Sie sind auf Seite 1von 27

# Chapter 6.

Fourier Series
6.1

Periodic functions

## A function f (x) is called periodic if it is defined for

all real x and if there is some positive number p such
that

(1)

6.1.1

## Graphs of periodic functions

The graph of such a function can be obtained by periodic repetition of its graph in any interval of length
p.

## For example, sine and cosine functions are periodic

2.
f (x) = c, c constant, is a periodic function of period
p for every positive number p.
x, x2, x3, , ex, ln x are not periodic.
6.1.2

## Some algebraic properties of periodic

functions

From (1),
f (x + 2p) = f ((x + p) + p) = f (x + p) = f (x).

## Thus (by induction) for any positive integer n,

f (x + np) = f (x), for all x.
Hence 2p, 3p, are also periods of f .
Further, if f and g have period p, then the function
h(x) = a f (x) + bg(x) with a, b constants also has
period p.
6.1.3

Trigonometric series

## Our aim is to represent various periodic functions of

period 2 in terms of simple functions
1, cos x, sin x, cos 2x, sin 2x, , cos nx, sin nx, (2)
which have period 2.
The series that arises in this connection will be of the

## MA1505 Chapter 6. Fourier Series

form
a0 + a1 cos x + b1 sin x + a2 cos 2x + b2 sin 2x +
= a0 +

+ bn sin nx)

## where a0, a1, a2, , b1, b2, are real constants.

Series (3) is called a trigonometric series, and an
and bn are called coefficients of the series.
The set of functions (2) is often called a trigonometric system.
We note that each term of the series (3) has period
2. Hence if the series converges, its sum will be a
periodic function of period 2.

(3)

6.2

Fourier Series

## Assume that f (x) is a periodic function of period

2 and that it can be represented by a trigonometric
series

X
f (x) = a0 +
(an cos nx + bn sin nx).

(4)

n=1

That is, we assume that the series on the right converges and has f (x) as its sum.
We say the right hand side of (4) is the Fourier series
of f (x).
Given f (x), our task now is to determine the coefficients an and bn.

6.2.1

Determine a0

## We integrate both sides of (4) from to :

Z

f (x)dx =

X
(a0 + (an cos nx+bn sin nx)) dx.

n=1

## Assuming that term by term integration is allowed,

we obtain
Z

f (x)dx

Z
Z
Z

X
= a0
dx +
(an cos nx dx + bn sin nx dx)

n=1

X
cos nx
sin nx
= 2a0 +
an
+ bn
n
n

n=1
= 2a0
1
So a0 =
2

f (x)dx.

## Determine am, m > 0

6.2.2

We multiply both sides of (4) by cos mx and integrate term by term from to :
Z

f (x) cos mx dx

=a0

cos mx dx +

X
n=1

Z
an

+ bn

cos nx cos mx dx

sin nx cos mx dx
(5)

## Computing the three integrations on the right hand

sides of (5):

Z
sin mx
(i)
cos mx dx =
= 0.
m

(ii)
sin nx cos mx dx = 0,

and
7

## MA1505 Chapter 6. Fourier Series

(iii)

cos nx cos mx dx

1
=
(cos(m + n)x + cos(m n)x) dx
2

1 sin(m + n)x sin(m n)x
+
m 6= n
=
2
m+n
mn

1
[mx + sin mx cos mx]
m=n
2m
0 m 6= n
=
m=n
Substituting the above results back in (5), we get
1
am =

6.2.3

to :

f (x) sin mx dx

=a0

sin mx dx +

X
n=1

an

cos nx sin mx dx

n=1

Z
+ bn
sin nx sin mx dx
(6)

bn

sin nx sin mx dx

## as the first two integrands on the right hand side of

(6) are odd functions.
Z
Now
sin nx sin mx dx

Z
1
(cos(n m)x cos(n + m)x)dx
=
2
i
( h
sin(n+m)x
sin(nm)x
1
n+m
m 6= n
2
nm
=

1
[mx

sin
mx
cos
mx]
m = n
2m
0 m 6= n
=
m=n
Z
1
Thus bm =
f (x) sin mx dx, m = 1, 2, .

