Control Engineering

Lecture-2

Muhammad Sajjad Sabir

1
 Section A

2
This question is taken from same book but sixth edition
 to provide the necessary total motive force.
19. Pa ab lic gh c llec . A set of parabolic mirrors Section B
can be used to concentrate the sun's rays to heat a fluid flowing
in a pipe positioned at the mirrors' focal points (Camach ,
2012). The heated fluid, such as oil, for example, is transported
to a pressurized tank to be used to create steam to generate
electricity or power an industrial process. Since the solar energy
varies with time of day, time of year, cloudiness, humidity, etc., a
control system has to be developed in order to maintain the fluid
temperature constant. The temperature is mainly controlled by
varying the amount of fluid flow through the pipes, but possibly
also with a solar tracking mechanism that tilts the mirrors at
appropriate angles.
Assuming fixed mirror angles, draw the functional block
diagram of a system to maintain the fluid temperature a
constant. The desired and actual fluid temperature difference is
fed to a controller followed by an amplifier and signal
conditioning circuit that varies the speed of a fluid circulating
pump. Label the blocks and links of your diagram, indicating all
the inputs to the system, including external disturbances such
as solar variations, cloudiness, humidity, etc.

oe

3
 direction. One very important feature of the PT is its safety: The
then without delay present. system will maintain its vertical position within a specified angle
despite road disturbances, such as uphills and downhills or even

7. A Segway®1 Personal Transporter (PT) (Figure P1.2) is a two- Section C

if the operator over-leans in any direction. Draw a functional
block diagram of the PT system that keeps the system in a
wheeled vehicle in which the human operator stands vertically vertical position. Indicate the input and output signals,
intermediate signals, and main subsystems
on a platform. As the driver leans left, right, forward, or ([Link]

backward, a set of sensitive gyroscopic sensors sense the desired

input. These signals are fed to a computer that amplifies them
and commands motors to propel the vehicle in the desired
direction. One very important feature of the PT is its safety: The
system will maintain its vertical position within a specified angle
despite road disturbances, such as uphills and downhills or even
if the operator over-leans in any direction. Draw a functional
block diagram of the PT system that keeps the system in a
vertical position. Indicate the input and output signals,
intermediate signals, and main subsystems
([Link]

FIGURE P1.2 The Seg a Pe al T a e (PT)

8. In humans, hormone levels, alertness, and core body
temperature are synchronized through a 24-hour circadian
cycle. Daytime alertness is at its best when sleep/wake cycles are
in sync with the circadian cycle. Thus alertness can be easily

4
corresponding variables at the input and output of every block.
18. H b id ehicle. The use of hybrid cars is becoming Section D
increasingly popular. A hybrid electric vehicle (HEV) combines
electric machine(s) with an internal combustion engine (ICE),
making it possible (along with other fuel consumption–reducing
measures, such as stopping the ICE at traffic lights) to use
smaller and more efficient gasoline engines. Thus, the efficiency
advantages of the electric drivetrain are obtained, while the
energy needed to power the electric motor is stored in the
onboard fuel tank and not in a large and heavy battery pack.
There are various ways to arrange the flow of power in a hybrid
car. In a serial HEV (Figure P1.7), the ICE is not connected to
the drive shaft. It drives only the generator, which charges the
batteries and/or supplies power to the electric motor(s)
through an inverter or a converter.

FIGURE P1.7 Se ial h b id-elec ic ehicle5

The HEVs sold today are primarily of the parallel or split-power
variety. If the combustion engine can turn the drive wheels as
well as the generator, then the vehicle is referred to as a
a allel hybrid, because both an electric motor and the ICE can
drive the vehicle. A parallel hybrid car (Figure P1.8) includes a
relatively small battery pack (electrical storage) to put out extra
power to the electric motor when fast acceleration is needed.
See (B ch, 5 h ed., 2000), (B ch, 7 h ed., 2007), (Edel ,
2008), and (A de , 2009) for more detailed information
about HEV.

