Ebook A First Course in Control System Design 2Nd Edition Kamran Iqbal Online PDF All Chapter

A First Course in Control System
Design 2nd Edition Kamran Iqbal

Visit to download the full and correct content document:
https://ebookmeta.com/product/a-first-course-in-control-system-design-2nd-edition-ka
mran-iqbal/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Robust Control System Design 3rd Edition Chia-Chi Tsui
https://ebookmeta.com/product/robust-control-system-design-3rd-
edition-chia-chi-tsui/
A First Course in Machine Learning 2nd Edition

Solutions Manual Simon Rogers
https://ebookmeta.com/product/a-first-course-in-machine-
learning-2nd-edition-solutions-manual-simon-rogers/
Abstract Algebra: A First Course 2nd Edition Stephen

Lovett
https://ebookmeta.com/product/abstract-algebra-a-first-
course-2nd-edition-stephen-lovett/
Nonlinear Functional Analysis - A First Course 2nd

Edition S. Kesavan
https://ebookmeta.com/product/nonlinear-functional-analysis-a-
first-course-2nd-edition-s-kesavan/
First Course in Statistical Programming With R 2nd
Edition Braun
https://ebookmeta.com/product/first-course-in-statistical-
programming-with-r-2nd-edition-braun/
System Reliability and Security: Techniques and

Methodologies 1st Edition Javaid Iqbal
https://ebookmeta.com/product/system-reliability-and-security-
techniques-and-methodologies-1st-edition-javaid-iqbal/
Advances in Power System Modelling, Control and

Stability Analysis 2nd Edition Federico Milano
https://ebookmeta.com/product/advances-in-power-system-modelling-
control-and-stability-analysis-2nd-edition-federico-milano/
A First Course in Spectral Theory 1st Edition Milivoje

Luki■
https://ebookmeta.com/product/a-first-course-in-spectral-
theory-1st-edition-milivoje-lukic/
A First Course in General Relativity 3rd Edition

Bernard Schutz
https://ebookmeta.com/product/a-first-course-in-general-
relativity-3rd-edition-bernard-schutz/
River Publishers Series in Automation, Control and Robotics
A First Course in
A First Course in Control System Design

Control System Design A First Course in
2nd Edition Control System Design
Kamran Iqbal 2nd Edition
Control systems are pervasive in our lives. Our homes have
2nd Edition
environmental controls. The appliances we use, such as the washing
machine, microwave, etc. carry embedded controllers in them. We
Kamran Iqbal
fly in airplanes and drive automobiles that extensively use control
systems. The industrial plants that produce consumer goods run on
process control systems. The recent drive toward automation has
increased our reliance on control systems technology.
This book discusses control systems design from a model-based
perspective for dynamic system models of single-input single-output Kp
type. The emphasis in this book is on understanding and applying
the techniques that enable the design of effective control systems
in multiple engineering disciplines. The book covers both time- e u(t) u
e(t)
domain and the frequency-domain design methods, as well as 1 Kd 1
controller design for both continuous-time and discrete-time
systems. MATLAB© and its Control Systems Toolbox are extensively
used for design. Technical topics discussed in the book include:
∙ Mathematical models of physical systems ∫udt Ki
Kamran Iqbal
∙ Analysis of transfer function and state variable models
∙ Control systems design objectives
dθ r θr − θa
∙ Control system design with root locus Cr
dt
+
Rw
= qi .
∙ Control system design in the state-space
∙ Control system design of sampled-data systems ℎ
∙ Compensator design with frequency response methods
d∆ θ
Rw Cr + ∆ θ = Rw qi .
dt
Foreword by
Professor Emeritus Hooshang Hemami,
River Publishers River Ohio State University River Publishers
A First Course in
Control System Design
2nd Edition
RIVER PUBLISHERS SERIES IN AUTOMATION,
CONTROL AND ROBOTICS
Series Editors:
ISHWAR K. SETHI
Oakland University
USA
TAREK SOBH
University of Bridgeport
USA
QUAN MIN ZHU

University of the West of England
UK
Indexing: All books published in this series are submitted to the Web of Science
Book Citation Index (BkCI), to SCOPUS, to CrossRef and to Google Scholar for
evaluation and indexing.
The “River Publishers Series in Automation, Control and Robotics” is a series of

comprehensive academic and professional books which focus on the theory and
applications of automation, control and robotics. The series focuses on topics ranging
from the theory and use of control systems, automation engineering, robotics and
intelligent machines.
Books published in the series include research monographs, edited volumes,
handbooks and textbooks. The books provide professionals, researchers, educators,
and advanced students in the field with an invaluable insight into the latest research
and developments.
Topics covered in the series include, but are by no means restricted to the
following:
• Robots and Intelligent Machines

• Robotics
• Control Systems
• Control Theory
• Automation Engineering
For a list of other books in this series, visit www.riverpublishers.com

