Beruflich Dokumente
Kultur Dokumente
5.1 Convergence and divergence of a sequence. A partial sum of a series. The concept
of a power series
Introduction
In this chapter we will expand our discussion of series to the case where the terms
of the series are functions of the variable x . This chapter shows how functions that are
infinitely differentiable generate power series called Taylor series. Here we investigate
more general problems:
Which functions have power series representation?
How can we find such representation?
In many cases, these series can provide useful polynomial approximations of the
generating functions.
So the question:
What is the Series?
Why we need the series?
Answers:
A series is represented as a list of numbers with addition operators between them;
The integrals and differential equations studied in earlier courses often possessed
solutions expressible in terms of elementary functions such as polynomials,
exponentials, sins, and cosines. However, many important equations and integrals
arise that do not have solutions that can be so conveniently expressed which are
ease solve by series.
For example:
1 2 3 4
99 100
A series may finite or infinite. So, now I am going to talk about series.
1, 2 , 3 , 4 , 5 , ...
We start add these numbers
1 2 3 4 5 ...
this will be example for series. Another series like:
(*)
But, if you adding small enough and fast enough maybe it will infect not going to
infinity.
The same thing plays series. If the formula series is just lode different you can
have very difference thing to showing converge and diverges. One of the few types
series you can enter yes converge not diverges. Is example we have here
1 1
1
1
1
1 1 1 1 1 1 ...
10 10
10
10
10
This types series when you keep multiply term by one term. We were multiplying the
same number to get the next term.
2
1 1
1
1
1
1 1 1 1 1 1 ...
10 10
10
10
10
n 1
we get an expression of
the form
a1 a2 a3 ... an ...
(1)
Which is called an infinite series (or just a series) and is denoted, for short, by the
symbol
an or an .
n 1
1 2 3 4 5
Because if we start adding the terms we get the cumulative sums 1, 3 , 6 ,10 ,15, 21,
and, after the
n n 1
2
1 1 1 1 1
2 4 8 32 64
We get
1
n
2
1 3 7 15 31 63
, , , , ,
2 4 8 16 32 64
1
,1 n ,
2
In fact, by adding sufficiently many terms of the series we can make the partial sums as
close as we like to 1 . So it seems reasonable to say that the sum of this infinite series is
1 and to write
n
2 4 8 32 64
n 12
1
1 1
1
n
2
1.
We use a similar idea to determine whether or not a general series (1) has a sum.
We consider the partial sums
S1 a1 ,
S2 a1 a2 ,
S3 a1 a2 a3 ,
S4 a1 a2 a3 a4 .
And, in general,
n
Sn a1 a2 a3 a4 ... an ai .
i 1
Given a series
i 1
Sn ai a1 a2 a3 a4 ... an .
If the sequence
S n
i 1
ai
i 1
a1 a2
an
S or
ai S .
n 1
The number S is called the sum of the series. Otherwise, the series is called
divergent.
Actually we divided the series to two types:
1. Arithmetic series;
2. Geometric series.
Arithmetic series
The formula of arithmetic series is
a, a , a 2 , a 3 , , a n 1
If we denoted last term by L, and it is equal L a n 1 , the sum of series S for
n
2
a L .
n
a a n 1 2a n 1
2
2
(2)
For example:
2 4 8 16 32
This is a Geometric Series, because each term is equal to twice of the previous
term.
I will start about the Geometric series formula
a, a r , a r 2 , a r 3 ,
, a r n 1
n 1
Again, I denoted last term by L, and it is equal L a r
the sum of series S is:
n
a L
2
(*)
So now, how to find terms Geometric Series or Geometric Progression (GP). If the first
(initial) term of a GP is a and the common ratio of successive number is r , then
the n
th
term of GR is
Common ratio
an a r n 1
(2)
nth term
First term
3, 6, 12, 24,
48,
a1 3
a 3 21
3, 6, 12, 24, 48, , and this segment is Geometric 2
a3 3 22
segment. In this case common ration is equal r 2 . I
a4 3 23
rewrite the series, plug the numerical value.
a5 3 24
a6 3 25
For
segment,
so
we
have
segment
an a1 r n1
th
For example, what is the 8 term?
a8 a1 r 81 3 2 81 3 27 3 128 384
We can check through segments:
3,
6, 12, 24,
a1, a2 , a3 , a4 ,
48,
a5 ,
Example 7.1:
Given series 4 16 64 256 1024
th
. What is the 8 term in GS?
Solution:
Here, first term: a1 4 , common ratio: r 4 , number of term:
n 8 . Use the
formula (2) and plug the numerical value of each term, we will take
a8 a1 r 81 4 4 81 4 47 4 16384 65536
Examples for converges or diverges of GS.
The condition of converges, is common ratio must be between 1, 1 , so 1 r 1 .
GS is one of few types of series, when you can answer yes converge never diverges.
So we again go to our general series and write much more compactly. We can
rewrite this series using the summation.
n1
If plug
1
1
10
n 1
(3)
In GS 10 number said common ration and denoted with r . Basically GS says and
sometimes GS module of converge is r 1 (less than one). Another way number
between rise to the power should be between 1 r 1 . From ratio with anything all
you can conclude GS are diverges.