5
ows the described process as well as the chapters in which t
The antenna azimuth position control system discusse
Design of
presentative Process
control (Total
systems six
thatsteps)
must be analyzed and

Step 1 Step 2 Step 3

Determine
a physical Draw a Transform
system and functional the physical b
specifications block system into
from the diagram. a schematic.
requirements. o
r

Analog: Chapter 1 C
Digital:
6
URE 1.11 The control system design process
 Te a ef ed i c l e
I F ci De c i i Ske ch U e

Impulse () Transient response

Modeling

Step () Transient response

Steady-state error

Ramp () Steady-state error

Parabola Steady-state error

Sinusoid sin Transient response

Modeling Steady-state
7 error
manageable and still approximate physical reality.
The next step is to develop mathematical models from schematics of physical systems. We will discuss
two methods: (1) transfer functions in the frequency domain and (2) state equations in the time domain.
These topics are covered in this chapter and in Chapter 3, respectively. As we proceed, we will notice that
Mathematical Model
in every case the
In first step
Chapter
of science and engineering.
1 wein developing
discussed
For example,
schematic and demonstrated
a mathematical
the analysis
when
this step for a we model
position
model
and design sequence is to apply
that included
electrical
control
the the
obtaining
system. Tonetworks,
fundamental
system's
Ohm's
obtain a schematic, thelaw
physical laws
and Kirchhoff's
control
laws, which are basic laws of electric networks, will be applied initially. We will sum voltages in a loop or
systems engineer must often make many simplifying assumptions in order to keep the ensuing model
manageable and still approximate physical reality.

sum currents at a node. When we study mechanical systems, we will use Newton's laws as the
Transfer functions in frequency domain
The next step is to develop mathematical models from schematics of physical systems. We will discuss
fundamental guiding principles. Here we will sum forces or torques. From these equations we will obtain
two methods: (1) transfer functions in the frequency domain and (2) state equations in the time domain.
the relationshipThese
between thecovered
topics are system's
in thisoutput andininput.
chapter and Chapter 3, respectively. As we proceed, we will notice that
 State
In Chapter equations
1 we saw that in time
a differential equation domain
in every case the first step in developing a mathematical model is to apply the fundamental physical laws
can describe the relationship between the input and
of science and engineering. For example, when we model electrical networks, Ohm's law and Kirchhoff's
output of a system. The form
laws, which oflaws
are basic theofdifferential equation
electric networks, and its
will be applied coefficients
initially. are
We will sum a formulation
voltages in a loop or or
sum currents at a node. When we study mechanical systems, we will use Newton's laws as the
description of the [Link]
fundamental Although the Here
principles. differential
we will sumequation relates
forces or torques. Fromthethese
system to its
equations we input and output,
will obtain it
 Modelling Electric systems
is not a satisfying
the representation
relationship betweenfrom a system
the system's outputperspective.
and input. Looking at Eq. (1.2), a general, th-order,
linear, time-invariant
In Chapterdifferential
1 we saw that equation,
a differential we see can
equation thatdescribe
the system parameters,
the relationship which
between the inputare
and the
coefficients, appear
outputthroughout
of a system. Thethe
formequation. In addition,
of the differential theitsoutput,
equation and c(are
coefficients ), aand the input,
formulation or ( ), also
 Kirchhoff's law, Ohm’s law
description of the system. Although the differential equation relates the system to its input and output, it
appear throughout the equation.
is not a satisfying representation from a system perspective. Looking at Eq. (1.2), a general, th-order,
linear, time-invariant differential equation, we see that the system parameters, which are the
We would prefer a mathematical representation such as that shown in Figure 2.1(a), where the input,
coefficients, appear throughout the equation. In addition, the output, c( ), and the input, ( ), also
output, and system
appearare distincttheand separate parts. Also, we would like to represent conveniently the
 Modelling Mechanical systems
interconnectionWeofwould
throughout
severalprefer
equation.
subsystems.
a mathematicalFor example,
representation weaswould
such likeinto
that shown represent
Figure ca the
2.1(a), where cadedinput,
interconnections, as shown
output, inare
and system Figure
distinct2.1(b), where
and separate a mathematical
parts. Also, we would likefunction,
to representcalled a transfer
conveniently the function, is
 block,
inside each Newton’s laws, Equilibrium equations
interconnection of several subsystems. For example, we would like to represent ca caded
and block functions can easily be combined to yield Figure 2.1(a) for ease of analysis
interconnections, as shown in Figure 2.1(b), where a mathematical function, called a transfer function, is
and design. This convenience
inside each block, andcannot be obtained
block functions with
can easily the differential
be combined equation.
to yield Figure 2.1(a) for ease of analysis
and design. This convenience cannot be obtained with the differential equation.

