Professional Documents
Culture Documents
INTEGRAL EQUATIONS
Chapter 7 INTEGRAL EQUATIONS
7.6 Connection between integral equations and initial and boundary value
problems
7.7 Exercises
Future Topics:
7.7 Fixed point theorem (see [Hochstadt Integral equations, p.25]) (added in 7.2)
Elementary existence theorems
7.8 Practical applications (see [Jerri Introduction to Integral Equations with Applications])
1. Euclidian vector space \ n The n-dimensional Euclidian vector space consists of all points
{
\ n = x = ( x1 ,x2 ,...,xn ) xk \ }
for which the following operations are defined:
2. Vector space C G( ) ( )
Vector space C G consists of all real valued continuous functions defined on
the closed domain G \ n :
( ) {
C G = f ( x ) : D \ n \ continuous }
Norm f C
= max f ( x )
xG
( )
C G is a complete vector space (Banach space) relative to defined norm f C
.
3. Vector space L2 ( G ) The space of functions integrable according to Lebesgue (see Section 3.1)
L2 ( G ) = f ( x ) : G \ n ^ f ( x)
2
dx <
G
Inner product ( f ,g ) = f ( x ) g ( x )dx
G
( f , f ) = f ( x)
2
Norm f 2
= dx
G
The following property follows from the definition of the Lebesgue integral
f (x )dx f (x )dx
G G
( f ,g ) f 2
g 2
for all f ,g L2 ( G )
Proof:
If f ,g L2 ( G ) , then functions f , g and any combination f + g are
also integrable and therefore belong to L 2 (G ) .
Consider
0 ( f + g ) dx = f dx + 2 fg dx + 2 g dx
2 2 2
G G G G
The right hand side is a quadratic function of . Because this function is non-
negative, its discrimenant ( D = b 2 4ac ) is non-positive
2
2 2
4 fg dx 4 f dx g dx 0
G G G
2
2 2
fg dx f dx g dx
G
G G
( f ,g)
2 2 2
f 2
g 2
( f ,g) f 2
g 2
5. Minkowski Inequality (3rd property of the norm Triangle Inequality), (see Example 10.7 on p.257)
f +g 2
f 2
+ g 2
for all f ,g L2 ( G )
Proof:
= ( f + g, f + g)
2
Consider f +g 2
= ( f , f ) + ( f , g ) + (g , f ) + (g , g )
+ ( f , g ) + (g , f ) + g
2 2
f 2 2
2 2
f 2
+2 f 2
g 2
+ g 2
from C-B inequality
= f ( 2
+ g 2
) 2
Note that the Minkowski inequality reduces to equality only if functions f and
g are equal up to the scalar multiple, f = g , R (why?).
Chapter 7 INTEGRAL EQUATIONS
L:M N
Proof:
Let operator L : M N be bounded, then according to the definition there
exists c > 0 such that Lf N c f M .
the limit it follows that for any > 0 there exists k N such that
fk f M
< for all k k .
such that Lf k Lf N
< for all k K .
Choose = , then
c
Lf k Lf N
= L( f k f ) N
c fk f M
< c = for all k k c = K .
c
Lf Lg k f g for all f ,g M
f0 S is a starting point
f1 = Lf0
f 2 = Lf1
#
f n +1 = Lf n
#
But they do not always converge to the fixed point of operator equation. This
example shows that even the choice of the starting point close to the fixed point
yields the divergent sequence of successive approximations (apparently they are
not very successive / ):
Assume for n : f n f n + 1 k n f 0 f1
Consider
fm fn f m Lf m + Lf m Lf n + Lf n f n
f m Lf m + k f m f n + Lf n f n Lipschitz condition
f m Lf m + Lf n f n
1 k
f m f m +1 + f n f n +1
definition of s.a.
1 k
km f 0 f1 + k n f0 f1
equation ()
1 k
km + kn
= f0 f 1 0 when m,n
1 k
Chapter 7 INTEGRAL EQUATIONS
(
lim f n +1 = L lim f n
n n
)
converges to
f = Lf
And therefore, f S is a fixed point.
