32 Chapter 2 Modeling and Simu lation

Applying the rotational version of Newton '8 second law to the motor shaft, we
obtain
J d20 (t) + J( dO(t) = J( i (t). (2. 13)
dt2 f dt t a

These two equations give the mathematical model of the DC motor system with
input va(t) and output ott).
In some applications, we arc concerned with the angular velocity (speed ) of
the motor, instead of the angular position. Such cases are called s peed control
cases. Replacing dOlt) by w(t) in (2.12) and (2.13) , we get the differential equation
dt
model of a DC motor system in the speed contro l case.

. dia (t)
Ra'a( t ) + La ~ + [(bW(t) = va(t) (2.14)

Jdw/(t ) + !(fw(t) = J(t ia(t). (2.15)

et
These two equations give the mathematical model of the DC motor system with
input va(t) and output w(t).
Another interesting electromechanical system is a magnetic-ball suspension
system shown in F igure 2.8. T he coil at the top, after being fed with current,
produces a magnetic field. The magnetic field generates an attracting force on the
steel ball .

p~/::r~r~~;

T (0

FIGURE 2.8: A magnetic-ball sus pension system.

Here, the voltage applied to t he coil v(t) is the input, the distance from the
ball to the coil y(t) is t he output, and the lift ing force generated by the magnetic
field on the ball is approxi mately given by I(t) = J( ~gi. The other parameters

1/ are mass of the ball M , winding resistance R, and winding inductance L.

I Applying KVL to the coil, we obtain

d(t)
Ri(t) + L~d = v(t).
t
 Section 2.2 State Spa ce Model and Lin ea rization 33

Applying Newton's second law to t he ball, we obtain

2 2
Md y(t) = _ J{i (t) M.
dt 2 y(t) + 9
These two equations then give the mathematical model of the system. It is noted
t hat t he variables y(t) and itt) are involved nonlinearly in t he equations a nd t his
system is called a nonlinear system.

2.2 STATE SPACE MODEL AND LINEARIZATION

The mathematical models obtained in the previous sections consist of sets of dif-
ferential equations with different orders and these equations in volve variables other
than the inpuls and outputs. For the sake of systematic studYI we need to put them
in standard forms . One of the commonly used standard forms is the state space
form.
D e finiti on 2.1. The state variables of a system are a set of iridepcndent
variables whose values at time tOI together with inpu t for all t 2: to, determine
the behavior of the system for all t 2: to.
This definition looks very abstract but in many situations the state variables
can be chosen intuitively. For electrical circuits we can always choose the voltages
l

across independent capacitors and the currents through independent inductors as

state variables. For mechanical systems, we can always choose the positions and
velocities of independent rigid bodies as state variables. Suppose that a difierential
eq \lation model of a system is already obtained 1 the variables in the differentia.l
equations are the input u(t) and the internal variables VI (t), ... , v,,(t), and t he
hi ghest order of the derivat ives of Vi(t) in the different ial equations is qi. Then we
can choose

Vi( t), Vi(t), Vi(t), ... , vi q

, - I) (t) i = 1,2 1", 1P
as the state var iables. In th is case, the total number of state variables is L~= l CJi·
After the state variables are chosen, usually named X, (t), X2(t), . . . , xn(t), we
put them into a vector

XI (t)
X2(t) 1 E {RH.
x(t) =
[
Xn(t)

T his vector is called a s tate vector. Here and in the sequel 1 we use bold font
letters to denote vectors (or matrices) and vector-valued (or matrix-valued) func-
tions, whereas we use normal font letters to denote scalars and scalar-valued
[ullctions. Then the set of mixed ordered differential equations can be cOl)verted
into a set of first-order differential equations plus an algebraic equation
x(t) = f[x(t) , u(t), t] (2.16)
y(t) = g[x(t), u(t), t] (2. 17)
 Sectio n 2.2 State Space Mod el and Linea rization 35

EXAMPLE 2.3
The magnetic suspension system has the differential equation model
d(i)
Ri(t) + L ~t = v(i)
2
Md y(t) = _ J(i2(t) + Mg.
cii 2 y(t;).
Choose XI(t) = i(t),x2(t) = y(i), x, (t) = y(t),u(t) = v(i), and we get the state
space model
u(t) - RXI (t)
L
x,(t)
f(xl(t)
=-'-7-'c:
MX2(t) +9
y(t) = X2(t)
with
XI(O)] [ i(O) ]
X2(0) = y(O) .
[ x,(O) y(O)