8
 Laplace Transform Review
The Laplace transform is
= + j , a complex variable. Thus, knowing f( ) and that the integral in Eq. (2.1) exists, w
defined as
unction, F( ), that is called the La lace a f m of f( ).1
ation for the lower
where = + limit
j , means
a complexthatvariable.
even if f(Thus,
where s )=isσdiscontinuous
knowing
+ at variable.
f( ) and
jω, a complex = 0,the
that we integral
can startinthe
Eq
tion prior to athe
find discontinuity
function, as long
F( ), that as the
is called theintegral
La laceconverges.
a f mThus, of f( ).we
1 can find the Lap
rm of impulse functions. This property has distinct advantages when applying the Laplace
rm to theThe notation
solution of for the lowerequations
differential limit means wherethatthe
even if f( conditions
initial ) is discontinuous at = 0, we
are discontinuous
integration
ifferential equations, priorweto the to
have discontinuity
solve for theasinitial
long as the integral
conditions converges.
after Thus, wekno
the discontinuity can
transform
ial conditions of impulse
before functions. Using
the discontinuity. This property has distinct
the Laplace transform advantages
we need when applyin
only know th
transform to the solution of differential equations where
ons before the discontinuity. See Kaila h (1980) for a more detailed discussion. the initial conditions are d
TheUsing
inverse differential equations, we
Laplace transform, haveallows
which to solve
us for the initial
to find conditions
f(t) given F(s), isafter the disc
erse Laplace transform,
the initial which
conditions allows
before theusdiscontinuity.
to find f( ) givenUsing F(the
), is
Laplace transform we need
conditions before the discontinuity. See Kaila h (1980) for a more detailed discussi
The inverse Laplace transform, which allows us to find f( ) given F( ), is

9where
 TABLE 2.2

La lace a f able La lace a f he e

Ie . The e Na e
Ie . f(t) F(s)
1. Definition
1. δ( ) 1
2. Linearity theorem
2. () 3. Linearity theorem
3. () 4. Frequency shift theorem
5. Time shift theorem
4. ()
6. Scaling theorem
5. e−a ( ) 7. Differentiation theorem
6. sin () 8. Differentiation theorem
7. cos ()
9. Differentiation theorem

10. Integration theorem

11. Final value theorem1

12. Initial value theorem2

1 For this theorem to yield correct finite results, all roots of the denominator of F( ) must have n
more than one can be at the origin.
2 For this theorem to be valid, f( ) must be continuous or have a step discontinuity at = 0 (i.e.,
derivatives at = 0).

In addition to the Laplace transform table, Table 2.1, we can use Laplace transfo
10 Table 2.2, to assist in transforming between f( ) and F( ). In the next example, w
of the Laplace transform theorems shown in Table 2.2 to find f( ) given F( ).
 In the following example we de
function.

f(t) F(s) TABLE 2.1

La lace a f able
2
t2 Ie . f(t) F(s)
s3
1. δ( ) 1
Ae 4t A 2. ()
s−4
3. ()

sin(t) 1 4. ()
s2 + 1
5. e−a ( )
s 6. sin ()
cos( 3t)
s2 + 3
7. cos ()

11
 TABLE 2.2

La lace a f able La lace a f he e

Ie . The e Na e
Ie . f(t) F(s) 1. Definition
2. Linearity theorem
1. δ( ) 1 3. Linearity theorem

2. () 4. Frequency shift theorem

5. Time shift theorem
3. () 6. Scaling theorem

7. Differentiation theorem
4. ()
8. Differentiation theorem

5. e−a ( ) 9. Differentiation theorem

6. sin () 10. Integration theorem

11. Final value theorem1

7. cos ()
12. Initial value theorem2

1 For this theorem to yield correct finite results, all roots of the denominator of F( ) must have ne
more than one can be at the origin.
1 or have a step discontinuity at
2 For this theorem to be valid, f( ) must be continuous
Find the inverse Laplace transform, f(t)=? F(s) = = 0 (i.e., n

(s + 3)2
derivatives at = 0).

In addition to the Laplace transform table, Table 2.1, we can use Laplace transfor
f (t)
Table 2.2, to assist = e −3t t between f( ) and F( ). In the next example, w
in transforming
Hint: of the Laplace transform theorems shown in Table 2.2 to find f( ) given F( ).

