Image Processing Reference
In-Depth Information
3.2.1 S
OLUTION OF
C
ONSTANT
-C
OEFFICIENTS
L
INEAR
D
IFFERENTIAL
E
QUATIONS
We
first consider zero-input response that is the response of the system de
ned by
Equation 3.1 when the input signal u
(
t
)
is zero. Consider an Nth-order constant-
, y
0
(
coef
cients linear DE with zero input driven by initial conditions y
(
0
)
0
)
,...,and
y
(
N
1
)
(
0
)
:
N
y
(
t
)
d
t
N
N
1
d
þ
a
N
1
d
y
(
t
)
d
t
N
1
þþ
a
0
y
(
t
) ¼
0
(
3
:
2
)
Assuming a solution of the form y
(
t
) ¼
Ce
Dt
, we would have
e
Dt
CD
N
þ
a
N
1
D
N
1
þ
a
N
2
D
N
2
þþ
a
0
¼
0
(
3
:
3
)
Since Ce
Dt
is nonzero, then
D
N
þ
a
N
1
D
N
1
þ
a
N
2
D
N
2
þþ
a
0
¼
0
(
3
:
4
)
This polynomial is called characteristic equation of the DE. It is a polynomial of
degree N that has N roots D
1
, D
2
,..., andD
N
. These roots are generally complex
and are called characteristic roots of the DE. The zero-input solution is given by
y
(
t
) ¼
C
1
e
D
1
t
þ
C
2
e
D
2
t
þþ
C
N
e
D
N
t
(
3
:
5
)
The constants C
1
, C
2
,..., andC
N
are found by applying the N initial conditions.
The stability of the system is de
ned in terms of the zero-input response. The
discrete LTI system described by DE given in Equation 3.1 is stable if the zero-input
response decays to zero as t
!1
. This implies that the characteristic roots must
have real parts less than zero, that is,
Re(
D
i
) <
for
i
¼
...
(
:
)
0
1, 2,
, N
3
6
Example 3.1
Solve the following second-order DE with initial conditions y(0)
¼
1 and
dy(t)
dt
j
t¼0
¼
4:
d
2
y(t)
dt
2
3
dy(t)
þ
dt
þ
2y(t)
¼
0
S
OLUTION
The characteristic equation is D
2
þ
3D þ
2
¼
0, which can be written as
D
2
þ
3D þ
2
¼
(D þ
1)(D þ
2)
¼
0
Therefore, the roots are D
1
¼
1 and D
2
¼
2. The zero-input response is
¼ C
1
e
D
1
t
þ C
2
e
D
2
t
¼ C
1
e
t
þ C
2
e
2t
y(t)
Search WWH ::
Custom Search