Hugo Steinhaus, the colleague and friend of Stefan Banach, formulated the fixed
point theorem in the following way:
Kf = K (x , y ) f ( y )dy xG Rn
G
K (x , y )
2
dxdy <
GG
1) K : L 2 (a , b ) C [a , b] Kf C
M ba f 2
for f L 2 (a , b )
2) K : L 2 (a , b ) L2 (a , b ) Kf 2
M (b a ) f 2
for f L 2 (a , b )
3) K : C [a , b ] C [a , b] Kf C
M (b a ) f C
for f C [a , b]
Proof:
Since K (x , y ) is continuous in the closed domain [a , b ] [a , b ] , there exists
M > 0 such that M = max K (x , y ) .
x , y[a ,b ]
b
= max
x[ a ,b ] K ( x, y ) f ( y )dy definition of integral operator
a
= max
x[ a ,b ]
( K ( x, y ) , f ( y ) ) inner product in L2 ( a,b )
12
b
K ( x, y ) dy definition of norm in L2 ( a,b )
2
2 x[ a ,b ]
f max
a
12
b
f max M 2 dy replace by M = max K (x , y )
2 x[ a ,b ] x , y[a ,b ]
a
= M ba f 2
calculating definite integral
Chapter 7 INTEGRAL EQUATIONS
12
b
= (Kf )(x ) dx inner product in L2 ( a,b )
2
12
b b 2
= K (x , y ) f ( y )dy dx definition of integral operator
a a
12
b 2
2
K 2
f 2 dx Cauchy-Bunyakowski inequality
a
12
b b
= f 2 K (x , y ) dy dx
2
factoring f 2
a a
12
b b
f 2 M 2 dy dx replace by M = max K (x , y )
x , y[a ,b ]
a a
12
b b
M f 2 dy dx calculating definite integral
a a
=M f 2
(b a )
b
= max
x[ a ,b ] K ( x, y ) f ( y )dy definition of integral operator
a
b
max K ( x, y ) f ( y ) dy
x[ a ,b]
a
b
max M f ( y ) dy replace by M = max K (x , y )
x[ a ,b] x , y[a ,b ]
a
b
= M f ( y ) dy does not depend on x
a
b
M max f ( y ) dy
y[ a ,b ]
a
b
M f C
dy definition of norm in C [ a,b ]
a
b
C
M f dy
a
= M (b a ) f C
calculating definite integral
Chapter 7 INTEGRAL EQUATIONS
Ku = K ( x, y ) u ( y ) dy
a
K ( x, y ) u ( y ) dy = f ( x ) Ku = f non-homogeneous eqn
a
b
K ( x, y ) u ( y ) dy = 0 Ku = 0 homogeneous eqn
a
b
u ( x ) = K ( x, y ) u ( y ) dy + f ( x ) u = Ku + f non-homogeneous eqn
a
b
u ( x ) = K ( x, y ) u ( y ) dy u = Ku homogeneous eqn
a
K (x , y )u( y )dy = f (x )
0
x
u (x ) = K (x , y )u ( y )dy + f (x )
0
0 a x
Chapter 7 INTEGRAL EQUATIONS
III Integro-Differential Equation includes an unknown function under the integral sign and also any derivative of
the unknown function. For example:
du
= u (x ) + K (x , y )u ( y )dy + f (x )
dx G
An important representation of the integro-differential equation is a Radiative
Transfer Equation describing energy transport in the absorbing, emitting and
scattering media (analogous equations appear in the theory of neutron transport).
u = Ku + f non-homogeneous equation
u = Ku homogeneous equation
The value of the parameter C for which the homogeneous integral equation
has a non-trivial solution u L 2 which is called an eigenvalue of the kernel
K (x , y ) , and the corresponding solution is called an eigenfunction of this
kernel.