12
 of terms and then find the inverse Laplace transform for each term. W
t multiply Eq. (2.8) by ( + 1), which
and isolates
show forK1. Thus,
each
case how an F( ) can be expanded into partial frac

Ca Te 1.I Roo
2.1 of he Denomina o of (2.9)
F(s) A e Real and Di inc
Partial fraction-Revision
An Use the following
example of an F( )MATLAB
with realand
andControl
distinct System
roots inToolbox statemen
the denominator
invariant
h −1 eliminates the last term and yields (LTI) transfer
K1 = 2. Similarly, K2 can be function
found by of Eq. (2.22).
.8) byCase
( + 2)[Link]
Roots
thenofletting
the Denominator
approach −2; ofhence,
F(s) AreK2 Real
= − [Link] Distinct
F= ([],[ 1 2 2],2)
part of Eq. (2.8) is an F( ) in Table 2.1. Hence, f( ) is the sum of the inverse Laplace
term, or
Ca
The eroots
2. Roo
of theof he Denomina
denominator o of(2.10)
F(s)
are distinct, A eeach
since Realfactor
and is
Reraised
ea edo
the example
An partial-fraction
of an F(expansion
) with realasand a sum of terms
repeated where
roots each
in the factor of
denominat
iven an F( ) whose denominatordenominator of each
has real and distinct roots,term, and constants, called residues, form the
a partial-fraction
Case 2. Roots of the Denominator of F(s) Are Real and Repeated

(2.11)

The roots of ( + 2)2 in the denominator are repeated, since the facto
(2.26)
order of N( higher
) is less than the order of D(than 1. In this
). To evaluate case,
each theKdenominator
residue, i, we multiply root at −2 is a m l i le
enominator of the multiplicity
r root is of higher corresponding partial
than [Link]
2, successive Thus, if we want to find
would Km, each
isolate we
by ( Case
ansion+of m )[Link]
the get of root.
Roots
multiple
We can write the partial-fraction
the Denominator of F(s) Are Complex or Imaginary expansion
as a sum of terms, where
forms the denominator of each term. In addition, each multiple root
consisting
(2.37) to Table 2.1 and Eq. (2.36), we find of denominator factors of reduced
(2.12) multiplicity. For examp

ng MATLAB statements to help you get Eq. (2.26).

(2.38)
2 2]);
e... This function can be expanded in the following form:
alize the 13
solution, an alternate form of f( ), obtained by trigonometric identities, is
then
h − m, all terms on the right-hand Kof
side
1 = 2, which can be found as previously
Eq. (2.12) go to zero except the term Km, described. K2 can be is
s begin
ction byalso
will writing
allowaus
partial-fraction
general
to algebth-order,
expansion and appliedthe
linear,
aicall combine time-invariant differential
mathematical representations
theconcepts
conceptstotothe
thesolution
solutionof
ofdifferential
equation,
of subsyste
differential equations.
equations. We are
partial-fraction expansion and applied
tal system
ready torepresentation.
ready to formulate
formulate thethe system
system representationshown
representation shownininFigure
Figure2.1
2.1by
byestablishing
establishing aa viable
viable definition
function
function
us begin that
that
by writing algebraically
algebraically
a general relates
relates a asystem's
th-order,system's output
output
linear, totoits
itsinput.
[Link]
time-invariant This function will
function
differentialwill allow separation
allow
equation,
thethe input, system, and output into threeseparate
separateandanddistinct
distinctparts,
parts,unlike
unlikethe
the differential
differential equation
Transfer Function
input,
function
function
system,
will
will also
also
and
allow
output
allow usus
toto
into three
algebaicall
algeb aicall combine
combinemathematical
mathematicalrepresentations
representations ofof subsystems to
a total
a total system
system representation.
representation.
e c( A)Let
isLet
thebegin
general
us us
output,
nth-order,
beginbyby
( )linear,
writing
writing
isa general
the input, and the
time-invariant
a general th-order,
th-order,
ai's,time-invariant
bi's, equation,
differential
linear,
linear,
and thedifferential
time-invariant form of the
differential differential eq
equation,
equation,
sent the system. Taking the Laplace transform of both sides,
ere c( ) is the output, ( ) is the input, and the ai's, bi's, and the form of the differential equa((
resent the system. Taking the Laplace transform of both sides,
where
where c( c(
) is) the
is the output,( )( is
output, ) is the
the input,and
input, andthe
theaa's,
's, bi's,and
andthe
theform
formof
ofthe
the differential
differential equation
equation
i i bi's,
Taking the Laplace
represent transform ofLaplace
both sides
represent thethe system.
system. Taking
Taking the the
Laplace transformofofboth
transform bothsides,
sides,
tion (2.51) is a purely algebraic expression.
FIGURE If2.1
[Link]
Bl ck that
diag all
a i ei ial c adi
e e i i f a
a e
e e e ai fa i e c ec i f b e
) reduces to
uation (2.51) is a purely algebraic expression. If we assume that all i i ial c di i a e e
2.2 La lace T an fo m Re ie
1) reduces to
Equation (2.51) is a purely algebraicAexpression. If we assume
system represented that allequation
by a differential i i ial cis difficult
di i toa model
e e as, Eq
ab
Equation (2.51) is
(2.51) reduces to a purely algebraic expression. If we assume that all i i ial c di i a e e , Eq
(2.51) reduces to lay the groundwork for the Laplace transform, with which we can represe
system as separate entities. Further, their interrelationship will be simply
form the ratio of the output transform, C( ),transform
the Laplace dividedand bythen
theshow
inputhowtransform,
it simplifies theR( ):
representation
(
1996). (
w form the ratio of the output transform, C( ), divided by the input transform, R( ):
The Laplace transform is defined as
Now form the ratio of the output transform, C( ), divided by the input transform, R( ):
Now form the ratio of the output transform, C( ), divided by the input transform, R( ):
(
(

