The Fourier Series

BET2533 – Eng. Math. III

R. M. Taufika R. Ismail
FKEE, UMP
Introduction
A Fourier series is an expansion of a
periodic function f (t) in terms of an infinite sum
of cosines and sines

a0 
f (t )    (an cos nt  bn sin nt )
2 n 1
In other words, any periodic function can be
resolved as a summation of
constant value and cosine and sine functions:
a0 
f (t )    (an cos nt  bn sin nt )
2 n 1
a0  (a cos t  b sin t )
 1 1
2
 (a2 cos 2t  b2 sin 2t )

 (a3 cos 3t  b3 sin 3t )  

The computation and study of Fourier series is
known as harmonic analysis and is extremely
useful as a way to break up an arbitrary
periodic function into a set of simple terms that
can be plugged in, solved individually, and
then recombined to obtain the solution to the
original problem or an approximation to it to
whatever accuracy is desired or practical.
f(t) Periodic Function a0
2

=
t

a1 cos t b1 sin t

+ +

a2 cos 2t b2 sin 2t

+ + + …
 a0 
f (t )    (an cos nt  bn sin nt )
2 n 1
2
where    Fundemental frequency
T
T
2
a0   f (t )dt
T 0
T T
2 2
an   f (t ) cos ntdt bn   f (t ) sin ntdt
T 0 T 0
T /2

*we can also use the integrals limit 

T / 2
.
Example 1
Determine the Fourier series representation of the
following waveform.
Solution
First, determine the period & describe the one period
of the function:

T=2

1, 0  t  1
f (t )   f (t  2)  f (t )
0, 1  t  2
 Then, obtain the coefficients a0, an and bn:
T 2 1 2
2 2
a0   f (t )dt   f (t )dt   1dt   0dt  1  0  1
T 0 20 0 1

Or, since  f (t )dt is the total area below graph

a
y = f(t) over the interval [a,b], hence

2  Area below graph  2

T
2
a0   f (t )dt       (11)  1
T 0 T  over [0, T ]  2
 2
2
an   f (t ) cos ntdt
T 0
1
 sin nt  sin n
1 2
  1 cos ntdt   0dt    
0 1  n  0 n

Notice that n is integer which leads sin n  0 ,

since sin   sin 2  sin 3    0

Therefore, an  0.
 2
2
bn   f (t ) sin ntdt
T 0
1
 cos nt  1  cos n
1 2
  1sin ntdt   0dt    
0 1  n  0 n

Notice that cos   cos 3  cos 5    1

cos 2  cos 4  cos 6    1
or cos n  (1) n
1  (1) n 2 / n , n odd
Therefore, bn  
n  0 , n even
Finally,
a0 
f (t )    (an cos nt  bn sin nt )
2 n 1
1 
1  (1) 
n
   sin nt
2 n 1  n 
1 2 2 2
  sin t  sin 3t  sin 5t   ♣
2  3 5
Some helpful identities

sin(  x)   sin x cos( x)  cos x

For n integers,
sin n  0 cos n  (1) n
sin 2n  0 cos 2n  1
[Supplementary]
 The sum of the Fourier series terms can
evolve (progress) into the original
waveform
 From Example 1, we obtain
1 2 2 2
f (t )   sin t  sin 3t  sin 5t  
2  3 5
 It can be demonstrated that the sum will
lead to the square wave:
(a) (b)

2 2 2
sin t sin t  sin 3t
  3

(c) (d)

2 2 2 2 2 2 2
sin t  sin 3t  sin 5t sin t  sin 3t  sin 5t  sin 7t
 3 5  3 5 7
(e)

2 2 2 2 2
sin t  sin 3t  sin 5t  sin 7t  sin 9t
 3 5 7 9

(f)

1 2 2 2
 sin t  sin 3t    sin 23t
2  3 23
Example 2
Given f (t )  t , 1  t  1
f (t  2)  f (t )

Sketch the graph of f (t) such that  3  t  3.

