Robust Adaptive Control by Petros A. Ioannou, Jing Sun - HTML preview

PLEASE NOTE: This is an HTML preview only and some elements such as links or page numbers may be incorrect.
Download the book in PDF, ePub, Kindle for a complete version.

G( s) = msm + bm− 1 sm− 1 + · · · + b 0 =

sn + an− 1 sn− 1 + · · · + a 0

U ( s)

where n > m. Then the system may be represented in the controller form

−an− 1 −an− 2 · · · −a 1 −a 0

1

1

0

· · ·

0

0

 0 

˙ x = 

0

1

· · ·

0

0

 . 

x +  ..  u

(2.3.6)

.

.

.

.

..

..

. .

.. 

 0 

0

0

· · ·

1

0

0

y = [0 , 0 , . . . , bm, . . . , b 1 , b 0] x

or in the observer form

−a

0

n− 1

1 0 · · · 0

 . 

−an− 2 0 1 · · · 0 

 . 

 . 

˙ x = 

.

.

.

. 

..

..

. . ..  x +  bm u

(2.3.7)

−a

 . 

1

0 0 · · · 1 

 .. 

−a 0

0 0 · · · 0

b 0

y = [1 , 0 , . . . , 0] x

2.3. INPUT/OUTPUT MODELS

37

One can go on and generate many different state-space representations

describing the I/O properties of the same system. The canonical forms in

(2.3.6) and (2.3.7), however, have some important properties that we will use

in later chapters. For example, if we denote by ( Ac, Bc, Cc) and ( Ao, Bo, Co)

the corresponding matrices in the controller form (2.3.6) and observer form

(2.3.7), respectively, we establish the relations

[adj( sI − Ac)] Bc = [ sn− 1 , . . . , s, 1] = αn− 1( s)

(2.3.8)

Co adj( sI − Ao) = [ sn− 1 , . . . , s, 1] = αn− 1( s)

(2.3.9)

whose right-hand sides are independent of the coefficients of G( s). Another

important property is that in the triples ( Ac, Bc, Cc) and ( Ao, Bo, Co), the

n+ m+1 coefficients of G( s) appear explicitly, i.e., ( Ac, Bc, Cc) (respectively

( Ao, Bo, Co)) is completely characterized by n + m + 1 parameters, which are

equal to the corresponding coefficients of G( s).

If G( s) has no zero-pole cancellations then both (2.3.6) and (2.3.7) are

minimal state-space representations of the same system. If G( s) has zero-

pole cancellations, then (2.3.6) is unobservable, and (2.3.7) is uncontrollable.

If the zero-pole cancellations of G( s) occur in Re[ s] < 0, i.e., stable poles are

cancelled by stable zeros, then (2.3.6) is detectable, and (2.3.7) is stabilizable.

Similarly, a system described by a state-space representation is unobservable

or uncontrollable, if and only if the transfer function of the system has zero-

pole cancellations. If the unobservable or uncontrollable parts of the system

are asymptotically stable, then the zero-pole cancellations occur in Re[ s] < 0.

An alternative approach for representing the differential equation (2.3.1)

is by using the differential operator

d( ·)

p( ·) = dt

which has the following properties:

( i) p( x) = ˙ x;

( ii) p( xy) = ˙ xy + x ˙ y

where x and y are any differentiable functions of time and ˙ x = dx( t) .

dt

The inverse of the operator p denoted by p− 1 or simply by 1 is defined

p

as

1

t

( x) =

x( τ ) + x(0) ∀t ≥ 0

p

0

38

CHAPTER 2. MODELS FOR DYNAMIC SYSTEMS

where x( t) is an integrable function of time. The operators p, 1 are related

p

to the Laplace operator s by the following equations

L {p( x) }|

= sX( s)

x(0)=0

1

1

L{ ( x) } |

X( s)

p

x(0)=0= s

where L is the Laplace transform and x( t) is any differentiable function of

time. Using the definition of the differential operator, (2.3.1) may be written

in the compact form

R( p)( y) = Z( p)( u)

(2.3.10)

where

R( p) = pn + an− 1 pn− 1 + · · · + a 0

Z( p) = bmpm + bm− 1 pm− 1 + · · · + b 0

are referred to as the polynomial differential operators [226].