Again this is one of the a few types can answer converge or diverge. If you know
a
converges of GS 1 r ,
a~
first term
1 (common ratio)
(4)
Again use the formula (3) I recognize the question the series converge or diverge. The
condition of converge
rewrite the formula (4)
1 1 10
9
1 1 9
10 10
10
we want to do point out very useful before this one. Again we looked origin series (*)
1 2 3 4 5 ...
(*)
lim a
n n
(5)
so series is diverges. This is very useful result. You must be careful what
lim a
n n
(equal to zero)
(6)
to say converge or diverges. This same times said test for diverges.
For example, we have this series. 1 .4 .16 .64 .256
multiply by 0.4 , this will be my common ratio,
, so each terms
1.66
1 0.4 6
6 3
10
Example 7.2:
2
5
Given next series
3
n3
n 1
Solution:
5
If you just take first term 5, 2 is incorrect. So when
1
3
2
5
3
31
4 20
2
5 5
9 9
3
n3:
this now, first term of the series, and series again definitely converge. This is the basic
idea to get first term.
Example 7.3:
n
n 2 , what is the first term this series?
Given series
n 1 3
Solution:
1
n 1
3
n
n 2
n 2
9
9
3
3
3
3
n 1
n 1
n 1
n 1
Again we must to think about converge, so our common ratio is should be between
3
1, 1 ,
divergence, because r
1.
3
Example 7.4:
Given series
n2
2
.
n 1 2n 1
Solution:
n2
If just look
it like almost limit problem. Let me check limit this series.
2n 2 1
n2
1 n2
1
lim
lim
0,
,
we
can
divided
everything,
and
look
separately.
2
2
n 2n 1
n 2 n 1 2
this limit dose not equal to zero. So we can conclude the original series is diverges by
Diverge Test divergence.
I will talk about partial sums and limits of series. Some series
an a1 a2 a3 ... an S
n 1
If you arranges
a1 a2 a3 ... an Sn
(7)
S lim Sn
n
(equal to zero)
(7*)
1
ln
1
n
n 1
(8)
Converge or diverge?
a
Here again I use test for diverges. nlim
n
0,
n 1
the limit.
1
1
lim ln 1 ln lim 1 ln(1) 0
n
n
n n
Let to think breaking up (7)
n 1
ln
ln n 1 ln n
n
n1
n 1
ln 2 ln 1 ln 3 ln 2
ln 4 ln 3 ln 5 ln 4 ln 6 ln 5 ...
where ln(1) 0 , if we opening brackets
n 1
ln
ln n 1 ln n
n
n1
n 1
ln 2 0 ln 3 ln 2 ln 4 ln 3
ln 5 ln 4 ln 6 ln 5 ...
1
l
n
~ diverges.
n
n 1
So now in next section we will to talk about power series.
Introduction
Now I want to talk about power series. More important and useful are series
whose terms are functions of x .
Definition 7.2 (power Series):
More generally, a series of the form
a x
n
a0 a1 x a2 x 2 a3 x3
(9)
n 0
or
x c n a0 a1 x c a2 x c 2 a2 x c 3
n 0
(9*)
x x 2 x3
1
2 4
8
b)
x 2 x3 x 4
2
3
4
c)
x3 x5 x 7
3! 5 ! 7 !
d)
( x 2)
2
( x 2) 2
3
( x) n
;
2n
(1)n 1 x n
;
n
(1)n 1 x 2 n 1
;
2n 1 !
( x 2) n
n 1
n 0, 1, 2, 3,
n 0,
n 1, 2, 3,
n 0,
n 1, 2, 3,
5.2 Simple tests for convergence of a series. A radius of convergence of a power series
Whether a power series converges or not depends on the value of x we are considering.
We often use the ratio test to find the value of x for which a series converges.
We begin a brief review of the basic definitions and properties of power series
and real analytic functions.
n 0
n 1
x n converges.
Solution:
a) Use ratio test to find when the ration is 1 ;
b) Check endpoint separately.
a
n 1
a) By using the Ratio Test nlim
an
an 1
n 1
3n 11
n 1
n 1 x n 1
n 1 x n 1
a
lim n 1 lim
n an
n
3n 2
n xn
we have,
lim
3n 2
an
n
3n 1
n 1 x n 1
3n 1
3n 2
n xn
n xn
3n 1
Since x n 1 x n x1
Since 3n 2 3n 1 31
3n 1
lim
n 1 x
3n
n 1 x
3.
3 n n
3
x
lim
for x 3, 3 x 3 . So, the series converges absolutely for 3 x 3 and diverges for
x 3 (i.e., for x 3 or x 3 ). Since the Ration Test gives no conclusion for the
3
n 0
n 1
n
n
3
n 1
3
3
n 0
n 0
x n
n
3
0 , the series diverges by the nth term test for divergence. The power
Since nlim
series converges for all x in the interval 3, 3 and diverges for all x outside this
interval.
n! x
convergent?
n 0
Solution:
an1 n 1 ! xn1 ,
an n ! xn
n
Let, an , as usual, denoted the nth term of the series, then an n ! x . If x 0 ,
we have
n 1 ! xn1 Since n 1 ! n 1 n !
an 1
lim
lim
lim n 1 x
n an
n
n
n ! xn
Since x n 1 x n x1
By the Ratio Test, the series diverges when x 0 . Thus the given series converges
only when x 0 .
n 1
lim
x n 1
n 1 x
n
n
an 1
,
an
an 1
xn
n
x n 1
n 1
an
xn
n
x
x lim
,
multiply by sepricol nlim
n
n 1
n 1
x 1, 1 x 1 .
x 1:
1n
n
n 1
1
n
1
n
is
decreasing.
