MEEN 651:

Control System Design

Lecture Notes 2
Laplace and z Transforms

Prabhakar R. Pagilla
[email protected]

Mechanical Engineering
Texas A&M University
College Station, TX 77843-3123
Phone: (979) 458-4829

1
 Laplace and z transforms

• Laplace and z transforms convert calculus operations

to algebraic operations.
• Powerful tools for linear system analysis and design.
1 • The Laplace transform can be used to transform
linear time-invariant (LTI) differential equations into
algebraic equations in the complex variable s.
• The z transform can be used to transform linear
time-invariant (LTI) difference equations into
algebraic equations in z.
• Laplace and z transforms are linear transforms.

Review of Laplace transformation

Definition: : Given a time function f (t), its Laplace

transform is defined as
2 Z ∞
F (s) := L{f (t)} := f (t)e−st dt
0
−at
Example: f (t) = e , F (s) = 1/(s + a).
If a = 0, then f (t) is a unit step function, and
F (s) = 1/s.

MEEN 651, Notes 2, Pagilla, Texas A&M University 2

 Differentiation (d/dt) in the time-domain corresponds to
multiplication by s in Laplace domain:
  Z ∞
df df −st
L = e dt
dt dt
0
Z ∞
−st ∞
= f (t)e 0
− (−s) f (t)e−st dt
0
3 = sF (s) − f (0)

First order system example:

dy(t)
+ ay(t) = bu(t); y(0) = 0
dt
the Applying Laplace transform
b
sY (s) − y(0) + aY (s) = bU (s) =⇒ Y (s) = U (s)
s+a

• If u(t) is a unit step function, Y (s) = b/s(s + a).

Taking partial fraction expansion and applying
inverse transformation we get y(t):
 
b 1 1 b
Y (s) = − =⇒ y(t) = (1−e−at )
a s s+a a
• Find y(t) if u(t) is a unit ramp function.
4
• Find y(t) if u(t) is a sinusoidal function.
• For the above input functions, find y(t) for the
following second-order system:
ÿ(t) + 2ζωn ÿ(t) + ωn2 y(t) = u(t)
where ζ and ωn are real numbers. Consider the
following cases: (1) 0 < ζ < 1; (2) ζ = 1; (3) ζ > 1.

MEEN 651, Notes 2, Pagilla, Texas A&M University 3

 Generally, n-th order transfer function is
Y (s) b 0 sm + · · · + b m B(s)
:= G(s) = :=
U (s) sn + a1 sn−1 + · · · + an A(s)

Characteristic equation: A(s) = 0.

5 Poles of G(s): Roots of A(s) = 0.
Zeros of G(s): Roots of B(s) = 0.
Realizability condition: m ≤ n.
Given the input function, u(t), to get y(t): factorize
A(s), find partial fraction expansion of G(s), apply
inverse transformation (Laplace transform tables).

Useful Theorems and Properties

• Initial Value Theorem:

limt→0 f (t) = lims→∞ sF (s)
• Final Value Theorem:
6
limt→∞ f (t) = lims→ 0 sF (s)
• Linearity:

L[αf1 (t) + βf2 (t)] = aF1 (s) + bF2 (s)

• Time delay:

L[f (t − td )] = e−td s F (s)

MEEN 651, Notes 2, Pagilla, Texas A&M University 4

 • Differentiation:
[ ]
dn f (t)
L dtn
= sn F (s) − sn−1 f (0) − sn−2 f 1 (0) − · · · − f n−1 (0)

• Integration:
Z t 
L f (τ )dτ = 1s F (s)
0
7
• Multiplication by an exponential function:
 
L e−αt f (t) = F (s + α)

• Convolution integral:
R∞
f1 (t) ∗ f2 (t) := 0 f1 (t − τ )f2 (τ )dτ ,

L [f1 (t) ∗ f2 (t)] = F1 (s)F2 (s)

The z transform

Definition: The z transform of a sampled signal x(kT ),

or a time sequence x(k), k = 0, 1, . . . is defined by
X
∞
8 X(z) = Z[x(kT )] = x(kT )z −k
k=0
or
X
∞
X(z) = Z[x(k)] = x(k)z −k
k=0
where z is a complex variable which is chosen such that
the above infinite sum converges.

MEEN 651, Notes 2, Pagilla, Texas A&M University 5

 Examples:
• x(k) = 1(k), k = 0, 1, . . . (unit-step signal), then
X
∞
1
X(z) = z −k = 1 + z −1 + z −2 + . . . =
k=0
1 − z −1
9
• x(k) = ak , k = 0, 1, . . . (polynomial function), then
X
∞
X(z) = ak z −k = 1 + az −1 + a2 z −2 + . . .
k=0
1 z
= =
1 − az −1 z−a