Eigenvalue problem We will distinguish eigenvalue problems for the integral kernel (integral
equation):
u = Ku
u = Ku + f
the following iterations of the method of successive approximations are set by:
u 0 (x ) = f (x )
u n (x ) = Ku n 1 + f n = 1,2 ,...
n
Lemma 7.1 u n (x ) = k K k f K (
where K k = K ("
K ))
k =0
k times
n=0 u 0 (x ) = 0 K 0 f = f ( x ) confirmed
n = n+1 un +1 ( x ) = Ku n + f by definition
n
= K k K k f + f by assumption
k =0
n
= f + k +1 K k +1 f linearity
k =0
n +1
= f + p K p f change of index p = k + 1
p =1
n +1
= 0 K 0 f + p K p f
p =1
n +1
= p K p f
p =0
n +1
= k K k f change of index p = k
k =0
Neumann Series k K k f is called to be the Neumann Series
k =0
Estimation of iterations Kn f
C
(
= K K n 1 f ) C
M 2 (b a ) K n 2 f
2
C
M n (b a ) f
n
C
Chapter 7 INTEGRAL EQUATIONS
k K k f k Kk f Cauchy-Bunyakovsky inequality
C
k =0 C k =0
M k (b a )
k k
f C
Theorem 7.2 3)
k =0
[ M (b a )]
k
= f C
geometric series
k =0
1
converges if <
M (b a )
f
= C
1 M (b a )
1
Therefore, the Neumann series k K k f converges for <
M (b a )
.
k =0
= K lim u n 1 (x ) + f
n
b
= K (x , y ) lim u n 1 ( y )dy + f
n
a
b
= K (x , y )u ( y )dy + f
a
f
And , recalling estimation, u (x ) C C
1 M (b a )
show that this solution is unique. For that, it is enough to show that the
homogeneous equation u = Ku has only a trivial solution. Indeed, if
u 0 = Ku 0 , then u 0 C [a , b] and , according to Theorem 6.2 3),
u0 C
M (b a ) u 0 C
, then
[1 M (b a )] u 0 C 0
Because <
1
M (b a )
[ ]
, 1 M (b a ) > 0 and, therefore, u 0 C = 0 . That
yields, that u (x ) = 0 for all x [a , b] . So, only the trivial solution exists for the
homogeneous equation.
1
, then there exists an inverse operator (I K ) .
1
Therefore, if <
M (b a )
Conditions of Theorem 7.3 are only just sufficient conditions; if these conditions
are not satisfied, solution of the integral equation still can exists and the
Neumann series can be convergent.
Identify: K (x , y ) = 1 f (x ) = e x ba =1
1
M =1 =
e
1 1 1
Check condition: < < <1
M (b a ) e 11
1) iterations:
u 0 (x ) = ex
u 1 (x ) = ex +
11
e 0
u 0 ( y )dy = ex +
11 x
e 0
e dy = ex +
e
[ ]
1 x
e
1
0 = ex + 1
1
e
11 1 1 x 1 1
u 2 (x ) u 1 ( y )dy
e 0 e 0
= ex + = ex + e + 1 dy = e + 1 2
x
e e
11 1
u n (x ) u n 1 ( y )dy
e 0
= ex + = ex + 1
en
This result can be validated by a direct substitution into the given integral
equation.
Chapter 7 INTEGRAL EQUATIONS
2) Neumann series:
u (x ) = k K k f = f ( x ) + 1 K 1 f + 2 K 2 f + "
k =0
f (x ) = ex
1
Kf = e x dy = e 1
0
1
K2 f = (e 1)dy = e 1
0
Kn f = e 1
1
u (x ) = ex + (e 1) + 12 (e 1) + " + 1n (e 1) + "
e e e
1 1 1
= e x ( e 1) + ( e 1) + ( e 1) + 2 ( e 1) + " + n ( e 1 ) + "
e e e
1
= e x (e 1) + (e 1) n
n =0 e
= ex e +1+
(e 1)
1
1
e
= ex e + 1+ e
= ex +1
2. The Method of Successive Substitutions for Fredholms Integral Equation (the Resolvent Method)
Iterated kernel Let integral operator K has a continuous kernel K (x , y ) , then define:
Repeated operator ( ) (
K n = K K n 1 = K n 1 K ) n = 2 ,3 ,...