The same time we must to check value of positive 1 also.
x 1:
n
,
n 1
x lim
~ power series,
1
n 1 2
n
1
. So in conclusion our limit will be
2
Example:
x 5 n
n 0
n 1
, an n3 x 5
so limit
lim
n 13 x 5n 1
n
n3 x 5
n 1
lim
n n
x 5n x 51
x 5 n
n 1
lim x 5
n
n
n 1
x 5 lim
x 5 1;
n n
1 x 5 1;
4 x 1;
1 x 6 ;
x 4:
n
n 0
x 5
4 5
n 0
n3 0 ;
a) nlim
3
b) n ~ is decreasing.
n
n 0
but in our case n3 0 , and is not decreasing. These dose not convergence.
x 6:
x 5
n 0
6 5
n 0
n 0
n 0
x c n a0 a1 x c a2 x c 2 a2 x c 3
n 0
(9*)
Where x is a variable and the an is a constants. We say that (9*) converges at the point
x c if the infinite series (of real numbers) an x c n converges; thats the limit of
n 0
lim
an x c
n 0
this limit is exists (as a finite number). If this limit does not exist, the power series is
said to diverge at the point x c . Observe that (9) converges at x c since
an x c a0 0 0
n
a0
n 0
n 0
The number in case (iii) Theorem 7.1 above is called the radius of
convergence of the power series, such that (9). By convention, the radius of
convergence is 0 in case (i) and
in case (ii).
The interval of convergence of a power series is the interval that consists of all
values of x for which the series converges.
In case (i) the interval consists of just a single point
c.
xc
can be rewritten as
c xc.
When
series might converge at one or both endpoints or it might diverge at both endpoints.
Notice that, Theorem 1 settles the question of convergence except at the endpoints
divergence for x c
?
convergence for x c
Radius of convergence
Interval of convergence
1, 1
n! x
2, 4
x
n 0
n 0
n 0
n 1
x2n
x 3
lim
an 1
L,
an
n 0
x c n is
1
, with if L 0 and 0 if
L
L.
Example 7.7:
n 0
n 1
x 3
(10)
Since an
n 1
, we have
n 1
2 n 1
2 n 1
a
lim n 1 lim
lim
2L.
n an
n
2 n n 2 n n 2
By the ratio test, the radius of convergence is 1 2 . Hence the series (10) converges
absolutely for x 3 1 2 and diverges when x 3 1 2 . It remains only to determine
what happens when x 3 1 2 , that is, when x 5 2 and x 7 2 .
Set x 5 2 , and the series (10) becomes the harmonic series
n 1 , which is
n 0
known to diverge. When x 7 2 , the series like (10) becomes an alternating harmonic
series, which is known to converge.
n 0
x c n converges, we get
a number that is the sum of the series. It is appropriate to denote this sum by f ( x ) ,
since its value depends on the choice of
x . Thus we write
f ( x) an x c
n 0
f ( x)
x
n 0
1
, that is
1 x
1
1 x1 x 2 x3
1 x
x
n 0
for x 1,
1 x 1
(11)
Example 7.8:
For example, given two power series:
a) f ( x)
a x c
n
n 0
b) g ( x)
, (**)
b x c
n
n 0
with nonzero radii of convergence, we want to find power series representations for the
sum, product, and quotient of the function f ( x ) and
g ( x) .
Solution:
a)
f ( x) g ( x) an bn x c
n 0
for all
b)
f ( x) g ( x) d n x c
n 0
were
dn an bn .
If f ( x) an x c
n 0
, then f is
xc .
n an x c
n 1
n 1
for
Furthermore, term wise integration gives the power series for the integral of f :
(ii)
f ( x)dx A a x c a
0
x c
2
a2
x c
3
A an
n 0
x c
n 1
n 1
for x c .
Example 7.9:
Express
1 x2
Starting with GS (11) for 1 , find a power series for each of the following functions:
1 x
a)
1 x
b)
1 x
Solution:
(a)
n
1
1
n 2n
2
1
x 1 x 2 x 4 x 6 x8
2
2
1 x 1 x n 0
n 0
2
Because this is a geometric series, it converges when x 1 , that is, x 2 1 , or x 1 .
(b)
Notice that
1 x
1 x
1 2 x 3x 2 4 x3
n x n 1
1
. Hence on
1 x
5.3 Taylor series. Maclaurin series for standard function. Bessels functions. Fourier
trigonometric series
TAYLOR SERIES
Taylor polynomials
Taylors remainder
Taylor series
Taylor Polynomials
Let f be a given function and let
arbitrary many derivatives at
c ). Let
Pn x a0 a1 x a2 x 2
an x n
(12)
c . If we write Pn x
Hence, we write
Pn x a0 a1 x c a2 x c
2
an x c
(13)
an yet to be determined.
Pn c f c
Pn c a0 f c .
Pn' c f c
Pn' c a1 f ' c
''
''
Equate the second derivatives. Since Pn c 2a2 f c we must have a2
1 ''
f c .
2
a3
1 '''
1
1
f c , a4
f ( 4) c and, in general an f ( n ) c .
23
2 3 4
n!
Pn x f c f ' c x c
1 ''
2
f c x c
2
1 ( n)
n
f c x c
n!