Equation (2.3.10) has the same form as

R( s) Y ( s) = Z( s) U ( s)

(2.3.11)

obtained by taking the Laplace transform on both sides of (2.3.1) and as-

suming zero initial conditions. Therefore, for zero initial conditions one can

go from representation (2.3.10) to (2.3.11) and vice versa by simply replacing

s with p or p with s appropriately. For example, the system

s + b

Y ( s) =

0 U( s)

s 2 + a 0

may be written as

( p 2 + a 0)( y) = ( p + b 0)( u)

with y(0) = ˙ y(0) = 0 , u(0) = 0 or by abusing notation (because we never

defined the operator ( p 2 + a 0) 1) as

p + b

y( t) =

0 u( t)

p 2 + a 0

Because of the similarities of the forms of (2.3.11) and (2.3.10), we will use

s to denote both the differential operator and Laplace variable and express

the system (2.3.1) with zero initial conditions as

Z( s)

y =

u

(2.3.12)

R( s)

2.3. INPUT/OUTPUT MODELS

39

where y and u denote Y ( s) and U ( s), respectively, when s is taken to be the Laplace operator, and y and u denote y( t) and u( t), respectively, when s is taken to be the differential operator.

We will often refer to G( s) = Z( s) in (2.3.12) as the filter with input u( t)

R( s)

and output y( t).

Example 2.3.1 Consider the system of equations describing the motion of the cart

with the two pendulums given in Example 2.2.1, where y = θ 1 is the only measured

output. Eliminating the variables θ 1 , θ 2, and ˙ θ 2 by substitution, we obtain the

fourth order differential equation

y(4) 1 . 1( α 1 + α 2) y(2) + 1 . 2 α 1 α 2 y = β 1 u(2) − α 1 β 2 u where αi, βi, i = 1 , 2 are as defined in Example 2.2.1, which relates the input u with

the measured output y.

Taking the Laplace transform on each side of the equation and assuming zero

initial conditions, we obtain

[ s 4 1 . 1( α 1 + α 2) s 2 + 1 . 2 α 1 α 2] Y ( s) = ( β 1 s 2 − α 1 β 2) U( s) Therefore, the transfer function of the system from u to y is given by

Y ( s)

β

=

1 s 2 − α 1 β 2

= G( s)

U ( s)

s 4 1 . 1( α 1 + α 2) s 2 + 1 . 2 α 1 α 2

For l 1 = l 2, we have α 1 = α 2 , β 1 = β 2, and

β

β

G( s) =

1( s 2 − α 1)

=

1( s 2 − α 1)

s 4 2 . 2 α 1 s 2 + 1 . 2 α 21

( s 2 − α 1)( s 2 1 . 2 α 1)

has two zero-pole cancellations. Because α 1 > 0, one of the zero-pole cancellations

occurs in Re[ s] > 0 which indicates that any fourth-order state representation of

the system with the above transfer function is not stabilizable.

2.3.2

Coprime Polynomials

The I/O properties of most of the systems studied in this book are repre-

sented by proper transfer functions expressed as the ratio of two polynomials

in s with real coefficients, i.e.,

Z( s)

G( s) =

(2.3.13)

R( s)

40

CHAPTER 2. MODELS FOR DYNAMIC SYSTEMS

where Z( s) = bmsm + bm− 1 sm− 1 + · · · + b 0 , R( s) = sn + an− 1 sn− 1 + · · · + a 0

and n ≥ m.

The properties of the system associated with G( s) depend very much

on the properties of Z( s) and R( s). In this section, we review some of

the general properties of polynomials that are used for analysis and control

design in subsequent chapters.