• x(kT ) = sin(kωT ), k = 0, 1, . . ., then

X
∞ X
∞
ejkωT − e−jkωT
X(z) = sin(kωT )z −k = z −k
k=0 k=0
2j
z −1 sin(kωT )
=
1 − 2z −1 cos(kωT ) + z −2
10
z sin(kωT )
=
z 2 − 2z cos(kωT ) + 1
Euler’s formula:

ejωt = cos ωt + j sin ωt

e−jωt = cos ωt − j sin ωt

MEEN 651, Notes 2, Pagilla, Texas A&M University 6

 Useful properties and theorems of z transform

• Z[αf (k) + βg(k)] = αF (z) + βG(z)

• Z[ak x(k)] = X(a−1 z)
11 • Shifting Theorem: Z[x(t − nT )] = z −n X(z);
X
n−1
Z[x(t + nT )] = z [X(z) −
n
x(kT )z −k ]
k=0

• Initial Value Theorem:

x(0) = lim X(z)

z→∞

• Final Value Theorem:

lim x(k) = lim (z − 1)X(z)
k→∞ z→1

• Convolution: Let X1 (z) = Z[x1 (k)] and

X2 (z) = Z[x2 (k)], then
"∞ #" ∞ #
X X
12 X1 (z)X2 (z) = x1 (i)z −i x2 (j)z −j
i=0 j=0
(" # )
X
∞ X
k
= x1 (k − i)x2 (i) z −k
k=0 i=0
= Z[x1 (k) ∗ x2 (k)]
where ∗ denotes the convolution sum in the
discrete-time domain.

MEEN 651, Notes 2, Pagilla, Texas A&M University 7

 Example

Consider the discrete sequence given by

x(k) = 1 − ak , |a| < 1, k = 0, 1, 2, · · ·

13 The z transform of this sequence is

X(z) = Z[1] − Z[ak ]

z z (1 − a)z
= − =
z−1 z−a (z − 1)(z − a)
(1−a)z
IVT: x(0) = limz→∞ X(z) = limz→∞ (z−1)(z−a)
= 0.

FVT:

lim x(k) = lim [(z − 1)X(z)]

k→∞ z→1
(1 − a)z
= lim (z − 1) = 1.
z→1 (z − 1)(z − a)
14 We can use the shift theorem and IVT to find the value
of x(k) for any k. Now x(1) can be found as follows:

x(1) = lim Z[f (k + 1)] = lim [zX(z) − zx(0)]

z→∞ z→∞
z (1 − a)
2
= lim =1−a
z→∞ (z − 1)(z − a)

MEEN 651, Notes 2, Pagilla, Texas A&M University 8

 The Discrete-time Transfer Function

Consider the discrete-time system given below.

y(k + 2) + a1 y(k + 1) + a2 y(k) = b2 u(k)

Taking z transform on both sides and using its properties,
15 z 2 Y (z) − z 2 y(0) − zy(1) + a1 [zY (z) − zy(0)] + a2 Y (z) = b2 U (z)

where Y (z) = Z[y(k)] and U (z) = Z[u(k)]. For

y(0) = 0, y(1) = 0, we obtain
Y (z) b2
G(z) := =
U (z) z 2 + a1 z + a2
G(z) is called the discrete-time transfer function.

For an n-th order system,

b0 z m + b1 z m−1 + · · · + bm
G(z) = (m ≤ n)
z n + a1 z n−1 + · · · + an
b0 (z − z1 )(z − z2 ) · · · (z − zm )
=
(z − p1 )(z − p2 ) · · · (z − pn
16 where pi ’s (i = 1, 2, . . . n) are the poles of the system
(G(z)) and zj ’s (j = 1, 2, . . . m) are the zeros of the
system. G(z) can also be written as
b0 z −(n−m) + b1 z −(n−m+1) + · · · + bm z −n
G(z) =
1 + a1 z −1 + a2 z −2 + · · · + an z −n
In the above, z is usually interpreted as an advance shift
operator and z −1 is interpreted as a unit delay operator.

MEEN 651, Notes 2, Pagilla, Texas A&M University 9

 The inverse z transform

Given X(z), the z transform of x(k), we use the inverse

z transform to find x(k). Four methods for obtaining the
inverse z transform are:
• Direct division method: Expand X(z) into an
17 infinite power series in z −1 and compare coefficients
P
using the definition of X(z) (:= ∞ k=0 x(k)z
−k
) to
obtain values of x(k) for different k.
• Computational method: Use software such as
MATLAB, FORTRAN.
• Partial-fraction-expansion method: Find the
partial-fraction-expansion, and use the z transform

tables.
Partial-fraction-expansion of a general X(z):
B(z) b0 z m + b1 z m−1 + · · · + bm
X(z) = =
A(z) z n + a1 z n−1 + · · · + an
where m < n. Then the expansion takes the form,
18 C11 C12 C1r1
X(z) = + + ··· +
z − p1 (z − p1 )2 (z − p1 )r1
Cqrq
+ ··· +
(z − pq )rq
X
q
X
ri
Cij
:=
i=1 j=1
(z − pi )j

where q represents the number of roots, some of

MEEN 651, Notes 2, Pagilla, Texas A&M University 10

 which may be multiple roots, and ri is the
multiplicity of the ith root. Note that if all roots are
distinct then q = n.
19 The constants Cij are given by