(14)
0
If we write f c f c , then
f (n) c
n 0
n!
Pn x
x c
(15)
The polynomial (15) is called Taylor Polynomial of degree n for function f about c .
Example 7.13:
Find the Taylor polynomial of degree 3 for the function f x sin 2 x about x
Solution:
We must compute the first three derivatives of f x about
f x sin 2 x
f ' x 2 cos 2 x
f '' x 4 sin 2 x
f ''' x 8 cos 2 x
:
2
f sin 0 ;
2
f ' 2 cos 2 ;
2
f '' 4 sin 0 ;
2
f ''' 8 cos 8 .
2
Hence, we get
Pn x a0 a1 x c a2 x c
an x c
8
0 2 x 0 x
2
3!
2
4
2 x x .
2 3
2
3
To find an approximation of f x sin by using P3 x , we get
4
3
3
sin 2 x sin x
8
4
.
2
3
P3 x P3
8
3 4 3
8 2 3 8 2
3
0.7074
4 384
3
The actual value for sin is 0.7074 .
4
Taylors Remainder:
Rn x f x Pn x
(16)
f x Rn x Pn x
expresses f x exactly in terms of the polynomial and the remainder. Thus, Rn x is
the error in approximating f x and Pn x .
Taylors Theorem:
Theorem 7.1 (Taylors Theorem)
f (k ) c
k 1
k!
f ( n 1) t
f x
Then,
Rn x
c
n!
x c
x t
Rn x
dt for all x a , b
Example 7.14:
Estimate the error in approximating sin 2x by Pn x .
Solution:
(17)
Rn x
f ( n 1) t
n!
x t
dt
If f t sin 2 x , then f
For both case above, f
n 1
n 1
t 2 n1 sin 2t
or f
t 2 n1 , because of
n 1
t 2 n1 cos 2t .
2( n 1) x t n 1
2
2( n 1)
n
Rn x
x
x t dt
n!
n 1
n 1!
2
n !
2
2
x
( n 1)
( n 1)
Taylor Series
The Taylor Polynomial Pn x
f (n) c
n 0
n!
Pn x
x c
th
f (n) c
n 0
n!
x c
(18)
c.
If the Taylors Remainder Rn x converges to 0 for all those value of x , then the
Taylor series for f about
converges to f x , or
f (n) c
n 0
n!
f x
Example 7.15:
x c
(19)
Solution:
f x ex
f ' x e x f '' x
f (n) x ;
f ( n ) 0 e0 1;
f (n) 0
n 0
n!
x 0
1 n
x 2 x3
x 1 x
2! 3!
n 0 n !
Example 7.16:
Construct the Taylor series for function f x ln x about 1 .
Solution:
f x ln x
1
f ' x ,
x
1
2
f '' x 2 , f ''' x 3 ,
x
x
(k )
k 1
f 1 1 k 1!.
f ( n ) x 1 k 1
k 1! ;
xk
1 n 1
n 0
n!
n 1
x 1
n 0
1
n
x 1
n 1
x 1
1
1
2
3
x 1 x 1
2
3
f x = x 1
f ' x = x 2
f '' x = 2!x 3
f ''' x = 3!x 4
(n)
x = 1
1
f 2 = 21 = ;
2
1
f ' 2 = 2 ;
2
f '' 2
1
= 23 = 3 ;
2!
2
f ''' 2
1
= 4;
3!
2
f ( n ) 2 1
= n1 .
n!
2
n
n!x
( n 1)
f '' 2
2
f 2 f 2 x 2
x 2
2!
'
f (n) 2
n
x 2
n!
1 x 2 x 2
n x 2
= 2
1
.
2
2
2
2
2n1
This is a geometric series with first term 1 2 and ratio r = x 2 2 . It converges
2
n 0
f ( n) 0
n!
xn
x 2 x3 x 4
e 1 x
2! 3! 4!
xn
;
n 0 n !
1 x 2 n 1
;
n 0 2n 1 !
n
1 x 2 n
x 2 x4 x6
cos x 1
;
2! 4! 6!
2
n
!
n 0
n 1
1 x n
x 2 x3 x 4
ln 1 x x
.
2!
3!
x3 x5 x7
sin x x
3! 5! 7 !
4!
n 1
Example 7.17:
Approximate sin
1
to four significant figures (three decimal places) using the
2
1 x 2 n 1
;
2
n
1
!
n 0
0.5211 ,
y'
1
,
x y 1
y(0) 0 .
(20)
Solution:
Lets assume the solution to this initial value problem is analytic at x 0 . Then
we can write
y n (0) n
y ( x)
x .
n
!
n 0
(21)
''
''
(21) with respect to x , thereby getting an expression for y ( x) in terms x, y ( x) and
1
x y x 1
1 y ' x
(22)
'
Substituting x 0 , y (0) 0 and y (0) 1 in equation (22), we obtain
y '' (0)
1
12
1 1) 2 .
y ''' (0)
1 y ' x x y x 1
x y x 1
2
2
2
y ''' (0) 3 1 1 1 2 10 .
1
3
y '' x ,
Repeating this procedure, we can determine y (0) for any value of n , although the
n
y ( x) x x 2
5 3 7 5
x x
3!
5!
