EE102 – SYSTEMS & SIGNALS
Download
Report
Transcript EE102 – SYSTEMS & SIGNALS
Review of methods for LTI systems
• Time domain methods (differential equations,
convolutions). Apply to all cases, but may be
cumbersome to compute.
• Laplace transforms. For systems starting at time t=0;
useful to study transients, instability.
• Fourier series. For studying stable systems in steady
state (after transients die down). Restricted to
periodic inputs. Key idea: superposition of sinusoids.
• Question: can we extend this steady-state analysis via
sinusoids to non-periodic inputs?
• Answer: Fourier transforms.
Fourier series as the period increases.
Consider the series f (t )
F e
n
in0t
n
, for period T
The frequencies involved in the expansion are n0 n .
20
0
0
0
2
0
20
Now consider a function with twice the period, T 2T . Then
0
0
2
. The frequencies involved in the expansion are n0
20
0
0
0 20 30
As the period gets longer, we get a more dense set of
frequencies. In the limit: include all frequencies.
n
Fourier Transform
We wish to extend the Fourier series concept to a non-periodic f(t).
Intuitively: f(t) takes infinitely long to repeat itself, so we think of
f
it as a function of infinite period.
This suggests replacing
1
Fn
T
T
2
t
f (t ) ein0t dt by an integral where T .
T
2
Definition: The Fourier transform of time-domain function
f (t ), t , is the function of frequency ,
defined by the integral F (i ) F f (t ) f (t )ei t dt ,
assuming it converges.
Example I: f (t ) e
F (i )
0
t i t
e e
t
F (i )
.
dt
0
et eit dt et eit dt
e(1i )t dt e(1i )t dt e
0
f (t )eit dt
0
(1i )t
1 i
0
e
(1i )t
(1 i )
0
(1 i ) (1 i )
2
1 i 1 i
(1 i )(1 i )
1 2
1
1
t
t it
Example II: f (t ) e u(t ). F (i ) e e
0
dt
1
1 i
Note: in case II, the Fourier transform F (i ) coincides with
the Laplace transform F ( s) evaluated at s i .
Representation of the Fourier transform
F (i ) is a complex function of the real variable . The "i "
in the argument of F (i ) plays no role, and could be omitted.
We only put it in for compatibility with Laplace transforms,
as in example II above (more on this later).
iF ( )
Magnitude and phase representation, F (i ) F (i ) e
Example:
F (i )
F (i )
F ( i )
1
1 i
1
2 1
F ( ) Arctan( ).
F ( )
Bode Plots
Represent the frequencies in a logarithmic scale
(i.e. proportional to log( ).).
Also plot the magnitude in "decibels (dB)": 20log F (i )
20log F (i )
F ( )
Widely used, but we will not emphasize them in this course.
The inverse Fourier transform
In Fourier series, we can reconstruct the function from
in0t
the Fourier coefficients via f (t )
Fn e
.
n
Similarly, for Fourier transforms we can reconstruct
the function using the inverse formula
1
f (t )
2
F (i )eit d
A complete derivation is mathematically involved, but we
can sketch a proof based on Fourier series case, letting the
period go to infinity.
F (i )
f (t )e
T
T
Let FT (i )
2
T
it
2
1
dt f (t )
2
f
T
F (i )eit d
t
2
f (t )eit dt . Then F (i ) lim FT (i )
T
T T
Considered on the interval t , , f (t ) (if "well-behaved")
2 2
in0t
has a Fourier series expansion f (t )
Fn e
, where
2
n
T
1
1
0
in0t
Fn
f ( t )e
dt FT (in0 )
FT (in0 )
T T
T
2
2
2
0
in0t
T T
So we have f (t )
FT (in0 ) e
, for t ,
2 2
n 2
f (t )
1
i t
F (i )e
2
0
in0t
d
F
(
in
)
F
(
in
)
e
T 0
0
n 2
0
1
in0t
F (in0 ) e
2
n 2
As T :
(I)
F (i )ei t d
( II)
Term (I) goes to zero since lim FT (in0 ) F (in0 )
T
Term (II) goes to zero as shown in the next slide.
1
it d
So the left-hand side must be zero: f (t )
F
(
i
)
e
2
Also, as T all t 's will be included in the interval.
1
Term (II):
2
n
in0t
F (in0 ) e
1
0
2
F (i )eit d
F (i )eit
0
n0
The sum on the left is a staircase approximation to the
integral on the right. As T , 0 0 and the
approximation becomes exact, provided the integral exists.
So the difference tends to zero.
Recap
Fourier Series
Fourier Transforms
(T-periodic functions)
1
Fn
T
f (t )
T
f (t ) e
in0t
dt
F (i )
0
in0t
Fn e
n
1
f (t )
2
f (t )eit dt
F (i )eit d
Note that the Fourier transform and its inverse are formally
almost identical: except for a sign change and the factor
2 , t and could be interchanged. This is called duality.
Example: F (i ) u( B) u( B).
F (i )
1
B
1
f (t )
2
1
i
t
F (i )e d
i t
1 e
2 it
B
2
B
B
e
2 it
1
iBt
B
i t d
e
B
e
iBt
sin( Bt )
t
sin(t )
Notation: we define the function sinc(t )
t
B
Then we have f (t ) sinc(Bt )
sin(t )
Plot of sinc(t )
t
sinc(t )
1
t
1
t
sin(t )
Note that sinc(0) lim
1
t 0
t
Example: f (t ) (t ).
F (i )
it
(
t
)
e
dt 1 for all .
1
i t d
The inverse formula would say that (t )
e
2
Now the function ei t has magnitude 1 for all .
Not an absolutely convergent integral!
We can interpret it as
B
1
i t d
(t ) lim
e
B 2
B
B
lim sinc(Bt )
B
B
sinc(Bt )
Table of Basic Fourier Transforms
f (t )
F (i )
(t )
1
1
B
sinc(Bt )
u(t a) u(t a)
(t )
i0t
e
2 ( )
u( B ) u( B ).
Homework
ei
2 ( 0 )
They can be easily obtained applying the Fourier definition,
or the inverse formula. Notice the time-frequency duality.
F (i )
f (t )
i0t
e
cos(0t )
i0t
e
e
2
i0t
2 ( 0 )
( 0 ) ( 0 )
( )
( )
0
sin(0t )
i0t
e
e
2i
i0t
0
i ( 0 ) ( 0 )
( i )
0
0
( i )