 p −1
0
0
2π
∂f ∂f
∂x
= p0 ∫ ∂x (t , x)dt
0
2π
= π ∫ ( 4 − 9 sin t ) dt
0
= π ( 4t − 9 cos t ) 0
2π
= π ( 4( 2π ) )
∂f
= 8π 2 < 9π 2
∂x
But using the condition of theorem 2.3 there exists a non- negative integer m
m 2 M 2 < a ≤ b < ( m + 1) M 1
2
are not satisfied
EXAMPLE (3)
2π
∂f
∫ ∂x (t , x) = (m + 2δ )(1 + 5 sin t ) + δ sin t. cos x
0
2π
∂f
∂x
(t , x ) = ∫ (( m + 2δ )(1 + 5 sin t ) + δ sin t. cos x)dt
0
= ( (m + 2δ )( 2π ) + δ cos x. cos t ) 0
2π
2π
∂f
∫ ∂x (t , x) = 2(m + 2δ )π
0
2π
∂f
2mπ < 2( m + δ )π < ∫ ∂x (t , x(t )) dt < 2(m + 2δ )π < 2(m +1)π
0
1
< 2( m + 3δ )π < 2( m +1)π For 0 <δ <
3
∴Corollary (4.2) implies that the equation (5.7) has a unique 2π - Periodic
solution
70
CONCLUSL
ON
CONCLUSION
and its applications lemma some examples are studied in details. The 2π -
and n-th order differential system are studied and its results are deduced.
71
REFERENC
E
REFERENCES