BESSELS FUNCTIONS
coordinate of ) is
(1)
y r sin r sin t
(2)
Figure 1
The back and forth motion of is called simple harmonic motion. By
definition, an object is executing (ispolnyt) simple harmonic motion if its displacement
from equilibrium (ravnovesie) can be written as
r sin( t ) , but these two functions differ from r sin t only in choice of origin;
such functions are called sinusoidal function). You can think of many physical
examples of this sort of simple vibration: a pendulum (mayatnik), a tuning fork (), a
weight bobbing up and down at the end of a spring.
The x and y coordinates of particle in Figure 1 are
x r cos t ;
y r sin t
(3)
z x iy r cos t i r sin t
r cos t i sin t
re
it
re
(4)
It is often worth while to use this complex notation even to describe the motion of :
we then understand that the actual position of is equal to the imaginary part of z .
For example, the velocity of is the imaginary part of
dz d
(5)
dy
The imaginary part of (5) is r cos t , which is dt from (2). It is useful to draw a
Figure 2
The period of the harmonic motion or the period of the functions is the time for
one complete oscillation, that is
(Figure 2).
dy
r cos t cos t
dt
(6)
Here is the maximum value of the velocity and is called the velocity amplitude. Note
that the velocity has the same period as the displacement. If the mass of the particle
is m , its kinetic energy is:
Example 1:
The number of hours of daylight at any particular location changes with the time
of year. The table shows the average number of hours of daylight for approximately
twoyear period.
Note: Day 15 is January 15 of year 1.
Day 74 is March 15 of year 1.
Day 411 is February 15 of year 2.
Day
15
Hrs of 6.7
74
135
166
258
349
411
470
531
561
623
653
684
714
11.7
17.2
18.8
12.9
5.9
9.2
14.6
18.8
18.1
12.9
10.2
7.5
5.9
Light
6 18
12 .
2
The
How many hours of daylight will there be on August first of the third year?
6 1
2.5 .
2
The
1 (6) 7
3.5 cm .
2
2
Periodic function:
Definition: A function whose graph repeats at regular intervals; the y values in the
table of values show a repetitive pattern when the x values change by the
same increment.
Lets look the basic sin function.
A periodic function is a function whose graph repeats itself identically from left to tight.
A periodic function can be repeated by f ( x) f ( x p) for all x .
The period of a function is the horizontal distance required for a complete cycle.
Actually have special name for the distance , its called period. We can see for sin
function, period going from zero to 360 degree, so basic sin function have period 2 .
Also for cos function is 2 . For the tan function the period is . We just focusing
for sin and cos .
1
We also need to talk about amplitude of function. The Amplitude of a function is the
half the distance between the maximum and minimum values of a periodic function. We
can also think about radius of periodic function.
And remember the Amplitude is always positive. We have negative radius, but with
positive amplitude. We can shift this function not just horizontally, but vertically
The vertical shift of a function is the vertical shift of a periodic function along y - axis
up or down.
Fourier Series
When the French mathematician Joseph Fourier (17681830) was trying to solve
a problem in heat conduction, he needed to express a function f as an infinite series of
sine and cosine functions:
f ( x) a0
an cos nx bn sin nx
n
1
(1)
Earlier, Daniel Bernoulli and Leonard Euler had used such series while
investigating problems concerning vibrating strings and astronomy.
The series in Equation 1 is called a trigonometric series or Fourier series and it
turns out that expressing a function as a Fourier series is sometimes more advantageous
than expanding it as a power series. In particular, astronomical phenomena are usually
periodic, as are heartbeats, tides, and vibrating strings, so it makes sense to express them
in terms of periodic functions.
We start by assuming that the trigonometric series converges and has a
continuous function f ( x ) as its sum on the interval [ 1, 1 ] , that is,
f ( x) a0
an cos nx bn sin nx
x .
(2)
Our aim is to find formulas for the coefficients an and bn in terms of f . Recall
that for a power series f ( x) cn x a we found a formula for the coefficients in
n
terms of derivatives:
f n ( x)
cn
. Here we use integrals.
n!
an cos nx bn sin nx dx
f ( x) dx a0 dx
n
1
2 a0
an
But
cos nx dx
bn
n
1
sin nx dx
1
1
cos nx dx sin nx
sin n sin n
n
n
x
1
sin n sin n 0 .
n
Because
is an integer. Similiraty,
sin nx dx 0 . So
f ( x) dx
(3)
0 an cos nx bn sin nx
n 1
f ( x) cos mx dx
a0
an
cos mx dx
n
cos mx cos nx dx
cos mx dx
bn cos mx sin nx dx .
(4)
cos mx sin nx dx 0
for n m ,
0
cos mx cos nx dx
for n m .
f ( x) cos mx dx am
Solving for am and then replacing m by n , we have
an
f ( x) cos nx dx,
n 1, 2, 3,
(5)
1
sin n m x dx
2
0 n m x dx
1
1
1
1
cos n m x
cos n m x
2 nm
2 nm
0
0
1 1
cos n m 2 cos n m 0
2 n m
1 1
cos n m 2 cos n m 0
2 n m
1 1
1 1
1 1
1 1 0 .
2 nm
2 n m
to , we get
bn
f ( x) sin nx dx,
n 1, 2, 3,
(6)
a, b
removable or jump discontinuities. (In other words, the function has no infinite
discontinuities. See Section 2.5 for a discussion of the different types of discontinuities.)
