WEBC01 10/28/2014 14:10:29 Page 16

16 Chapter 1 Introduction

Step 4: Develop a Mathematical Model (Block Diagram)

Once the schematic is drawn, the designer uses physical laws, such as Kirchhoff’s laws for
electrical networks and Newton’s law for mechanical systems, along with simplifying
assumptions, to model the system mathematically. These laws are

Kirchhoff's voltage law The sum of voltages around a closed path equals zero.
Kirchhoff's current law The sum of electric currents flowing from a node equals zero.
Newton's laws The sum of forces on a body equals zero;3 the sum of moments on a
body equals zero.

Kirchhoff’s and Newton’s laws lead to mathematical models that describe the relationship
between the input and output of dynamic systems. One such model is the linear, time-invariant
differential equation, Eq. (1.2):

d m c…t† d m 1 c…t† d m r…t† d m 1 r…t†

‡ dn 1 ‡ ∙ ∙ ∙ ‡ d 0 c…t† ˆ bm ‡ bm 1 ‡ ∙ ∙ ∙ ‡ b0 r…t†
dt n dt n 1 dt m dt m 1

…1.2†4

Many systems can be approximately described by this equation, which relates the output,
c(t), to the input, r(t), by way of the system parameters, ai and bj . We assume the reader is
familiar with differential equations. Problems and a bibliography are provided at the end of
the chapter for you to review this subject.
Simplifying assumptions made in the process of obtaining a mathematical model
usually leads to a low-order form of Eq. (1.2). Without the assumptions the system model
could be of high order or described with nonlinear, time-varying, or partial differential
equations. These equations complicate the design process and reduce the designer’s insight.
Of course, all assumptions must be checked and all simplifications justified through analysis
or testing. If the assumptions for simplification cannot be justified, then the model cannot be
simplified. We examine some of these simplifying assumptions in Chapter 2.
In addition to the differential equation, the transfer function is another way of
mathematically modeling a system. The model is derived from the linear, time-invariant
differential equation using what we call the Laplace transform. Although the transfer
function can be used only for linear systems, it yields more intuitive information than the
differential equation. We will be able to change system parameters and rapidly sense the
effect of these changes on the system response. The transfer function is also useful in
modeling the interconnection of subsystems by forming a block diagram similar to
Figure 1.8(d) but with a mathematical function inside each block.
Still another model is the state-space representation. One advantage of state-space
methods is that they can also be used for systems that cannot be described by linear
differential equations. Further, state-space methods are used to model systems for
simulation on the digital computer. Basically, this representation turns an nth-order
differential equation into n simultaneous first-order differential equations. Let this
description suffice for now; we describe this approach in more detail in Chapter 3.

3 P
Alternately, forces ˆ Ma. In this text the force, Ma, will be brought to the left-hand side of the equation to yield
P
forces ˆ 0 (D’Alembert’s principle). We can then have a consistent analogy between force and voltage, and
P P
Kirchhoff’s and Newton’s laws (that is, forces ˆ 0; voltages ˆ 0).
4
The right-hand side of Eq. (1.2) indicates differentiation of the input, r(t). In physical systems, differentiation of
the input introduces noise. In Chapters 3 and 5 we show implementations and interpretations of Eq. (1.2) that do not
require differentiation of the input.
WEBC01 10/28/2014 14:10:30 Page 17

1.5 The Design Process 17

Finally, we should mention that to produce the mathematical model for a system,
we require knowledge of the parameter values, such as equivalent resistance, inductance,
mass, and damping, which is often not easy to obtain. Analysis, measurements, or
specifications from vendors are sources that the control systems engineer may use to
obtain the parameters.

Step 5: Reduce the Block Diagram Angular Angular

input output
Subsystem models are interconnected to form block diagrams of larger systems, Mathematical
as in Figure 1.8(d), where each block has a mathematical description. Notice that description
many signals, such as proportional voltages and error, are internal to the system.
FIGURE 1.11 Equivalent block
There are also two signals—angular input and angular output—that are external
diagram for the antenna azimuth
to the system. In order to evaluate system response in this example, we need to
position control system
reduce this large system’s block diagram to a single block with a mathematical
description that represents the system from its input to its output, as shown in
Figure 1.11. Once the block diagram is reduced, we are ready to analyze and
design the system.

