2 Coprime Factorization and 2.

Coprime Factorization and

( ) ( )
{ }
p g p p

Definition 2.1 Definition 2.1

( )
( ) ( ) ( )
( )
1
( ) det 0 (i) is square and
(ii)
M s M s
G s M s N s

=
=



(2.1)
( ) ( )
( )
( ) ( ) ( ) ( )
, (iii) s.t. V s U s H H M s V s N s U s I

- e + =

( ) ( ) ( ) ( )
, is a right coprime factorization of M s N s H H G s

e
(2.2)

( )
( ) ( ) ( )
1
( ) det 0
if and only if
(i) is square and
(ii)
M s M s
G s N s M s

=
= (2 3)
Bezout identity

2
( ) ( ) ( )
( ) ( ) ( ) ( ) ( ) ( ) ( )
,
(ii)
(iii) s.t.
G s N s M s
V s U s H H U s N s V s M s I

=
- e + =
(2.3)
(2.4)
Example
1
(1)
( )
1
1
G s
s
=
+
: left coprime factorization
( ) ( ) ( ) ( )
1
1,
1
M s M s N s N s = = = =

(2)
: right coprime factorization
( )
1
1
G s
s
=

l ft i f t i ti
( ) ( ) ( ) ( )
,
1 s+
: left coprime factorization
: right coprime factorization
( ) ( ) ( ) ( )
(i) and (ii) hold trivially. For (iii), 1 and 2. V s V s U s U s = = = =

( ) ( ) ( ) ( )
1 1
,
1 1
s
M s M s N s N s
s s

= = = =
+ +

(3)
( )
1 1
1 1
G s
s s
(
=
(
+

1 1 1 s s
(
( ) ( )
( )
( ) ( )
2
1 1 1
,
1 1
1
1 0
1 1
s s
M s N s
s s
s
(

= = (
+ +
+
(

(
(
(

: left coprime factorization
i ht i f t i ti
( ) ( )
1 1
,
1
1 1 0
1
M s N s
s
s s
s
(
(
= =

(
(
+ +

+
: right coprime factorization
1 0
0 0
3
(
( (
+
(
( (
3
( ) ( ) ( )
( )
( )
2
3
4 3 For (iii), , , ,
4 4
1
1
1
1 1
s
s V s U s V s U s
s
s
s
s s
+
(
( (
+ = = = =
(

( (
+
+ (
+
+ +


Remark 2.2
The left and right coprime factorizations of G(s) can be grouped together The left and right coprime factorizations of G(s) can be grouped together
to form a Bezout double identity as
( ) ( ) ( ) ( )
0 V s U s I M s U s
( (
(

( ) ( )
( ) ( )
( ) ( )
( ) ( )
0
0
V s U s I M s U s
N s M s I N s V s
( (
(
=
( (
(

(2.5)
Suppose that satisfy (2.1),(2.2),(2.3) and (2.4),
i.e.,
, , , , , , , M N V U M N V U

VM UN I NU MV I

1 1
,
0
VM UN I NU MV I
M N NM NM MN

+ = + =
= + =

define ( ) and ( ) as VU UV Q V s U s
' '
+ =

and
define ( ) and ( ) as
,
VU UV Q V s U s
V V QN U U QM
+ =
' '
+


Then, (2.5) holds as follows
( ) ( ) ( ) ( )
( ) ( )
0
0
V QN M U QM N V QN U U QM V V U
M U
N M
N V
I
VM UN Q NM MN VU UV Q NU MV
(
' ' + + + + (
(
=
(
(
(

(

(
(

4
( ) ( )
0

0
0
VM UN Q NM MN VU UV Q NU MV I
I
I
(
+ + + +
(
= =
(
(
(

Theorem 2.3
( ) ( ) ( )
, , where :stabilizable, :detectable
A B
G s A B A C
C D
(
=
(

Let constant matrices F and H be such that A+BF and A+HC are both stable. Let constant matrices F and H be such that A BF and A HC are both stable.
Then the transfer function matrices defined in the following
hold the Bezout double identity.
, , , , , , , M N V U M N V U

( )
(
( ) ( )
( ) ( )
( )
0
A HC B HD H
V s U s
F I
N s M s
C D I
+ + (
(
(
=
(
(

(


(2.6)
C D I
(

( ) ( )
0
A BF B H
M s U s
F I
+
(
(

(
(

(2 7)
( ) ( )
( ) ( )
0 F I
N s V s
C DF D I
(
=
(
(

( +

(2.7)
Furthermore the above hold
M N M N

Furthermore, the above hold
( ) ( ) ( ) ( ) ( )
1 1
G s M s N s N s M s

= =

(2.8)
, , , M N M N
5
Exercise 2.1 Prove the equation (2.8).
Hint: Use the formula (1.6).
<Proof of Theorem 2.3>
Because A+BF and A+HC are stable it holds that M N V U M N V U H e