Definition 1:
and
f ( x) dx ,
an
f ( x) cos nx dx ,
bn
f ( x) sin nx dx
is a
Example 1:
Find the Fourier coefficients and Fourier series of the square-wave function f
0
defined by f ( x) 1
if
if
x 0
and
0 x
f ( x 2 ) f ( x) . So f is periodic
Solution:
We found the Fourier series of the squarewave function, but we dont know yet
whether this function is equal to its Fourier series. Lets investigate this question
graphically. Figure 2 shows the graphs of some of the partial sums
Sn x
1 2
2
sin x
sin 3x
2
3
2
sin n x
n
f a lim f x ,
xa
f a lim f x
xa
series is the average of the right and left limits, that is 2 f x f x
xa
xa
and similarly for the other points at which f is discontinuous. The average of these left
and right limits is
1
2
f ( x) if n n ,
1
2
sin 2k 1 x 1
2 k 1 2k 1
if x n .
x
L
and
Lt
g t f x f
then, as you can verify, g has period 2 and x L corresponds to t . The
Fourier series of g is
a0
an cos n t bn sin n t
n1
where
g (t ) dt ;
1
a0
2
an
g (t ) cos n t dt ,
Lt
bn
, then t
x
L
g (t ) sin n t dt
, dt
dx , and we have
the following:
If f is a piecewise continuous function on L, L , its Fourier series is
a0
n x
n x
bn sin
L
L .
an cos
n 1
(7)
where
1
a0
2L
L f ( x) dx
And, for n 1 ,
1
an
L
n x
f x cos
dx ,
L
1
bn
L
n x
f x si n
dx
L
(7*)
Of course, the Fourier Convergence Theorem (1) is also valid for functions with
perio 2L .
Example 2:
Calculate the Fourier series of
1,
f x 0,
1,
for 0 x ,
for
x 0, ,
for x 0,
with
for all x .
Solution:
Our f has a period 2L 2 and so L .
f ( x) dx 0 ,
so
an
f x cos n x dx 0 ,
also
bn
2
2
cos n x 0
cos n x 1
n
n
0 , n ~ even ,
2
n
4
, n ~ odd .
n
n
Thus
f x f x 2
S f x
n ~ odd
4
sin n x
n
4
sin(2k 1) x.
(2
k
1)
k 1
Example 3:
Find the Fourier series of the triangular wave function dened by f x x for
f x on the interval
a, b
f x
of
on
a, b
is
approximately equal to
y=f(x)
f x1 f x2
f xn
(1)
n
x1=a
x2
x3
Figure 3.
xn=b
be x apart.
Multiply the numerator and the denominator of the approximate average by x . Then
(1) becomes:
f x1 f x2
f xn x
(2)
n x
Now n x b a , the length of the interval over which we are averaging, no matter
what n and are x . If we let n and x 0 , the numerator approaches
b
f ( x) dx , and we have
a
Average of f ( x) on a, b
f ( x) dx
a
ba
(3)
In applications, it often happens that the average value of a given function is zero.
For example, the average of sin x over any number of periods is zero. The average
value of the velocity of a simple harmonic oscillator over any number of vibrations is
zero. In such cases the average of the square of the function may be of interest. For
example, if the alternating electric current (a-c) flowing through a wire is described by a
sin x function, the square root of the average of the sin x squared is known as the root-
mean-square or effective value of the current, and is what you would measure with an ac ammeter. In the example of the simple harmonic oscillator, the average kinetic energy
(average of
1
1
m 2 ) is m times the average of
2
2
2 .
2
Now you can, of course, find the average value of sin x over a period (say
yourself that the area under them is the same for any quarter period from 0 to
to , etc.
,
2 2
cos 2
sin 2
Figure 5
Then
sin
x dx
cos
x dx
(4)
sin nx dx
2
cos 2 nx dx
(5)
2
2
But since sin nx cos nx 1 ,
sin
nx cos nx dx
2
dx 2
(6)
sin nx dx
2
cos 2 nx dx
(7)
1
2
1
sin nx dx
2
2
cos 2 nx dx
2 2
(8)
2
We can say all this more simply in words. By (5), the average value of sin nx equals
2
2
2
the average value of cos nx . The average value of sin nx cos nx 1 is 1.
2
2
Therefore the average value of sin nx or cos nx is
1
2
sin x and cos x have period 2 : so do sin nx and cos nx for any integral n since
sin n x 2 sin nx 2n sin nx . Then, given a function
f x
of period 2 ,
a0
a1 cos x a2 cos 2 x a3 cos 3x
2
b1 sin x b2 sin 2 x b3 sin 3 x
(1*)
1
sin m x cos n x
2
sin mx cos nx dx 0 ,
(9)
sin m x sin n x
sin mx sin nx dx
2
0 , when m n ,
1
, when m n 0 ,
2
0 , when m n 0.
(10)
cos m x cos n x
cos mx cos nx dx
2
0 , when m n ,
1
, when m n 0 ,
2
1, when m n 0 .
2
2
We have already shown that the average value of sin nx or cos nx is
(11)
1
2
. The last
integral in (11) is the average value of 1 which is 1. To show that the other average
values in (9), (10), and in (11) are zero, we could use the trigonometry formulas for
products like sin cos and then integrate.
ei cos isin
e
cos isin
(13)
This two equations can be solved for sines and cosines , we get
einx einx
cos nx
2
einx einx
sin nx
.