Then compute the Fourier series expansion of f (t).
Solution
The function is described by the following graph:

T=2
2
We find that   
T
Then we compute the coefficients:
1
2
a0   f (t )dt
T 1
1
2
1
t  2
11
  tdt     0
2 1  2  1 2
 1 1
2
an   f (t ) cos ntdt   t cos ntdt
T 1 1
1
 t sin nt  sin nt
1
   dt
 n  1 1 n
1
sin n  [ sin(n )]  cos nt 
  2 2 
n  n   1
cos n  cos( n )
 0
n 2 2
cos n  cos n cos( x)  cos x
  0 since
n 2 2
 1 1
2
bn   f (t ) sin ntdt   t sin ntdt
T 1 1
1
 t cos nt  cos nt
1
    dt
 n  1 1 n
1
 cos n  [ cos( n )]  sin nt 
  2 2 
n  n   1
2 cos n sin n  sin(n )
 
n n 2 2
2 cos n 2(1) n 2(1) n 1
  
n n n
Finally,

a0 
f (t )    (an cos nt  bn sin nt )
2 n 1

2(1) n 1
 sin nt
n 1 n
2 2 2
 sin t  sin 2t  sin 3t   ♣
 2 3
Example 3
2  t , 0  t  2
Given v(t )  
 0 , 2t 4
v(t  4)  v(t )

Sketch the graph of v (t) such that 0  t  12.

Then compute the Fourier series expansion of v (t).
Solution
The function is described by the following graph:

v (t)

2 4 6 8 10 12 t
0
T=4

2 
We find that   
T 2
Then we compute the coefficients:
4
2
a0   v(t )dt
T 0
2  2 4

  (2  t )dt   0dt 
4 0 2 
2
1 t 
2 2
1
  (2  t )dt  2t    1
20 2 2 0
 4 2 4
2 1
an   v(t ) cos ntdt   (2  t ) cos ntdt   0
T 0 20 2
2
1  (2  t ) sin nt  1 sin nt
2
     dt
2 n  0 2 0 n
2
1  cos nt 
 0   2 2 
2  n  0
1  cos 2n 2(1  cos n ) 2[1  (1) ] n
  
2n 
2 2
n2 2
n2 2
 4 2 4
2 1
bn   v(t ) sin ntdt   (2  t ) sin ntdt   0
T 0 20 2
2
1   (2  t ) cos nt  1 cos nt
2
     dt
2 n  0 2 0 n
2
1 1  sin nt 
   2 2 
n 2  n   0
1 sin 2n 1 2
   
n 2n 2 2
n n

since sin 2n  sin n  0

 Finally,

a0 
v(t )    (an cos nt  bn sin nt )
2 n 1
1   2[1  (1) n ]  nt  2  nt 
   cos  sin  ♣
2 n 1  n 
2 2
 2  n  2 
Symmetry Considerations
 Symmetry functions:
(i) even symmetry
(ii) odd symmetry
Even symmetry
 Any function f (t) is even if its plot is
symmetrical about the vertical axis, i.e.
f (t )  f (t )
Even symmetry (cont.)
 The examples of even functions are:
f (t )  t 2 f (t ) | t |

t t
f (t )  cos t

t
Even symmetry (cont.)
 The integral of an even function from −A to
+A is twice the integral from 0 to +A

f even (t )

A A

t f
A
even (t )dt  2  f even (t )dt
0
−A +A
Odd symmetry
 Any function f (t) is odd if its plot is
antisymmetrical about the vertical axis, i.e.
f (t )   f (t )
Odd symmetry (cont.)
 The examples of odd functions are:
f (t )  t 3 f (t )  t

t t

f (t )  sin t

t
Odd symmetry (cont.)
 The integral of an odd function from −A to
+A is zero
f odd (t )

A

−A +A
t f
A
odd (t )dt  0
Even and odd functions
The product properties of even and odd
functions are:
 (even) × (even) = (even)
 (odd) × (odd) = (even)
 (even) × (odd) = (odd)
 (odd) × (even) = (odd)
Symmetry consideration
From the properties of even and odd
functions, we can show that:
 for even periodic function;
T /2
4
an 
T  f (t ) cos ntdt
0
bn  0

 for odd periodic function;

T /2
4
a0  a n  0 bn 
T  f (t ) sin ntdt
0
 How?? [Even function]
f (t )

t
T T

2 2

T /2
2
T /2
4
T /2
2
an  
T T / 2
f (t ) cos ntdt 
T  f (t ) cos ntdt bn  
T T / 2
f (t ) sin ntdt  0
0