9

10

6.2.4

Euler formulas

## Given a periodic function f (x) of period 2 with

Fourier series

X
a0 +
(an cos nx + bn sin nx).
n=1

## Its coefficients are known as Fourier coefficients and

are given by
Z
1
a0 =
f (x)dx
2
Z
1
f (x) cos nx dx, n = 1, 2, (7)
an =

Z
1
bn =
f (x) sin nx dx, n = 1, 2,

(7) are known as Euler formulas.

10

11

6.2.5

## If a periodic function f (x) with period 2 is piecewise

continuous in the interval x and has
a left hand derivative and right hand derivative at
each point of the interval, then the Fourier series with
coefficients (7) is convergent. Its sum is f (x) except
at a point x0 at which f (x) is discontinuous and the
sum of the series is the average of the left hand and
right hand limits of f at x0.
6.2.6

Example

f (x) =

k, if < x < 0
k, if 0 < x <
11

12

## and f (x) = f (x + 2).

Functions of this kind occur as external forces acting on mechanical systems, electromotive forces in
electric circuits, etc.
f (x) is piecewise continuous and the value of f at
a single point does not affect the integral. We can
therefore leave f undefined at x = 0, x = .

## Solution. We observe that over the interval (, ),

f is an odd function. Thus f (x) cos nx is also an odd
12

13

and
bn =
=
=
b1 =
b4 =

Z
1
f (x) sin nx dx

Z
2
2k
k sin nx dx =
(1 cos n)
0
n
2k
(1 (1)n).
n
4k
4k
, b2 = 0, b3 = ,

3
4k
0, b5 = , .
5

## The Fourier series for the square wave is, therefore,

4k
1
1
(sin x + sin 3x + sin 5x + ).

3
5
6.2.7

An approximation for

f (x) in (0, ).
13

14

## MA1505 Chapter 6. Fourier Series

Setting x = 2 , we get

k=
i.e.

4k
1 1
(1 + )

3 5

= 1 13 + 15 (Leibniz).

## Note that at all the points of discontinuity (0, , etc)

of f , the sum of the series is equal to 0, which is the
average of the left hand and right hand limits of f
(e.g. they are k and k respectively at x = 0).
6.2.8

## Let f (x) be a periodic function of period p = 2L.

We set v =

x
L .

Then x =

vL

and at x = L, v =

.
We now view f as a function of v and put f (x) =
14

15

## g(v). Then g becomes a periodic function of period

2. If f (x) has a fourier series, then so has g(v). We
have

X
g(v) = a0 +
(an cos nv + bn sin nv)
n=1

with
Z

1
1
a0 =
g(v)dv =
2
2
Z L
1
=
f (x)dx
2L L

g(v) dx
L
L

and for n = 1, 2, 3,
Z

1
an =
g(v) cos nv dv

Z
1 L
nx
=
f (x) cos
dx,
L L
L
Z
nx
1 L
f (x) sin
dx.
bn =
L L
L

15

16

## Since g(v) = f (x), we get

X
n
n
f (x) = a0 +
an cos x + bn sin x
L
L
n=1

## with a0, an and bn as given above.

The interval of integration in the above formula can
be replaced by any interval of length p = 2L, for
example, by 0 x 2L or L x 3L.
6.2.9

Example

## Let f be a periodic square wave of period p = 2L = 4

defined as follows :

0, if 2 < x < 1
f (x) = k, if 1 < x < 1

0, if 1 < x < 2

16

17

## To find the Fourier series of f , we compute

Z
1 2
f (x)dx
a0 =
4 Z2
1 2
nx
an =
f (x) cos
dx
2 2
2
and since f is even,
bn = 0 for n = 1, 2, .
Z

1
1 1
k
a0 =
f (x)dx =
kdx =
4 2
4 1
2
Z 2
Z 1
1
nx
1
nx
an =
f (x) cos
dx =
k cos
dx
2 2
2
2 1
2
2k
n
=
sin
n
2

17

18

## Hence an = 0 if n is even and

2k if n = 1, 5, 9,
an = n2k

if n = 3, 7, 11,
n
Hence
k 2k

1
3
1
5
f (x) = + (cos x cos x+ cos x ).
2
2
3
2
5
2
Fourier cosine and sine series

6.2.10

Using
Z

f (x)dx =
L

if f is odd

RL
2 0 f (x)dx if f is even.