2
(14)
If we substitute equations (14) into a Fourier series like (12) we get a series of
terms of the forms e
inx
and e
inx
also find the complex form directly; this is often easier than finding the sine cosine
form. We can then, if we like, work back the other way and [using Eulers formula]
ei cos isin
z x iy r (cos isin ) r e
Eulers formula
(15)
Cn einx
(16)
einx
inx
e dx in
x
x
1 in in ein ein
e
e
0
in
in
inx
on ,
is zero when
(17)
is an integer not
equal to zero.
To find C0 , we find the average values of the terms in (16)
1
1
f x dx C0
dx
2
2
average values of terms of the
C0 0
inx with n an int eger 0
f
orm
e
(18)
and
1
C0
f x dx
2
So to find Cn , we multiply (16) by e
(19)
inx
each term. [Note: the minus sign in exponent. In finding an , the coefficient of cos nx
inx
inx
.]
1
inx dx C 1
inx dx C 1
f
x
e
e
einx eix dx
0
1
2
2
2
1
C1
einx eix dx
2
Then we get
1
inx dx
Cn
f
x
e
2
Note that this formula contains the one for C0 (no
(20) is valid for negative as well as positive
(20)
1
2 to worry about here!). Also, since
Example 4:
Let us expand the same
0 , when x 0 ,
0 x .
1, when
f x
Solution:
We have from (20)
1
Cn
2
0 e
inx
1 einx
2 in
1
dx
1 einx dx
2
0
1
, n ~ is odd ;
1
inx
e
1 in
2 in
0, n ~ is even 0,
(21)
1
C0
2
dx
1
.
2
Then
f x Cn e
inx
1
3
5
i 1
3
5
1 1
2 i
(22)
It is interesting to verify that is the same as the sinecosine series we had before. We
could use Eulers formula for each exponential, but it is easier to collect terms like this:
2
2i
3
2i
5
2i
1 2
1
1
sin x sin 3x sin 5 x
2
3
5
(23)
2 1
3
5
(24)
In lecture we will took about Fourier Transform in Pair complex number express
in Polar coordinates
In Cartesian (Rectangular) Coordinates, any complex number Ck can be
expressed as:
Ck Rk i I k
Remember again the definition, that i 1 . However if we express the
2
A time e rest to the power i tetra, ~ angle tetra is the angle or some people called
first angle. Now official according e rest to the power i tetra time A, is the same
Horizontal axis you have here, represent real component of the complex number, let we
say Rk and vertical axis represent imaginary component of complex number I k , so if
you have any C.N. it can be represented by, let me say any complex number Ck or Ck
(C tuta), that can be represented by the the amplitude A , which is like show another
radius of this sogo with amplitude A and also express by first angler tetra ( ). So in
polar coordinate the complex number Ck can be express in term of the amplitude A ,
which radius in this case and first angle . On the hand the same complex number if
you express in Cartesian coordinate, then that complex number see can be by the reyou
component which Rk and imaginary component which is I k . So complex number we
can be express in the polar and the Cartesian coordinates.
Thus, one obtains the following relations between the Cartesian and polar
coordinate
systems,
Rk A cos
and
similarity
imaginary
component
be express in term of polar component such as amplitude A and . And the two system
rectangular and polar coordinates they are related by two equation
Rk A cos , I k A sin
(25)
Ok according to the definition, we can say Rk2 I k2 equal to what? From the previous
picture you can see the real part Rk2 plus the imaginary part I k2 should be equal to A2 . Is
many fact if you looked to formula (1), what will be
=1
(26)
From this we can be say what Rk I k A or we can say the amplitude (is
2
Rk2 I k2 .
If you we know the real part Rk and the imaginary part I k of the complex
number then we can find of the amplitude A very easily. Now next, how to get first
angle tetra . Well again it is very easy, according previous formula (25) the angle, we
computed best on Rk divide by A or can be best on I k divide by A . So from formula
(25) we can find angle . And exactly what I show here:
cos
Rk
R
cos 1 k
A
A
(27)
sin
Ik
I
sin 1 k
A
A
(28)
General Interval
The functions sin nx and sin nx and einx have period 2 . We have been
considering , as the basic interval of length 2 . Given f x on , , we
have first sketched it for this interval, and then repeated our sketch for the intervals
, 3 , 3 , 5 , 3 , ,
length 2 , any one of which could serve as the basic interval. If we are given f x on
any interval of length 2 , we can sketch f x for that given basic interval and then
repeat it periodically with period 2 . Recall that in evaluating the Fourier Coefficients,
we used average values over a period. The formulas for the coefficients are then
unchanged (except for the limits of integration) if we use other basic intervals of length
2 . In practice, the intervals , and
For f x defined on
0, 2
0, 2
read
an
f ( x) cos nx dx ,
bn
f ( x)sin nx dx ,
1
1
cn
2
f x e
inx
dx
(29)
f(x)
f(x)
Figure 6
Figure 7
, 2 ,
you can easily verify by sketching that the answer from either
sin
n
n x
n x
2n sin
x 2l sin
.
l
l
l
n x
Similarly, cos
and e
l
n x
l
replaced by
f ( x)
a0
x
2 x
3 x
a1 cos
a2 cos
a3 cos
2
l
l
l
x
2 x
3 x
b1 sin
b2 sin
b3 sin
l
l
l
a
0
2
f ( x)
in x
e l
n 1
cos
n x
n x
bn sin
,
l
l
(30)
We have already found the average values over a period of all the functions we
need to use to find an , bn , and cn here. The period is now of length 2l l to l , so
in finding average values of the terms we replace
1
2
1
by 2l
.