(even) × (even) (even) × (odd)

|| ||

(even) (odd)
 How?? [Odd function]
f (t )

t
T T

2 2
T /2
2

T /2 T /2
a0  f (t )dt  0 2 4
T T / 2 bn  
T T / 2
f (t ) sin ntdt 
T  f (t ) sin ntdt
0
(odd)
T /2 (odd) × (odd)
2
an  
T T / 2
f (t ) cos ntdt  0
||
(even)
(odd) × (even)

||
(odd)
Example 4
 1 ,  2  t  1

Given f (t )   t ,  1  t  1
1 , 1 t  2

f (t  4)  f (t )

Sketch the graph of f (t) such that  6  t  6.

Then compute the Fourier series expansion of f (t).
Solution
The function is described by the following graph:
f (t)

1
−6 −4 −2 0 2 4 6 t
−1
T=4

2 
We find that   
T 2
Then we compute the coefficients. Since f (t) is
an odd function, then
2
2
a0   f (t )dt  0
T 2
and
2
2
an   f (t ) cos ntdt  0
T 2
 2 2
2 4
bn   f (t ) sin ntdt   f (t ) sin ntdt
T 2 T 0
4  1 2

   t sin ntdt   1sin ntdt 
4 0 1 
1 2
 t cos nt  cos nt  cos nt 
1
    dt   
 n  0 0 n   n 1
1
cos n  sin nt  cos 2n  cos n
  2 2  
n  n  0 n
cos 2n sin n 2 cos n
  2 2 
n n n
since sin 2n  sin n  0
Finally,
a0 
f (t )    (an cos nt  bn sin nt )
2 n 1

 2 cos n  nt
   sin
n 1  n  2

(1) n 1 nt
 2 sin ♣
n 1 n 2
Example 5
Compute the Fourier series expansion of f (t).
Solution
The function is described by
T=3
1 , 0  t  1

f (t )  2 , 1  t  2
1 , 2  t  3


f (t  3)  f (t )
T=3

2 2
and  
T 3
 Then we compute the coefficients.

2  2
3 1 2 3
2 8
a0   f (t )dt    1dt   2dt   1dt    (1  0)  2(2  1)  (3  2) 
T 0 3 0 1 2  3 3

Or, since f (t) is an even function, then

4  4
3 3/ 2 1 3/ 2
2 4  3  8
a0   f (t )dt   f (t )dt    1dt   2dt   (1  0)  2  1 
T 0 T 0
3 0 1  3  2  3

Or, simply
2  Total area below graph  2
3
2 8
a0   f (t )dt       4 
T 0 T  in a period  3 3
 3 3/ 2
2 4
an   f (t ) cos ntdt   f (t ) cos ntdt
T 0 T 0

4 
1 3/ 2
   1cos ntdt   2 cos ntdt 
3 0 1 
1 3/ 2
4  sin nt  4  2 sin nt 
   
3  n  0 3  n 1


4   3n 
 sin n  2 sin  sin n 
3n   2 
4  3n  2
  2 sin  sin n  ; 
3n  2  3
2  2n  2 2 n
  2 sin n  sin  sin
n  3  n 3
and bn  0 since f (t) is an even function.

Finally,
a0 
f (t )    (an cos nt  bn sin nt )
2 n 1
4   2 2n  2nt
   sin  cos
3 n 1  n 3  3
4 2  1 2n  2nt
    sin  cos
3  n 1  n 3  3 ♣
Function defines over a finite interval

 Fourier series only support periodic functions

 In real application, many functions are non-
periodic
 The non-periodic functions are often can be
defined over finite intervals, e.g.
y=2

y=1 y=1
t2
y=
t
y=
 Therefore, any non-periodic function must be extended to a
periodic function first, before computing its Fourier series
representation
 Normally, we prefer symmetry (even or odd) periodic extension
instead of normal periodic extension, since symmetry function
will provide zero coefficient of either an or bn
 This can provide a simpler Fourier series expansion
 Periodic extension f (t )
Non-periodic
function f (t )  y (t ) , 0  t  l
y (t ) f (t  l )  f (t )
t
T l  3l  2l  l 0 l 2l 3l
T

t Even periodic extension f even (t )