Step 6: Analyze and Design

The next phase of the process, following block diagram reduction, is analysis and design. If
you are interested only in the performance of an individual subsystem, you can skip the
block diagram reduction and move immediately into analysis and design. In this phase, the
engineer analyzes the system to see if the response specifications and performance
requirements can be met by simple adjustments of system parameters. If specifications
cannot be met, the designer then designs additional hardware in order to effect a desired
performance.
Test input signals are used, both analytically and during testing, to verify the
design. It is neither necessarily practical nor illuminating to choose complicated input
signals to analyze a system’s performance. Thus, the engineer usually selects standard test
inputs. These inputs are impulses, steps, ramps, parabolas, and sinusoids, as shown in
Table 1.1.
An impulse is infinite at t ˆ 0 and zero elsewhere. The area under the unit impulse is 1.
An approximation of this type of waveform is used to place initial energy into a system so
that the response due to that initial energy is only the transient response of a system. From
this response the designer can derive a mathematical model of the system.
A step input represents a constant command, such as position, velocity, or acceleration.
Typically, the step input command is of the same form as the output. For example, if the
system’s output is position, as it is for the antenna azimuth position control system, the step
input represents a desired position, and the output represents the actual position. If the system’s
output is velocity, as is the spindle speed for a video disc player, the step input represents a
constant desired speed, and the output represents the actual speed. The designer uses step
inputs because both the transient response and the steady-state response are clearly visible and
can be evaluated.
The ramp input represents a linearly increasing command. For example, if the
system’s output is position, the input ramp represents a linearly increasing position, such as
that found when tracking a satellite moving across the sky at constant speed. If the system’s
output is velocity, the input ramp represents a linearly increasing velocity. The response to
an input ramp test signal yields additional information about the steady-state error. The
previous discussion can be extended to parabolic inputs, which are also used to evaluate a
system’s steady-state error.
Sinusoidal inputs can also be used to test a physical system to arrive at a mathematical
model. We discuss the use of this waveform in detail in Chapters 10 and 11.
WEBC01 10/28/2014 14:10:32 Page 18

18 Chapter 1 Introduction

TABLE 1.1 Test waveforms used in control systems

Input Function Description Sketch Use
Impulse δ…t† δ…t† ˆ ∞ for 0 < t < 0‡ f(t) Transient response
ˆ 0 elsewhere Modeling
Z 0‡
δ…t†dt ˆ 1 δ(t)
0

Step u…t† u…t† ˆ 1 for t > 0 f(t) Transient response

ˆ 0 for t < 0 Steady-state error

Ramp tu…t† tu…t† ˆ t for t  0 f(t) Steady-state error

ˆ 0 elsewhere

Parabola 1 2 1 2 1 f(t) Steady-state error

t u…t† t u…t† ˆ t 2 for t  0
2 2 2
ˆ 0 elsewhere

Sinusoid sin ωt f(t) Transient response

Modeling
Steady-state error

We conclude that one of the basic analysis and design requirements is to evaluate the
time response of a system for a given input. Throughout the book you will learn numerous
methods for accomplishing this goal.
The control systems engineer must take into consideration other characteristics
about feedback control systems. For example, control system behavior is altered by
fluctuations in component values or system parameters. These variations can be caused
by temperature, pressure, or other environmental changes. Systems must be built so that
expected fluctuations do not degrade performance beyond specified bounds. A sensitivity
analysis can yield the percentage of change in a specification as a function of a change in a
system parameter. One of the designer’s goals, then, is to build a system with minimum
sensitivity over an expected range of environmental changes.
In this section we looked at some control systems analysis and design considerations.
We saw that the designer is concerned about transient response, steady-state error,
stability, and sensitivity. The text pointed out that although the basis of evaluating system
performance is the differential equation, other methods, such as transfer functions and state
space, will be used. The advantages of these new techniques over differential equations will
become apparent as we discuss them in later chapters.
 1.6 Computer-Aided Design 19

1.6 Computer-Aided Design

Now that we have discussed the analysis and design sequence, let us discuss the use of the
computer as a computational tool in this sequence. The computer plays an important role in
the design of modern control systems. In the past, control system design was labor intensive.
Many of the tools we use today were implemented through hand calculations or, at best,
using plastic graphical aid tools. The process was slow, and the results not always accurate.
Large mainframe computers were then used to simulate the designs.
Today we are fortunate to have computers and software that remove the drudgery
from the task. At our own desktop computers, we can perform analysis, design, and
simulation with one program. With the ability to simulate a design rapidly, we can easily
make changes and immediately test a new design. We can play what-if games and try
alternate solutions to see if they yield better results, such as reduced sensitivity to
parameter changes. We can include nonlinearities and other effects and test our models
for accuracy.