Definit ion 2.4. A system is said to be linear if it can be described by linear

differential equations, in particular, if the functions f and 9 in its state space
model are linear functions of x(t) ancl u(t).
For a linear system, the state space model takes the following matrix form:
x(t) = A(t)x(t) + b(t)u(t)
y(t) = c(i)x(t) + d(t)u(t)
where A(t) E ~n x n is an n x n matrix, possibly depending on time t, and b(t) E
lR nxl and c(t) E IR 1XIl are, respectively, column and row vectors depending possibly
on time t . For example , the RLC circuit in Section 2.1 is a linear system and its
state space equations were already in the matrix form as in (2.7) and (2.8)
Theorem 2.5 (Supe rposition Principle). Assume that a linear system
has zero initial condition. If input UI(t) produces output yI(t) ancl input
U2(t) produces output Y2(t), then input "'111.1 (t) + "'2U2(t) produces output
"'IYI (t ) + "'2Y2(t) for all "'I , "'2 E ~.
Proof. With zero initial condition, if input UI(t) produces output YI(t) and
input U2(t) produces output Y2(t), then there are XI(t) and X2-(t) with
Xl(a) = a ancl X2(0) = a satisfying
XI(t) = A(t)xI (i) + b(I,)uI(t)
YI (t) = C(t)XI (t) + d(t)uI (t)
38 Chapter 2 Modelin g and Simulation

where
oj, oj, oj,
ox, DX n OU
of of
ox
oJn &jn
OU
=
oIn
09
Ox [ ox,og og
Dxu 1
ox, OXn OU
Since i,(t), x(l), jj(t) arc small , we can neglect the high-order terms and approximate
the origi nal system by the [allowing linear system :
:tit) = Ax(t) + bu(t)
jj(t) = exit) + du(t)
where

e - -og d- Ogl
- ax 1
x = xo '
u=uo
- au X = Xu'
u = uo

This linear system is called a linearized system of the original nonli near system.

EXAMPLE 2.10
The magnetic suspension system is a nonlinear system described by state space
equation
u(t) - Rx, (t)
L
XI(t) ]
X2(t) = X3(t)
[ X3 (t)
xi(t)
MX2(t) +9
'y(t) = X2(t).
A usual control problem is to lift t he ball to a certain height and suspend it at that
height. J-fence we wish to linear ize it around an operating point with y(t} = Yo .
To get the operating point, solve equations
0 = UO - Rx10
L
0 = X30
X2
0 = __1_0_+ g
MX20
Yo = X2Q·
 Section 2.3 T ra nsfer Fun ction s and Impul se Responses 31

T his gives

Uo = R JMgyo, XlO
X20 ] -_ [ JlIil9Yo
Yo ], Yo = Yo
[
x'o 0
which means that to suspend t he ball at height Yo in t he steady state, one need
to apply a constant voltage u(t) = R.,j Mgyo to t he coi l. Denote t he deviations 0
the input , state, and output variables from the operating poin t by
u(t) = u(t) - Uo = u(l) - RJMgyo

Xl (t) - "fl'Vl!iYo ]
5:(t) = x(t) - Xo = X2(t) - Yo
[
X3(t)
fi(t) = y(t) - Yo·
Now, the linearized model of the deviation variables is
:i:(t) = A5:(t) + burl)
fi(t) = ci(t) + du(t)
where
R 1
0 0
L L
A -
- 8x
af l _
:C = :Co -
0 0 1 b _ af l
- 8u :C=Xo -
_ 0

-2) ~YO
U = Uo U=Uo
.9 0 0
Yo

e= a9 I
8x
- [0
:C = :Vo -
1 o J, d = agl
au :C=Xo
= o.
U = Uo U = Uo

2.3 TRANSFER FUNCTIONS AND IMPULSE RESPONSES

Consider an LT I system descri bed by state. space equation
5:(t) = Ax(t) + burt)
y(t) = eX(I) + du(t).
Take the Laplace transform with zero init ial conditions:
sX(s) = + bUrs)
AX(s) (2. 18)
Y(s) - eX(s) + dU(s) . (2 .19)
Now a set of differential equations in the time domain becomes a set of alge-
braic equations in the frequency domain. There are a total of n + 1 equations in
. (2.18) - (2 .19) and we call use them to elim inate the n variables in X(s) to obtain
an equation relating t he inp ut U(s) and the output Y (s) . Linear algebra now gives

...
 Section 2.3 Transfer Functions and Impulse Responses 41

For systems with a small number of state variables, it is probably more conve-
nient to obtain the transfer function by directly manipulat ing the Laplace trans[orm
of the state space model (2. 18) a nd (2.19). Por example, in the speed control case,
the Laplace transform of t he state space model is

sX I(s)
Ra
= - -L XI(s)
J~
- -[ X2(S)
1
+ -L U(s)
a Ja a

J(, J( f
sX 2(s) = J XI (S) - j X2(S)

Y (s) = X2(S),
Substitute Xl(S) from the second equation into the first equation and note that
Y (s) = X2 (S) from the third equation. We then get

J (
[ J(, S + La
Ra) (S + j J( f) + La1V (s)
J(o 1
= La U(s).

Consequently,

the same result as the one obtained by matrix in version.

EXAMPLE 2.13
Let us continue with Example 2.10, the magnetic suspension system . The transfer
function of the linearized model is
- I
R 1
S+ - 0 0
L L
G(s) = [0 1 OJ 0 s - 1 0
9
2 [Myo
;£ s 0
Yo

- 2[ ;Myo
£1 L - 2v?iYO

(s+ ~) (S2 - :J VM( Ls + R )(Yos2 - g)

The transfer function of an LTI system with <'I, state space model is always a
ratio of two polynomials
G(s) = b(s)
a(s)
where b(s) is called t he numera tor polynomial and a(s) is called t he denominator
polynomial. We assume that polynomials b(s) and a( s) are coprime , i.e., t hey do
not have common factors.