0 l
 y (t ) , 0  t  l
f (t )  
 y ( t ) ,  l  t  0 t
f (t  2l )  f (t )  3l  2l  l 0 l 2l 3l
T  2l T

Odd periodic extension f odd (t )

 y (t ) , 0t l
f (t )  
 y (t ) ,  l  t  0 t
f (t  2l )  f (t )  3l  2l  l 0 l 2l 3l
T  2l
T
Half-range Fourier series expansion

 The Fourier series of the even or odd

periodic extension of a non-periodic
function is called as the half-range Fourier
series
 This is due to the non-periodic function is
considered as the half-range before it is
extended as an even or an odd function
 If the function is extended as an even
function, then the coefficient bn= 0, hence
a0 
f (t )    an cos nt
2 n 1
which only contains the cosine harmonics.
 Therefore, this approach is called as the
half-range Fourier cosine series
 If the function is extended as an odd
function, then the coefficient an= 0, hence

f (t )   bn sin nt
n 1

which only contains the sine harmonics.

 Therefore, this approach is called as the
half-range Fourier sine series
Example 6
Compute the half-range Fourier sine series expansion
of f (t), where
f (t )  1 , 0  t  
Solution
Since we want to seek the half-range sine series,
the function to is extended to be an odd function:
f (t) f (t)

1 1
0 π t −2π −π 0 π 2π t
−1
T = 2π
2
 1
T
 Hence, the coefficients are
a0  a n  0

and
T /2 
4 4
bn 
T 
0
f (t ) sin ntdt  
2 0
1sin ntdt

2  cos nt  2 4 / n , n odd
    (1  cos n )  
 n  0 n  0 , n even

Therefore,
 
2 4
f (t )   (1  cos n ) sin nt   sin nt ♣
n 1 n n 1 n
n odd
Example 7
Determine the half-range cosine series expansion
of the function
f (t )  2t  1 , 0  t  1
Sketch the graphs of both f (t) and the periodic
function represented by the series expansion for
−3 < t < 3.
 Solution
Since we want to seek the half-range cosine series,
the function to is extended to be an even function:
f (t) f (t)
1 1
2 2
1 1 t 3 1 1 3 t
1 2
1

T=2

2
 
T
Hence, the coefficients are
 
T /2 1
4 4 1
a0   f (t )dt   (2t  1)dt  2 t  t 0  0
2

T 0 20
T /2 1
4 4
an 
T 
0
f (t ) cos ntdt   (2t  1) cos ntdt
20
1
 (2t  1) sin nt  sin nt
1
 2   2  2dt
 n 0 0
n
1
2 sin n  cos nt 
  4 2 2 
n  n  0
4(cos n  1)  8 / n 2 2 , n odd
 
n2 2
 0 , n even
 bn  0

Therefore,


f (t )  a0   an cos nt
n 1
 
 8  8 1
 0     2 2  cos nt   2  cos nt ♣
n 1  n   n 1 n
2

n odd n odd
Parseval’s Theorem
 Parserval’s theorem states that the
average power in a periodic signal is equal
to the sum of the average power in its DC
component and the average powers in its
harmonics
 Pdc
f(t) Pavg a0
2

=
t
Pa1 Pb1
a1 cos t b1 sin t

+ +
Pa2 P
a2 cos 2t b2 2t
b2 sin

+ + + …
 For sinusoidal (cosine or sine) signal,
2
Vpeak 
  2
2
Vrms  2 V
1 peak
P  
R R 2 R

 For simplicity, we often assume R = 1Ω,

which yields
1 2
P  Vpeak
2
 For sinusoidal (cosine or sine) signal,
Pavg  Pdc  Pa1  Pb1  Pa2  Pb2  
2
 a0  1 2 1 2 1 2 1 2
    a1  b1  a2  b2  
 2 2 2 2 2