Because A+BF and A+HC are stable, it holds that . , , , , , , , M N V U M N V U H

e
( ) ( )
1
0 V U F I
sI A HC B HD H
N M C D I
( ( (
= + + (
( ( (



( ) | |
1
0 F I
M U
sI A BF B H
C DF D I
N V

(
( (
= +
(
( (
+

(
( ) ( ) ( ) | |
( ) ( ) ( ) | |
1 1
1 1
0 0
V U F F
M U
sI A HC B HD H sI A BF B H
N M C C DF
N V
F I I F
I A HC B HD H I A BF B H I


( ( ( (
= + (
(
( ( (
+


( ( ( (
(
( ( ( (

( ) ( ) ( ) | |
( ) ( )( ) | |
1 1
sI A HC B HD H sI A BF B H I
C D I D I C DF
F
sI A HC BF HC sI A BF B H
C

+ + + + (
( ( ( (
+

(
= +
(

C
F
sI
C

(
+
(

( ) | | ( ) | |
1 1
F
A HC B H sI A BF B H I
C
(
+ +
(

(
( ) ( ) ( ) ( ) { }( ) | |
1 1
F
sI A HC BF HC sI A BF sI A HC sI A BF B H I
C
I
(
= + + +
(

=
6
To complete the proof, you have to prove the equation (2.8), and this is left as
the exercise 2.1.
(q.e.d.)
2.2 Stabilizing Controllers
d
( ) ( ) ( )
1
G s M s N s

=

g
( )
K s
( )
G s
r
e u y
+
-
+
+
( ) ( ) ( )
( ) ( )
1
N s M s

=
: l.c.f and r.c.f.
-
( ) ( ) ( ) ( )
, , , V s U s V s U s

the Bezout double identity (2.5).
Lemma 2 4
satisfy
Lemma 2.4
( ) ( ) ( ) ( ) ( )
1 1
K s U s V s V s U s

= =

is a stabilizing controller in the interconnected system of G(s) and K(s).
(2.9)
( )
1 1
1
1
1
0

0
I G
I M N M N M
M s
K I
V U I U V V

( ( (
(
= = =
( ( (
(



)
( )
1
1
1 1 1
0
0 0 0 0
0 0
0 0 0 0
0 0
K I
V U I U V V
I V U I I V U I
M M
I N M I I N M I
V V

( (
( ( ( ( ( (
= =
` ( (
( ( ( ( ( (


)


7
1
0
0
I
M U
I
N V

)
(
(
=

(

1
0
0
0
0
I
M
H
I
V

(
(
e
( (
(

r
e
d
+
( )
K s
( )
G s
r
e u y
+
-
+
+
Theorem 2.5
The set of all stabilizing controller K(s) is given by The set of all stabilizing controller K(s) is given by
( ) ( ) ( ) ( )
( )
( ) ( ) ( )
( )
( )
{ }
{ }
1
K s U s M s Q s V s N s Q s Q s H

= = + e K

(2.10)
( ) ( ) ( ) ( )
( )
( ) ( ) ( )
( )
( )
{ }
1
K s V s Q s N s U s Q s M s Q s H

= = + e

( 0)
Exercise 2.2
Prove that every controller stabilizes G(s)
( )
K s eK y ( )
in the interconnected system.
( )
8
Remark 2.6
The controller K(s) in (2.10) can be represented as follows.
e u
( )
1
V s

( )
U s

( )

+
-
+
-
e u
( )
Q s

( )
N s
( )
M s
( )
N s
( )
M s
e u
( )
U s
( )
1
V s

+
-
e u
( )
Q s
( )
M s

( )
N s

9
( )
M s
( )
N s
+ +
Remark 2.7
When we choose in (2.10),
the stabilizing controller K(s) is given by
( ) ( ) ( )
0 or, 0 Q s Q s = =

( ) ( ) ( ) ( ) ( )
1 1
K s U s V s V s U s


( ) ( ) ( ) ( ) ( )
K s U s V s V s U s = =
and this controller is just a state feedback + observer .
In fact, from Theorem 2.3,
( ) ( )
and .
0
A BF H A BF H
U s V s
F C DF I
+ +
( (
= =
( (


( ) ( )
0 F C DF I
( (
+

Therefore
( )
1
A BF HC HDF H
V s

+ + +
(
=
(

( )
( )
V s
C DF I
=
(
+

and
( ) ( ) ( ) ( )
1 1
. K s U s V s F sI A BF HC HDF H

= =

( ) ( ) ( ) ( ) ( ) ( )
, t A BF HC HDF t He t u t F t = + + + + =

This state space model is described as