Indeed, ( K f )( x)
1
= K ( x, y ) f ( y ) dy
G
K1 ( x ,y )
(K f )(x )
2
= [K (Kf )](x )
= K (x , y ) K ( y , y ) f ( y )dy dy
G G
= K ( x, y ) K ( y , y )dy f ( y ) dy
G G
K 2 ( x ,y )
Kernel K n (x , y ) = K (x , y )K n 1 ( y , y )dy
G
= K n 1 (x , y )K ( y , y )dy
G
b
u (x ) = f (x ) + R(x , y , ) f ( y )dy
a
23
Identify: K (x , y ) = xy f (x ) = x ba =1
6
1
M =1 =
8
1 1 1
Check condition: < < <1
M (b a ) 8 11
Iterated kernels:
K 1 (x , y ) = xy
1
1 1
y3 xy
K 2 (x , y ) = K ( x, y )K1 ( y , y ) dy = xy y ydy = xy =
0 0 3 0 3
1
1 1
xy xy y 3 xy
K 3 (x , y ) = K 2 ( x, y )K 2 ( y , y ) dy = y ydy = =
0 0 3 3 3 0 32
xy
K n (x , y ) =
3 n 1
Resolvent:
R(x , y , ) = k K k +1 (x , y )
k =0
1 xy 1 xy 1 xy 1 xy
= xy + + 2 2 + 3 3 + ... + n n + ...
8 3 8 3 8 3 8 3
11 1 1 1 1 1 1
= xy 1 + + 2 2 + 3 3 + ... + n n + ...
8 3 8 3 8 3 8 3
1
= xy
1
1
24
24
= xy
23
Solution:
b
u (x ) = f (x ) + R(x , y , ) f ( y )dy
a
23 1 1 24 23
= x + xy ydy
6 8 0 23 6
23 1 1
= x + x y 2 dy
6 2 0
1
23 1 y3
= x+ x
6 2 3 0
= 4x
Chapter 7 INTEGRAL EQUATIONS
3. The Method of Successive Approximations for the Volterra Integral Equation of the 2nd kind
u 0 (x ) = f (x )
n
u n (x ) = n K k f = Ku n 1 + f
k =0
Identify: K (x , y ) = 1 f (x ) = 1
M =1 =1
K0 f = f (x ) = 1
( )
x x
= K (x , y ) K 0 f ( y )dy = 1 1dy = [ y ]0
x
K1 f =x
0 0
x
y2
( )
x x
x2
K2 f = K (x , y ) K 1 f ( y )dy = 1 ydy = =
0 0 2 0 2
x
1 y3
( ) y2
x x
x3
K3 f = K (x , y ) K 2 f ( y )dy = 1 dy = =
0 0 2 2 3 0 23
xn
Kn f =
n!
( ) xk
Solution: u (x ) = n K k f (x ) = = ex
k =0 k =0 k!
Chapter 7 INTEGRAL EQUATIONS
7.6 Connection between integral equations and initial and boundary value problems
u ( y ) 3 y u ( y )dy = 0
2
0
x x
( u )dy ( 3 y u )dy = 0
2
0 0
x
u ( x ) u ( 0 ) 3 y 2 u ( y )dy = 0 use the initial condition u (0 ) = 1
0
x
u ( x ) = 1 + 3 y 2 u ( y )dy is a Volterra equation with K ( x, y ) = y 2
0
d
b(x ) b( x )
g ( x, y ) db ( x ) da ( x )
g (x , y )dy dy + g x,b ( x ) g x,a ( x )
dx a ( x )
=
a( x) x dx dx
In particularly,
d x
g ( y )dy = g (x )
dx 0
d x x
g (x , y )
g (x , y )dy dy + g (x , x )
dx 0
=
0 x
0
initial value problem.