e that Eq. (2.53) separates the output, C( ), the input, R( ), and the system, which is the
omials in
14
ice thatNoticeon
thatthe
Eq. (2.53) [Link].
(2.53) We
separates calloutput,
separates
the this ratio,
the output,
C( ),G(
C( ), ),
the thethe aR( ),fe
input,R(
input, f the
), and
and c system,
theisystem,
and evaluate
which is the
which it wiro
isratio
the
di i .
er function can be represented as a block diagram, as shown in Figure 2.2, with
e output on the right, and the system transfer function inside the block. Notice t
tor ofTransfer Function
the transfer function is identical to the characteristic polynomial of the dif
Also, we can find the output, C( ) by using

R(s) C(s)
G(s)

Two conditions to compute the transfer function of a system

(i) system must be in s-domain

(ii) Initial conditions are zero.

15
 a.

b.

7. A system is described by the following differential

equation:

Find the expression for the transfer function of the system,

Y( )/X( ). [Section: 2.3]

Check Answer!
3 2
8. WriteG(s)
Y(s) s + 4s + 6s + 8
=
the differential =
equation that corresponds to each of the
following transferX(s)
functions.s 3[Section:
+ 3s 2 +2.3]5s + 1

a.
16
The complete solution is at [Link]/go/Nise/ControlSystem
ANSWER:

Skill-A e men E e ci e 2.5

The complete solution is at [Link]/go/Nise/ControlSystemsEngin

PROBLEM:
Skill-A
Find the ramp e men
response E e ci whose
for a system e 2.5 transfer function is
PROBLEM:
Find the ramp response for a system whose transfer function is

ANSWER:
ANSWER:

The
17 complete solution
The complete is at [Link]/go/Nise/ControlSystem
solution is at [Link]/go/Nise/ControlSystemsEngin
 Immediately, we can solve for K1 if we let approach − 1. We can solve for K2 if we differentiate Eq.
(2.28) with respect to and then let approach − 1. Subsequent differentiation will allow us to find K3
through K . The general expression for K1 through K for the multiple roots is

(2.29)

Ca e 3. Roo of he Denomina o of F( ) A e Com le o Imagina

T I 2.3
Use the following MATLAB and Control System Toolbox statement to form the LTI transfer
function of Eq. (2.30).
F= ([3],[1 2 5 0])

An example of F( ) with complex roots in the denominator is

18
E am le 2.5 S em Re on e f om he T an fe F nc ion
PROBLEM:
Use the result of Example 2.4 to find the response, c( ) to an input, ( ) = ( ), a unit step, assuming
zero initial conditions.

T I 2.6
Use the following MATLAB and Symbolic Math Toolbox statements to help you get Eq. (2.60).

C=1/( *( + 2))
C= a ace(C)

T I 2.7
Use the following MATLAB statements to plot Eq. (2.60) for from 0 to 1 sat intervals of 0.01
s.
=0:0.01:3; ...
( ,(1/2 -1/2*e (-2* )))

SOLUTION:
To19solve the problem, we use Eq. (2.54), where G( ) = 1/( + 2) as found in Example 2.4. Since ( )
= ( ), R( ) = 1/ , from Table 2.1. Since the initial conditions are zero,