1 2 1  2
 Pavg  a0   (an  bn )
2

4 2 n 1
Exponential Fourier series
 Recall that, from the Euler’s identity,
e  jx  cos x  j sin x
yields
e e
jx  jx
and e jx  e  jx
cos x  sin x 
2 j2
Then the Fourier series representation becomes
a0 
f (t )    (an cos nt  bn sin nt )
2 n 1
a0    e jnt  e  jnt   e jnt  e  jnt 
   an    bn  
2 n 1   2   j2 
a0    e jnt  e  jnt   e jnt  e  jnt 
   an    jbn  
2 n 1   2   2 
a0   an  jbn  jnt  an  jbn   jnt 
    e   e 
2 n 1  2   2  
a0   an  jbn  jnt   an  jbn   jnt
   e    e
2 n 1  2  n 1  2 
 an  jbn an  jbn
Here, let we name cn  , c n 
2 2
a0
and c0  . Hence,
2
a0   an  jbn  jnt   an  jbn   jnt
f (t )     e    e
2 n 1  2  n 1  2 
c0 cn c−n
 
 c0   cn e jnt   c n e  jnt
n 1 n 1
 
 c0   cn e jnt   cn e jnt
n 1 n  1
1  
 n
c e jnt

n  
 c0  n 
c e jnt

n 1
n
c e jnt

n  
Then, the coefficient cn can be derived from
an  jbn
cn 
2
T T
12 j 2

2T 0 f (t ) cos ntdt 
2T 0 f (t ) sin ntdt

1 
T T
   f (t ) cos ntdt  j  f (t ) sin ntdt 
T 0 0 
T
1

T  f (t )[cos nt  j sin nt ]dt
0
T
1

T 
0
f (t )e  jnt dt
 In fact, in many cases, the complex
Fourier series is easier to obtain rather
than the trigonometrical Fourier series
 In summary, the relationship between the
complex and trigonometrical Fourier series
are:
T
a0 1
T
1
c0  
2 T  f (t )dt cn 
T 
0
f (t )e  jnt dt
0

an  jbn
cn 
2
an  jbn
c n  or c n  cn
2
Example 8
Obtain the complex Fourier series of the following
function f (t )

e 2

e t
)=
f (t
1
t
 4  2 0 2 4
Solution
Since T  2 ,   1 . Hence
T
1
c0 
T  f (t )dt
0
2
1
  e dt
t

2 0

e 2  1

1 t
2
e   2
0 
2
 T
1
cn 
T 
0
f (t )e  jnt dt

2 2
1 1
e e e
t  jnt (1 jn ) t
 dt  dt
2 0
2 0
2
1 e 
(1 jn ) t
  
2  1  jn  0
e 2 (1 jn )  1 e 2 e  j 2 n  1 e 2  1
  
2 (1  jn) 2 (1  jn) 2 (1  jn)

since e  j 2 n  cos 2n  j sin 2n  1  0  1

 2 2
e 1 e 1
cn    c0
n 0
2 (1  jn) n 0 2

Therefore, the complex Fourier series of f (t) is

2
 
e  1 jnt
f (t )  c e
n  
n
jnt
 
n   2 (1  jn )
e ♣

*Notes: Even though c0 can be found by substituting

cn with n = 0, sometimes it doesn’t works (as shown
in the next example). Therefore, it is always better to
calculate c0 alone.
cn is a complex term, and it depends on nω.
Therefore, we may plot a graph of |cn| vs nω.

2
e 1
cn 
2 1  n 2

In other words, we have transformed the function

f (t) in the time domain (t), to the function cn in the
frequency domain (nω).
Example 9
Obtain the complex Fourier series of the function in
Example 1.
Solution
 
T 1
1 1 1
c0   f (t )dt   1dt 
T 0 20 2
T 1 2
1 1
 f (t )e dt   1e  jnt dt   0
 jnt
cn 
T 0
20 1
1
1e 
 jnt
j  jn
    (e  1)
2   jn  0 2n
 But e  jn  cos n  j sin n  cos n  (1) n

j  jn
Thus, cn  (e  1)
2n
j  j / n , n odd
 [(1)  1]  
n

2 n  0 , n even

*Here notice that cn n 0  c0 .


1  j jnt
Therefore, f (t )  c e
n  
n
jnt
 
2 n   n
e ♣
n0
n odd
The plot of |cn| vs nω is shown below
 1
1  , n odd
c0  cn   n
2  0, n even

0.5