 
1 dri −j
Cij = lim [(z − pi ) X(z)]
ri
(ri − j)! z→pi dz ri −j

Examples

Partial-Fraction-Expansion:
1. Consider the X(z) given below
1 C11 C21
X(z) = := +
(z − 1)(z − 0.5) z−1 z − 0.5
20 The coefficients C11 and C21 can be computed as
 
1
C11 = lim [(z − 1)X(z)] = lim (z − 1)
z→1 z→1 (z − 1)(z − 0.5)
=2
C21 = lim [(z − 0.5)X(z)]
z→0.5
= −2

To find x(k), we use the z transform tables. First, write X(z)

MEEN 651, Notes 2, Pagilla, Texas A&M University 11

 in terms of the unit delay operator,
2 2 2z −1 2z −1
X(z) = − = −
z−1 z − 0.5 1− z −1 1 − 0.5z −1
From the z transform tables,
 
−1 1
y1 (k) := Z = 1(k)
1 − z −1
 
21 1
y2 (k) := Z −1 = (0.5)k
1 − 0.5z −1
Therefore,

x(k) = 2[y1 (k − 1) − y2 (k − 1)], k = 0, 1, 2, 3, . . .

Since y1 (k) and y2 (k) is assumed to be zero for all k < 0, we

get
x(k) = 2[1 − (0.5)k−1 ], k = 1, 2, 3, . . .

2. Consider the X(z) given below,

1 C11 C12 C21
X(z) = := + +
(z − 1)2 (z − 0.5) z−1 (z − 1)2 z − 0.5
We know C11 , C12 and C21 can be computed as
 
  1
C12 = lim (z − 1)2 X(z) = lim (z − 1)2
z→1 z→1 (z − 1)2 (z − 0.5)
22
=2
    
d   d 1
C11 = lim (z − 1) X(z) = lim
2
z→1 dz z→1 dz z − 0.5
= −4
 
1
C21 = lim [(z − 0.5)X(z)] = lim
z→0.5 z→0.5 (z − 1)2

=4

MEEN 651, Notes 2, Pagilla, Texas A&M University 12

 To find x(k), write X(z) as

C11 z −1 C12 z −2 C21 z −1

X(z) = + +
1 − z −1 (1 − z −1 )2 1 − 0.5z −1
From the z transform tables,
" #
23 z −1
y(k) = Z −1 = kak−1
(1 − az −1 )2

Therefore,

 2 − 4(k − 1) + 4(0.5)k−1 , k = 1, 2, 3, . . .
x(k) =
 0 otherwise

• Inversion integral method:

I
−1
1
x(k) = Z [X(z)] = X(z)z k−1 dz
2πj C
= K 1 + K2 + · · · + K q

where K1 , K2 , . . . , Kq denote the residues of

24 X(z)z k−1 at poles p1 , p2 , . . . , pq , respectively. The
residues are computed depending on pi being simple
or multiple,
– If z = pi is a simple pole of X(z)z k−1 , then
corresponding residue K is

K = lim [(z − pi )X(z)z k−1 ]

z→pi

MEEN 651, Notes 2, Pagilla, Texas A&M University 13

 – If z = pi is a multiple pole of X(z)z k−1 of order
r, then corresponding residue K is
1 dr−1
K= lim [(z − pi )r X(z)z k−1 ]
(r − 1)! z→pi dz r−1
25 • Note: The inversion integral method is a simple
technique to find the inverse z transform if X(z)z k−1
has no poles at the origin. In most cases you can use
the partial fraction expansion method but it may be
simpler in some cases to use the inversion integral
method.

Solving difference equations using the z transform

Example: Consider the following difference equation

x(k + 2) + 5x(k + 1) + 6x(k) = u(k), x(0) = 0, x(1) = 0

Let the input u(k) be

26 
 0, k<0
u(k) =
 1, k = 0, 1, 2, . . .

Take the z transform on both sides using the properties of z

transform
z
z 2 X(z) − z 2 x(0) − zx(1) + 5zX(z) − 5zx(0) + 6X(z) =
z−1

MEEN 651, Notes 2, Pagilla, Texas A&M University 14

 Substituting initial data and simplifying gives
z z
X(z) = =
+ 5z + 6)(z − 1)
(z 2 (z + 3)(z + 2)(z − 1)
C11 C21 C31
= + +
z−1 z+2 z+3
z −1 z −1 z −1
= C11 + C 21 + C 31
1 − z −1 1 + 2z −1 1 + 3z −1
27
Using the z transform tables, we get

x(k) = C11 1(k−1)+C21 (−2)k−1 1(k−1)+C31 (−3)k−1 1(k−1),

or

 0, k≤0
x(k) =
 C11 + C21 (−2)k−1 + C31 (−3)k−1 , k = 1, 2, . . .

MEEN 651, Notes 2, Pagilla, Texas A&M University 15