010.141
FOURIER SERIES, INTEGRALS, AND TRANSFORMS
MODULE 3
ADVANCED ENGINEERING MATHEMATICS
Fourier Series, Integrals, and Transform
Ø Fourier series: Infinite series designed to represent general periodic functions in terms of simple ones (e.g., sines and cosines).
Ø Fourier series is more general than Taylor series because many
discontinuous periodic functions of practical interest can be developed in Fourier series.
Ø Fourier integrals and Fourier Transforms extend the ideas and techniques of Fourier series to non-periodic functions and have basic applications to PDEs.
PERIODIC FUNCTIONS
Ø A periodic function is a function such that f(x + p) = f(x) where p is a period.
Ø The smallest period is called a fundamental period
.FOURIER SERIES
Assume that f(x) is periodic with period 2p and is integrable over a period.
f(x) can be represented by a trigonometric series:
i.e., the series converges and has sum f(x)
( ) ( )
f x = a +
0a cos nx + b sin nx
n nn = 1
¥
å
Euler Formulas
Fourier coefficients
a , a , b are the Fourier coefficients.
( ) ( )
( )
0
1 2
1 cos
1 sin
n
n
a = f x dx
a = f x nx dx
b = f x nx dx
p
p p
p p
p
p
p
p
-
-
-
×
×
ò ò ò
TRIGONOMETRIC SERIES WITH A PERIOD OF 2p
f x =
( )
k, < x < 0 k, < x <- -
ìí î
p
p 0
Example: Rectangular wave
FOURIER SERIES (cont)
( ) [ ]
0 0
0
n
1 1
a = k dx+ dx 0
2 2
1 1
a = cos x dx= 0 =0
k
f x n
p
p p
p
p p
p p
-
-
- =
ò ò
ò
Fourier Coefficients
( ) ( )
( )
n
0
0 0
0
b = 1 sin x dx
1 1
= sin x dx + sin x dx
s nx cos nx
= n n
= 1 1 1 1 = 4k odd n
f x n
k n k n
k co
k for n
n
p
p
p
p
p
p
p
p p
p
p p
-
-
-
-
æ é ù é ù ö
ç ê ú + -ê ú ÷
ç ë û ë û ÷
è ø
+ + +
ò
ò ò
EXAMPLE 1: RECTANGULAR WAVE
( )
( )
( ) ( )
0 n n
n = 1 n = 1
2n+1 n = 0
n = 0
x =a + a cos cos
= sin 2 1
4 sin 2 1 2 1
4 1 1
sinx+ sin 3 sin 5
2 1+1 5
f nx b nx
b n x
k n x
n
k x x
p
p
¥ ¥
¥
¥
+ +
= +
+
æ ö
= ç + + ÷
è × ø
å å
å å
L Fourier series:
ORTHOGONALITY OF SINE, COSINE FUNCTIONS
Two functions fm and fn of the same form are orthogonal if ò fm fn dx = 0 for all m ¹ n and ò fm fn dx = a for all m = n.
( )
( )
2
2
cos mx cos nx dx 0 for all m n
cos nx dx 0 for all m n
sin mx sin nx dx 0 for all m n
sin nx dx 0 for all m n
cos mx sin nx dx
p
p p
p p
p p
p p
p -
-
-
-
-
= ¹ ±
¹ =
= ¹
¹ =
ò ò ò ò
ò
= 0 for all m = nCONVERGENCE AND SUM OF FOURIER SERIES
CONVERGENCE AND SUM OF FOURIER SERIES
Proof: For continuous f(x) with continuous first and second order derivatives.
( ) ( ) ( )
n
0
a 1 cos nx dx
f sin nx
1 1 sin nx
dx dx
f sin nx
1 1
'sin nx dx
1 'sin nx dx f x
x f
n n
x f
n n
n f
p
p
p p
p p
p p
p p
p
p
p
p p
p p
p
-
- -
- -
-
=
æ ö¢
= ç ÷ - ¢
è ø
= -
= -
ò
ò ò
ò ò
1442443
Repeating the process:
Since f" is continuous on [– π, π ]
CONVERGENCE AND SUM OF FOURIER SERIES (cont)
( ) x cos nx dx
"
n f a 1
n 2
ò
p
p
p
--
=
( )
( )
2n 2
M 2 M
2
dx M n
dx 1 nx cos x
"
f n
a 1
M x
"
f
p
p
<
p
=
<
ò ò
p
p - p
p -
Similarly for
Hence
which converges. (see Page 670)
CONVERGENCE AND SUM OF FOURIER SERIES (cont)
n 2
n M b < 2
( )
÷ø ç ö
è
æ + + + + + +
+
< L
2 2
2 0 2
3 1 3
1 2
1 2
1 1 1 M 2 a
x f
FUNCTIONS OF ANY PERIOD
P = 2L
FUNCTIONS OF ANY PERIOD (cont) P = 2L
( )
( ) ( ) ( )
( ) ò ( ) ò ( )
ò
ò ò
ò
å å
- -
p
p -
- -
p
p -
¥
=
¥
=
p =
= p
= p
= p
× p p
= p
= p
+ +
=
= p
L
L L
L 0
L
L L
L n
1 n
n 1
n
n 0
dx x L f
2 dx 1 x L
2 f dv 1 v 2 g
a 1
L dx x cos n
x L f
dx 1 L L
x cos n
x 1 f
dv nv cos v 1 g
a
nv sin b nv
cos a
a v g
L v x
Proof: Result obtained easily through change of scale.