MATLAB
The computer is an integral part of modern control system design, and many computational
tools are available for your use. In this book we use MATLAB and the MATLAB Control
System Toolbox, which expands MATLAB to include control system–specific commands.
In addition, presented are several MATLAB enhancements that give added functionality to
MATLAB and the Control Systems Toolbox. Included are (1) Simulink, which uses a
graphical user interface (GUI); (2) the LTI Viewer, which permits measurements to be made
directly from time and frequency response curves; (3) the SISO Design Tool, a convenient
and intuitive analysis and design tool; and (4) the Symbolic Math Toolbox, which saves
labor when making symbolic calculations required in control system analysis and
design. Some of these enhancements may require additional software available from
The MathWorks, Inc.
MATLAB is presented as an alternate method of solving control system problems.
You are encouraged to solve problems first by hand and then by MATLAB so that insight is
not lost through mechanized use of computer programs. To this end, many examples
throughout the book are solved by hand, followed by suggested use of MATLAB.
As an enticement to begin using MATLAB, simple program statements that you can
try are suggested throughout the chapters at appropriate locations. Throughout the book,
various icons appear in the margin to identify MATLAB references that direct you to the
proper program in the proper appendix and tell you what you will learn. Selected end-of-
chapter problems and Case Study Challenges to be solved using MATLAB have also been
marked with appropriate icons. The following list itemizes the specific components of
MATLAB used in this book, the icon used to identify each, and the appendix in which a
description can be found:
MATLAB/Control System Toolbox tutorials and code are found in
Appendix B and identified in the text with the MATLAB icon shown in
the margin.

Simulink tutorials and diagrams are found in Appendix C and

identified in the text with the Simulink icon shown in the margin.
MATLAB GUI tools, tutorials, and examples are in Appendix E at
www.wiley.com/college/nise and identified in the text with the GUI
Tool icon shown in the margin. These tools consist of the LTI
Viewer and the SISO Design Tool.
20 Chapter 1 Introduction

Symbolic Math Toolbox tutorials and code are found in Appendix F

at www.wiley.com/college/nise and identified in the text with the
Symbolic Math icon shown in the margin.

MATLAB code itself is not platform specific. The same code runs on PCs and workstations
that support MATLAB. Although there are differences in installing and managing
MATLAB files, we do not address them in this book. Also, there are many more commands
in MATLAB and the MATLAB toolboxes than are covered in the appendixes. Please
explore the bibliographies at the end of the applicable appendixes to find out more about
MATLAB file management and MATLAB instructions not covered in this textbook.

LabVIEW
LabVIEW is a programming environment presented as an alternative to MATLAB. This
graphical alternative produces front panels of virtual instruments on your computer that
are pictorial reproductions of hardware instruments, such as waveform generators or
oscilloscopes. Underlying the front panels are block diagrams. The blocks contain
underlying code for the controls and indicators on the front panel. Thus, a knowledge of
coding is not required. Also, parameters can be easily passed or viewed from the front panel.
A LabVIEW tutorial is in Appendix D and all LabVIEW material is
identified with the LabVIEW icon shown in the margin.

You are encouraged to use computational aids throughout this book. Those not using
MATLAB or LabVIEW should consult Appendix H at www.wiley.com/college/nise for a
discussion of other alternatives. Now that we have introduced control systems to you and
established a need for computational aids to perform analysis and design, we will conclude
with a discussion of your career as a control systems engineer and look at the opportunities
and challenges that await you.

1.7 The Control Systems Engineer

Control systems engineering is an exciting field in which to apply your engineering
talents, because it cuts across numerous disciplines and numerous functions within those
disciplines. The control engineer can be found at the top level of large projects, engaged at
the conceptual phase in determining or implementing overall system requirements. These
requirements include total system performance specifications, subsystem functions, and
the interconnection of these functions, including interface requirements, hardware and
software design, and test plans and procedures.
Many engineers are engaged in only one area, such as circuit design or software
development. However, as a control systems engineer, you may find yourself working in a
broad arena and interacting with people from numerous branches of engineering and the
sciences. For example, if you are working on a biological system, you will need to interact
with colleagues in the biological sciences, mechanical engineering, electrical engineering,
and computer engineering, not to mention mathematics and physics. You will be working
with these engineers at all levels of project development from concept through design and,
finally, testing. At the design level, the control systems engineer can be performing
hardware selection, design, and interface, including total subsystem design to meet specified
requirements. The control engineer can be working with sensors and motors as well as
electronic, pneumatic, and hydraulic circuits.
The space shuttle provides another example of the diversity required of the systems
engineer. In the previous section, we showed that the space shuttle’s control systems cut
across many branches of science: orbital mechanics and propulsion, aerodynamics,