Definition 2.3.1 Consider the polynomial X( s) = αnsn + αn− 1 sn− 1 + · · · +

α 0 . We say that X( s) is monic if αn = 1 and X( s) is Hurwitz if all the roots of X( s) = 0 are located in Re[ s] < 0 . We say that the degree of X( s) is n if the coefficient αn of sn satisfies αn = 0 .

Definition 2.3.2 A system with a transfer function given by (2.3.13) is

referred to as minimum phase if Z( s) is Hurwitz; it is referred to as stable

if R( s) is Hurwitz.

As we mentioned in Section 2.3.1, a system representation is minimal

if the corresponding transfer function has no zero-pole cancellations, i.e., if

the numerator and denominator polynomials of the transfer function have

no common factors other than a constant. The following definition is widely

used in control theory to characterize polynomials with no common factors.

Definition 2.3.3 Two polynomials a( s) and b( s) are said to be coprime (or

relatively prime ) if they have no common factors other than a constant.

An important characterization of coprimeness of two polynomials is given

by the following Lemma.

Lemma 2.3.1 (Bezout Identity) Two polynomials a( s) and b( s) are co-

prime if and only if there exist polynomials c( s) and d( s) such that

c( s) a( s) + d( s) b( s) = 1

For a proof of Lemma 2.3.1, see [73, 237].

The Bezout identity may have infinite number of solutions c( s) and d( s)

for a given pair of coprime polynomials a( s) and b( s) as illustrated by the

following example.

2.3. INPUT/OUTPUT MODELS

41

Example 2.3.2 Consider the coprime polynomials a( s) = s + 1 , b( s) = s + 2. Then the Bezout identity is satisfied for

c( s) = sn + 2 sn− 1 1 , d( s) = −sn − sn− 1 + 1

and any n ≥ 1.

Coprimeness is an important property that is often exploited in control

theory for the design of control schemes for LTI systems. An important the-

orem that is very often used for control design and analysis is the following.

Theorem 2.3.1 If a( s) and b( s) are coprime and of degree na and nb, re-

spectively, where na > nb, then for any given arbitrary polynomial a∗( s) of

degree na∗ ≥ na, the polynomial equation

a( s) l( s) + b( s) p( s) = a∗( s)

(2.3.14)

has a unique solution l( s) and p( s) whose degrees nl and np, respectively,

satisfy the constraints np < na, nl ≤ max( na∗ − na, nb − 1) .

Proof From Lemma 2.3.1, there exist polynomials c( s) and d( s) such that

a( s) c( s) + b( s) d( s) = 1

(2.3.15)

Multiplying Equation (2.3.15) on both sides by the polynomial a∗( s), we obtain

a∗( s) a( s) c( s) + a∗( s) b( s) d( s) = a∗( s) (2.3.16)

Let us divide a∗( s) d( s) by a( s), i.e.,

a∗( s) d( s)

p( s)

= r( s) +

a( s)

a( s)

where r( s) is the quotient of degree na∗ + nd − na; na∗, na, and nd are the degrees of a∗( s) , a( s), and d( s), respectively, and p( s) is the remainder of degree np < na.

We now use

a∗( s) d( s) = r( s) a( s) + p( s)

to express the right-hand side of (2.3.16) as

a∗( s) a( s) c( s) + r( s) a( s) b( s) + p( s) b( s) = [ a∗( s) c( s) + r( s) b( s)] a( s) + p( s) b( s) 42

CHAPTER 2. MODELS FOR DYNAMIC SYSTEMS

and rewrite (2.3.16) as

l( s) a( s) + p( s) b( s) = a∗( s)

(2.3.17)

where l( s) = a∗( s) c( s) + r( s) b( s). The above equation implies that the degree of l( s) a( s) = degree of ( a∗( s) − p( s) b( s)) max {na∗, np + nb}. Hence, the degree of l( s), denoted by nl, satisfies nl ≤ max {na∗ − na, np + nb − na}. We, therefore, established that polynomials l( s) and p( s) of degree nl ≤ max {na∗ − na, np +

nb − na} and np < na respectively exist that satisfy (2.3.17). Because np < na