x 0
u (x ) = x 3 + (x y ) u ( y )dy u ( 0 ) = 0 3 + ( x y ) u ( y ) dy u (0 ) = 0
2 2
0
0
0
Chapter 7 INTEGRAL EQUATIONS
x 0
u (x ) = 3 x 2 + 2(x y )u ( y )dy u (0 ) = 30 2 + 2(x y )u ( y )dy u (0 ) = 0
0 0
x x
u (x ) = 3 x 2 + 2 u ( y )dy u (0 ) = 30 2 + 2 u ( y )dy u (0 ) = 0
0 0
u (x ) = 6 x + 2u (x )
Therefore, the integral equation is reduced to IVP for 3rd order ODE:
u 2u = 6 x u (0 ) = 0
u (0 ) = 0
u (0 ) = 0
x tn t3 t 2 x
1
" f ( t1 )dt1 dt2 " dtn 1 dtn = ( x t ) f ( t ) dt
n 1
0 0 0 0 ( n 1) ! 0
Set y (x ) = u (x )
x x
integrate y (t )dt = u (t )dt
0 0
x
y (x ) y (0 ) = u (t )dt
0
x x t 2
integrate [y (t 2 ) y (0 )]dt 2 = u (t 1 )dt 1 dt 2
0
0 0
x t 2
y (x ) y (0 ) y (0 )x = u (t 1 )dt 1 dt 2
0
0
x
y (x ) y (0 ) y (0 )x = (x t )u (t )dt repeated integration
0
y ( ) = 1 + y (0 ) + ( t )u (t )dt
0
1 = 1 + y (0 ) + ( t )u (t )dt
0
Solve for y (0 )
2 1
y (0 ) = 1 ( t )u(t )dt
0
Then
2 1 x
y (x ) = 1 + 1 ( t )u (t )dt x + (x t )u (t )dt
0 0
x
2 x
= 1 + x x ( t )u (t )dt + (x t )u (t )dt
0 0
x
2 x
u +1+ x
x
( t )u (t )dt + (x t )u(t )dt = x
0 0
x
2 x
u +1
x
( t )u (t )dt + (x t )u (t )dt = 0
0 0
x
2 x
u = 1 +
x+
( t )u (t )dt (x t )u (t )dt
0 0
x x
2 x x
u = 1 +
x+
( t )u (t )dt + ( t )u (t )dt (x t )u(t )dt
0 x 0
2 x x x
x
u= x 1 + ( t )u (t )dt (x t )u (t )dt + ( t )u (t )dt
0 0 x
2 x x x
x( t )
u= x 1 + ( t )u (t )dt (x t )u (t )dt + u (t )dt
0 0 x
2 x
x
x( t )
u= x 1 + ( t ) (x t )u (t )dt + u (t )dt
0 x
2 x
t ( x )
x( t )
u= x 1+ u (t )dt + u (t )dt
0 x
2
u= x 1 + K ( x,t )u ( t ) dt
0
with a kernel
t ( x )
0t x
K (x , t ) =
x( t )
x t
Chapter 7 INTEGRAL EQUATIONS
7.7 Exercises
x1 1 1
e) u ( x) = 1 + dy
4 0 x + y u( y )
x
u ( x ) = x + ( y x ) u ( y ) dy
0
4. Find the equivalent Volterra integral equation to the following initial value
problem
y ( x ) + y ( x ) = cos x y (0 ) = 0 y (0 ) = 1
y + y = x x ( 0,1) y (0 ) = 1 y ( 1) = 0
1
a) u ( x ) = x + xyu ( y ) dy
0
12
b) u ( x ) = x + cos xu ( y ) dy
40
x
u ( x ) = 1 ( y x ) u ( y ) dy
0
Chapter 7 INTEGRAL EQUATIONS
x
a) u ( x ) = sin 2x + u ( x s ) sin ( s ) ds
0
x
b) u ( x ) = x 2 + u ( x s ) e as ds u (0 ) = 0
0
9. Using mathematical induction prove identity for iterated kernel (7.5 2):
K n (x , y ) = K (x , y )K n 1 ( y , y )dy
G
10. Using mathematical induction verify the following estimate for iterated
kernels (7.5 2):
K n (x , y ) M n (b a )
n 1
11. Verify result of Example 7.4 by solving both IVP and derived integral
equation.
Chapter 7 INTEGRAL EQUATIONS
The Scottish caf in Lvov (Ukraine) was a meeting place for many mathematicians including Banach, Steinhaus,
Ulam, Mazur, Kac, Schauder, Kaczmarz and others. Problems were written in a book kept by the landlord. A
collection of these problems appeared later as the Scottish Book. R D Mauldin, The Scottish Book, Mathematics
from the Scottish Caf (1981) contains the problems as well as some solutions and commentaries.
Fredholm is best remembered for his work on integral equations and spectral theory.
Find out more at: http://www-history.mcs.st-andrews.ac.uk/history/Mathematicians/Fredholm.html
Volterra published papers on partial differential equations, particularly the equation of cylindrical
waves. His most famous work was done on integral equations. He published many papers on what is
now called 'an integral equation of Volterra type'.
Find out more at: http://www-history.mcs.st-andrews.ac.uk/history/Mathematicians/Volterra.html