Same for a0, bn
ANY INTERVAL (a, a + P) P = PERIOD = 2L
( )
( )
( )
( )
dxP x sin 2n
x P f
b 2
P dx x cos 2n
x P f
a 2
dx x P f
a 1
P x in 2n
s P b
x cos 2n
a a
x f
P a
a n
P a
a n
P a
a 0
1 n
n n
0
= p
= p
=
÷ø ç ö
è
æ p
p + +
=
ò ò ò
å
+ + +
¥
=
a a + P
EXAMPLE OF APPLICATION
Find the Fourier series of the function
Solution:
( )
ïî ïí ì
<
<
<
<
-
=
=
<
<
=
7 x 1 7 x 4 1
1 a
6 P 4 x 1 1 x
f
1 4 7
x f(x)
( ) ( )
( ) ú
û ù êë
é p
p - p -
- p ú= ú û ù ê
ê ë
é p
p - p =
=
ú ¹ û ù êë
é p
p - p -
= p ú ú û ù ê
ê ë
é p
p -
=
ú= ú û ù ê
ê ë é
- +
=
=
ò ò
ò
ò ò
ò ò
ò
3 s 7n 3 co
s n co 3
cos 4n n 2
dx 1 3
x sin n 3 dx
x sin n 3
dx 1 3
x sin n x 3 f
b 1
0 n 3 sin 7n 3
sin n 3
sin 4n n 2
dx 1 3
x cos n 3 dx
x cos n 3
a 1
0 dx 1 dx
6 1 dx 1 x 6 f
a 1
7
4 4
1 7
1 n
7
4 4
1 n
7
4 4
1 7
1 0
EXAMPLE OF APPLICATION (cont)
3 cos n n
n cos 2 1
b
3 in n n s
n cos 2 1
a
0 a
3 cos n 3
s 7n co
3 cos n n
3 cos cos 4n
3 in n 3 s
sin 7n
3 in n s n 3 cos
sin 4n
n n 0
÷ p ø ç ö
è æ
p p
= -
÷ p ø ç ö
è æ
p p - -
=
=
= p p p p
p =
= p p p p
p =
x
-3 1 4 7
f(x)
EVEN AND ODD FUNCTIONS
Examples: x4, cos x
f(x) is an even function of x, if f(-x) = f(x). For example, f(x) = x sin(x), then
f(- x) = - x sin (- x) = f(x)
and so we can conclude that x sin (x) is an even function.
Properties
1. If g(x) is an even function
2. If h(x) is an odd function
3. The product of an even and odd function is odd
EVEN AND ODD FUNCTIONS
( ) ( )
( )
x dx 0h
dx x g 2 dx x g
L
L
L
0 L
L
=
=
ò
ò ò
- -
EVEN AND ODD FUNCTIONS
EVEN AND ODD FUNCTIONS
SUM
SUM
SUM
SUM
Forced Oscillations
Forced Oscillations
Forced Oscillations
Forced Oscillations
Forced Oscillations
Consider a function f(x), periodic of period 2π. Consider an approximation of f(x),
The total square error of F
is minimum when F's coefficients are the Fourier coefficients.
APPROXIMATION BY
TRIGONOMETRIC POLYNOMIALS
( )
0n 1
( ) x cos sin
N
n n
f x F A A nx B nx
=
» = +
å
+( )
ò
p
p -
-
= f F dx
E 2
PARSEVAL'S THEOREM
( ) ( )
2 2 2 2
0
n 1
2a an bn 1 f x dx
p
p
p¥
= -
+
å
+ £ò
The square error, call it E*, is
Where an, bn are the Fourier coefficients of f.
FOURIER INTEGRALS
Since many problems involve functions that are nonperiodic and are of
interest on the whole x-axis, we ask what can be done to extend the method of Fourier series to such functions. This idea will lead to “Fourier integrals.”
FOURIER INTEGRALS
FOURIER INTEGRALS
FOURIER COSINE AND SINE INTEGRALS
FOURIER COSINE AND SINE INTEGRALS
FOURIER COSINE AND SINE INTEGRALS
( ) ( ) ( )
f x =
ò
0¥éëA w coswx+ B w sinwxùûdw( ) ( )
( )
0
1 2
1 cos
1 sin
n
n
a = f x dx
a = f x nx dx
b = f x nx dx
p
p p
p p
p
p
p
p
-
-
-
×
×
ò ò ò
( )
( )
f x = a +0 a cos nx + b sin nxn n
n = 1
¥
å
( ) ( ) ( )
( ) ( ) ( )
1 cos d
1 sin d
A f v v v
B f v v v
w w
p
w w
p
¥
- ¥
¥
- ¥
=
=
ò
ò
EXISTENCE
If f(x) is piecewise continuous in every finite interval and has a right hand and left hand derivative at every point and if
exists, then f(x) can be represented by the Fourier integral.