1
Recall that the average of the square of either the sine or the cosine over a period is
2
and the average of e
in x
l
in x
l
coefficients become
1
an
l
n x
1
f ( x) cos
dx , bn
l
l
n x
1
f ( x)sin
dx , cn
l
2l
f x
in x
e l dx
(31)
For the basic interval 0, 2l we need only change the integration limits to 0 to 2l . the
Direchlet theorem just needs replaced by l in other to apply here.
Example 5:
Given
0 x l,
l x 2l .
0 , when
f x
1, when
Expand
f x
in an exponential
Solution:
First we sketch a graph of f x repeated with period 2l (Figure 8). By
equations (31), we find
f(x)
1
-2l
-l
2l
Figure 8
We have from (20)
3l
1
1
Cn
0 dx
1 e
2l
2l
0
in x
l dx
in x
e l
1
2l in x
l
2l
1
e 2in e in
2in
1
,
1
in
1 e
in
2l
0,
1
C0
2l
2l
dx
when n is odd;
(32)
when n is even 0,
l
1
.
2l 2
Then
i x
i x
3i x
3i x
5i x
5i x
1 1 l
1 l
1 l
1 l
1 l
l
f x e e
e
e
e
e
2 i
3
3
5
5
1 2
i x 1
3 x 1
5 x
sin
sin
sin
.
2
l
3
l
5
l
(33)
Perceval theorem
We shall now find a relation between the average of the square (or absolute
square) of f x and the coefficients in the Fourier series for f x , assuming that
2
f x dx
relation.
You should understand that the point of the theorem is not to get average of the
square of a given f x by using its Fourier series. [Given f x , it is easy to get its
average square just by doing the integration]. The point of the theorem is to show
relation between the average of the square of f x and the Fourier coefficients. We
can derive a form of Parsevals theorem from any of the various Fourier expansions we
have made; let us use
f x
f x
a0
a1 cos x a2 cos 2 x a3 cos 3 x
2
b1 sin x b2 sin 2 x b3 sin 3 x
a0
an cos nx bn sin nx
2 n1
n 1
(34)
f x dx
2
(35)
When we square f x , we get many terms. To avoid writing out a large number of
them, consider instead what types of terms there are in f x and when the averages
2
The average of
an cos nx 2
The average of
b0 sin nx 2
1
we have:
2
2
1
is a0 ,
2
1
is an2 ,
2
1
is nn2 .
2
1
2
(36)
1
2
Then there are cross-product terms of the forms 2 a0 an cos nx , 2 a0bn sin nx ,
and 2anbn cos nx sin mx , with m n (we write n in the every cosine factor and m in the
sine factor since every sine term must be multiplied times every cosine term).
So by (5.2), the average values of terms of all these types are zero. Then, we have
1 1 2 1 2
The average of f x (over a period) a0 an bn .
2 n 1
2 2 n 1
2
(37)
Problem 1:
Prove (4) for a function of period 2L expanded in a sine - cosine series.
Solution (Proof):
You can also verify (Problem 2, see below) that if f x is written as a complex
exponential Fourier series, and if in addition we include the possibility that f x itself
may be complex, then we find:
The average of f x (over a period)
2
(38)
Problem 2:
If f x is complex, we usually want the average of the square of the absolute
value of f x . Recall that
in x
e L
Solution (Proof):
It seems plausible physically, and it can be proved mathematically, that with one or
more harmonics left out, we would not be able to represent sound waves containing the
omitted harmonics. We say that the set of function sin nx , cos nx is a complete set of
functions on any interval of length 2 ; that is, any function (satisfying Dirichlet
conditions) can be expaned in a Fourier series whose terms are constants times sin nx
and cos nx .
Example:
In an application of sound we said that the intensity (energy per square centimeter per
second) of a sound wave is proportional to the average value of the square of the excess
pressure.
Solution:
1 sin 524 t 30sin(524 2 t ) sin(524 3 t )
4
1
2
3
sin(524 5 t ) 30sin(524 6 t ) sin(524 7 t )
5
6
7
P t
(39)
If for simplicity we write (39) with letters instead of numerical values, we have
P t b sin 2 nft
n 1
(40)
2
1
P t bn2 the average of bn2 sin 2 2 nft .
2
n 1
n 1
Now the intensity or energy of the sound wave is proportional to the average of
P t , and the energy associated with the nth harmonic is proportional to the average
2
of bn sin 2 nft . Thus Parsevals theorem says that the total energy of the sound
2
wave is equal to the sum of the energies associated with the various harmonics.
Example:
Let us use Parsevals theorem to find the sum of an infinite series. Let f x x
on the interval 1 x 1 . Sketch the corresponding function of period 2 and expand it
in a complex exponential Fourier series of period 2 .
Solution:
Form problem (8.15) we get:
The function f x x of period 2 which is equal to x on 1, 1 :
i
1
1
1
1
ei x ei x e2i x e2i x e3i x e3i x
x
2
2
3
3
1
1
1 x2
1
2
The average of f x x dx .
2 1
2 3 1 3
2
cn
, so we have
1
1
1 1 1 1
2
cn 2 1 1
3
4 4 9 9
2 1
2 2 .
n1 n
1 2 1 2
2
2
3 6
n
n 1