## we obtain the following two series.

The Fourier series of an even function f (x) of period
2L is the Fourier cosine series
f (x) = a0 +

X
n=1
18

an cos

nx
,
L

19

## MA1505 Chapter 6. Fourier Series

with
Z

1 L
a0 =
f (x)dx,
L Z0
2 L
nx
an =
f (x) cos
dx, n = 1, 2, .
L 0
L
The Fourier series of an odd function f (x) of period
2L is a Fourier sine series
f (x) =
2
with bn =
L
6.2.11

bn sin

n=1

f (x) sin
0

nx
,
L

nx
dx.
L

## The Fourier coefficients of f1 + f2 are the sums of

corresponding Fourier coefficients of f1 and f2.
The Fourier coefficients of cf (c a constant) are c
times the corresponding Fourier coefficients of f .
19

20

6.2.12

Example

## Saw tooth function

f (x) = x + , < x < ,
f (x) = f (x + 2).

## We note that f = f1 + f2, where f1 = x, f2 = .

The Fourier coefficients for f2 are a0 = and
an = 0 = bn, n 1.
The function f1 = x is odd.

20

21

## Thus an = 0 for all n, and

Z

bn =
=
=
=

2
x sin nx dx

0

2
cos nx
cos nx
x

dx

n
n
0
0

2 (1)n
sin nx

n
n2
0
n+1
(1) 2
n

So
f (x) = f1(x) + f2(x)
=

X
(1)n+12
n=1

= +2

nx
sin
+

X
(1)n+1
n=1

21

sin nx

22

6.3

## MA1505 Chapter 6. Fourier Series

Half-range Expansions

## In various applications there is a practical need to use

Fourier series in connection with functions f that are
given on some interval only, say, 0 x L.

Figure (a)
6.3.1

Extension of f (x)

## We could extend f (x) as a periodic function with

period L and then represent the extended function
by a Fourier series, which in general would involve
both sine and cosine terms. We can do better and

22

23

## always get a cosine series by first extending f (x) from

0 x L as an even function on the interval L
x L as in figure (b) and then extend this new
function as a periodic function of period 2L, and since
it is even, we can represent it by a Fourier cosine
series.

Figure (b)
Also, we can extend f (x) from 0 x L as an
odd function on L x L as in figure (c) and
then extend this new function as a periodic function
of period 2L, and since it is odd, it is represented by
a Fourier sine series.
23

24

Figure (c)
6.3.2

## Half range expansion

Such two series are called the two half range expansions of the function f which is given only on half
the range.
The cosine half range expansion is
f (x) = a0 +

X
1

nx
an cos
L

with
Z
1 L
a0 =
f (x)dx,
L Z0
2 L
nx
an =
f (x) cos
dx, n = 1, 2,
L 0
L
24

25

f (x) =

X
1

bn sin

n
x
L

with
2
bn =
L
6.3.3

L
0

nx
f (x) sin
dx, n = 1, 2, .
L

Example

## Find the two half range expansions for

0, 0 < x < 2
f (x) =

1, 2 < x < .
For the cosine half range expansion, we have
(Z
)
Z
/2
1
1
a0 =
0dx +
1dx =

2
0
/2

25

26

## MA1505 Chapter 6. Fourier Series

and
an
(Z
)
Z
/2
2
0 cos nxdx +
1 cos nxdx
=

0
/2
2 sin n sin 12 n
= [
]

n
Thus an simplifies to
2
n
an =
sin
n
2
Indeed,
2
a1 =
,

2
a3 = ,
3

2
a5 =
,
5

## and an = 0 if n 1 and n is even.

The cosine half range expansion is

X
1
(1)m
f (x) = + 2
cos(2m 1)x
2
(2m

1)
m=1

26

27

## For the sine half range expansion, we have

bn
2
=

(Z

/2

0 sin nxdx +

1 sin nxdx

/2

2 cos n + cos 12 n
= [
]

n
Thus bn simplifies to
2 h
n i
n+1
bn =
(1)
+ cos
n
2
Indeed,
b1 = 2 ,
b5 =

2
5 ,

b2 =

4
2 ,

b6 =

4
6 ,

b3 =

2
3 ,

b7 =

b4 = 0,

2
7 ,

b8 = 0, .

f (x) =
2

P n sin(4m3)x
m=1

4m3

2 sin(4m2)x
4m2
27

sin(4m1)x
4m1