Where f(x) is discontinuous, the F.I. equals the average of the left-hand and right-hand limit of f(x).
( ) ò ( )
ò
®¥¥ -
®
+
b
b 0 0
a a
dx x
f lim
dx x
f lim
For an even or odd function the F.I. becomes much simpler.
If f(x) is even
If f(x) is odd
( ) ( )
( ) ( ) ( )
( ) ( )
( )
x B( )
ω sin(
ωx)
dω Fourier sineintegralf
dv ωv sin v
π f ω 2
B
integral cosine
Fourier
dω ωx
cos ω
A x
f
dv ωv cos v
π f ω 2
A
0 0
ò ò ò
ò
¥
¥
¥ -
¥
¥
¥ -
=
=
=
=
FOURIER COSINE AND SINE INTEGRALS
EXAMPLE
Consider
Evaluate the Fourier cosine integral A(w) and sine integral B(w).
( )
x e x 0, k 0f = -kx > >
Integration by parts gives For Fourier cosine integral,
Fourier cosine integral is
EXAMPLE
Integration by parts gives For Fourier sine integral,
Fourier sine integral is
Laplace integrals
FOURIER SINE AND COSINE TRANSFORMS
For an even function, the Fourier integral is the Fourier cosine integral
( ) ( ) ( ) ( ) ( )
ˆ( )
π dv 2
ωv cos v π f
ω 2 A dω
ωx cos ω A x
f
0
c w f
=
=
=
ò ò
¥
¥ -
¥
( ) ( ) ( ) ( )
( ) ( ) ( )
( )
ˆ
. of transform cosine
Fourier the
as defined is
ˆ
d ωx ˆ cos
π 2
d ωx ˆ cos
π x 2
f
dv ωv cos v π f
dv 2 ωv cos v π f
2 2 π
ω 2 A
π ˆ
0 0
0 0
c
c c
c
f f
f f
f
w
w w
w w
w
ò ò
ò ò
¥
¥
¥
¥
=
=
=
=
=
Then
FOURIER SINE AND COSINE TRANSFORMS (cont)
Similarly for odd function
( ) ( )
( )
fˆ( )
sin ωx dπ x 2
f T
sine F
Inverse
dx ωx sin x
π f 2
fˆ T
sine F
0 c 0
c
ò ò
¥
¥
w w
=
®
= w
®
NOTATION AND PROPERTIES
{ } { }
{ } { }
{ } { } { }
{ } { } { }
( )
{ } { } ( )
( )
{ } { }
{ } { } ( )
2
0 π f
f 2 ω
f (5)
f ω x
f (4)
0 π f
f 2 ω
x f (3)
g b
f a
bg af
(2)
g b
f a
bg af
(1)
ˆ ˆ ,
ˆ ˆ ,
c 2 c
c s
s c
s s
s
c c
c
s 1 s c
1 c
s s
c c
- ¢ -
¢¢ =
-
¢ =
-
¢ =
+
= +
+
= +
=
=
=
=
- -
F F
F F
F F
F F
F F
F F
F
F F
F
f f
f f
f f
f f
FOURIER TRANSFORM
( ) [ ( ) ( ) ]
( ) ( )
4 4 4
4 3
4 4 4
4 2
1
w
¥
¥ -
¥
¥
¥ -
¥
ò ò
ò ò
w - w p w
=
w w
+ w w
p w
=
in Even 0
0
v d x v
cos v
f 1 d
v d x sin v sin v f x cos v cos v
f 1 d
x f
( ) [ ( ) ( ) ]
( ) ( )
( ) ò ( )
ò ò
¥
¥ -
¥
¥ -
¥
= w
= w
+
=
v d ωv sin v
π f B 1
v d ωv cos v π f
A 1
dx ωx sin ω
B ωx cos ω
A x
f
0
The F.I. is:
Replacing
FOURIER TRANSFORM
( ) ( ) ( )
( )
d f( )
v sin(
x v)
dvl2 x 1
f
v d v x cos v
f 2 d
x 1 f
in odd
4 4 4
4 3
4 4 4
4 2
1
w
¥
¥ -
¥
¥ -
¥
¥ -
¥
¥ -
ò ò
ò ò
- w p w
=
- w p w
=
Now,
( ) ( ) ( )
( )
v e ( )dv Complex Fourier Integralf 2 d
1
dω G
i ω 2 F
x 1 f
v x
ò
iωò ò
¥
-
¥
¥
¥ -
p w
=
w p +
=
FOURIER TRANSFORM (cont)
( ) ( )
( )
( ) ( )
w w= p
w p
= p
ò
ò ò
¥
¥ -
º w
¥
¥ -
-
¥
¥ -
d e
2 fˆ 1 x f
dv e
v 2 f
d 1 2 e
x 1 f
x iω
f of Transform Fourier
fˆ
v iω x
iω
4 4
4 3
4 4
4 2
1
NOTATION AND PROPERTIES
{ } { } { }
{ } { }
{ } { }
{ f g } 2 { } f { } g
f ω
f
f iω
f
g b
f a
bg af
2