A First Course in
Control System Design
2nd Edition
Kamran Iqbal
University of Arkansas
Little Rock, USA
River Publishers
Published, sold and distributed by:
River Publishers
Alsbjergvej 10
9260 Gistrup
Denmark
www.riverpublishers.com
ISBN: 978-87-7022-152-8 (Hardback)

978-87-7022-151-1 (Ebook)
©2020 River Publishers
All rights reserved. No part of this publication may be reproduced, stored in

a retrieval system, or transmitted in any form or by any means, mechanical,
photocopying, recording or otherwise, without prior written permission of
the publishers.
Contents
Foreword xi
Preface xiii
Acknowledgement xxi
List of Figures xxiii
List of Tables xxix
List of Abbreviations xxxi
1 Mathematical Models of Physical Systems 1

1.1 Modeling of Physical Systems . . . . . . . . . . . . . . . . 2
1.1.1 Model Variables and Element Types . . . . . . . . . 3
1.1.2 First-Order ODE Models . . . . . . . . . . . . . . . 4
1.1.3 Solving First-Order ODE Models with Step Input . . 8
1.1.4 Second-Order ODE Models . . . . . . . . . . . . . 10
1.1.5 Solving Second-Order ODE Models . . . . . . . . . 12
1.2 Transfer Function Models . . . . . . . . . . . . . . . . . . . 15
1.2.1 DC Motor Model . . . . . . . . . . . . . . . . . . . 16
1.2.2 Industrial Process Models . . . . . . . . . . . . . . 20
1.3 State Variable Models . . . . . . . . . . . . . . . . . . . . . 21
1.4 Linearization of Nonlinear Models . . . . . . . . . . . . . . 24
1.4.1 Linearization About an Operating Point . . . . . . . 25
1.4.2 Linearization of a General Nonlinear Model . . . . . 27
Skill Assessment Questions . . . . . . . . . . . . . . . . . . . . . 29
2 Analysis of Transfer Function Models 31

2.1 Characterization of Transfer Function Models . . . . . . . . 32
2.1.1 System Poles and Zeros . . . . . . . . . . . . . . . 32
2.1.2 System Natural Response . . . . . . . . . . . . . . 34
v
vi Contents
2.2 System Response to Inputs . . . . . . . . . . . . . . . . . . 36

2.2.1 The Impulse Response . . . . . . . . . . . . . . . . 36
2.2.2 The Step Response . . . . . . . . . . . . . . . . . . 38
2.2.3 Characterizing the System Transient Response . . . 44
2.2.4 System Stability . . . . . . . . . . . . . . . . . . . 46
2.3 Sinusoidal Response of a System . . . . . . . . . . . . . . . 49
2.3.1 Sinusoidal Response of Low-Order Systems . . . . . 50
2.3.2 Visualizing the Frequency Response . . . . . . . . . 52
3 Analysis of State Variable Models 63

3.1 State Variable Models . . . . . . . . . . . . . . . . . . . . . 64
3.1.1 Solution to the State Equations . . . . . . . . . . . . 65
3.1.2 Laplace Transform Solution and Transfer Function . 66
3.1.3 The State-Transition Matrix . . . . . . . . . . . . . 68
3.1.4 Homogenous State Equation and Asymptotic
Stability . . . . . . . . . . . . . . . . . . . . . . . . 70
3.1.5 System Response for State Variable Models . . . . . 74
3.2 State Variable Realization of Transfer Function Models . . . 77
3.2.1 Simulation Diagrams . . . . . . . . . . . . . . . . . 78
3.2.2 Controller Form Realization . . . . . . . . . . . . . 80
3.2.3 Dual (Observer Form) Realization . . . . . . . . . . 83
3.2.4 Modal Realization . . . . . . . . . . . . . . . . . . 83
3.2.5 Diagonalization and Decoupling . . . . . . . . . . . 85
3.3 Linear Transformation of State Variables . . . . . . . . . . . 86
3.3.1 Transformation into Controller Form . . . . . . . . 86
3.3.2 Transformation into Modal Form . . . . . . . . . . 88
4 Feedback Control Systems 93

4.1 Static Gain Controller . . . . . . . . . . . . . . . . . . . . . 95
4.2 Dynamic Controllers . . . . . . . . . . . . . . . . . . . . . 96
4.2.1 First-Order Phase-Lead and Phase-Lag
Controllers . . . . . . . . . . . . . . . . . . . . . . 97
4.2.2 The PID Controller . . . . . . . . . . . . . . . . . . 99
4.2.3 Rate Feedback Controllers . . . . . . . . . . . . . . 103
Contents vii
5 Control System Design Objectives 111

5.1 Stability of the Closed-Loop System . . . . . . . . . . . . . 112
5.1.1 Closed-Loop Characteristic Polynomial . . . . . . . 112
5.1.2 Stability Determination by Algebraic Methods . . . 114
5.1.3 Stability Determination from the Bode Plot . . . . . 116
5.2 Transient Response Improvement . . . . . . . . . . . . . . . 117
5.2.1 System Design Specifications . . . . . . . . . . . . 119
5.2.2 The Desired Characteristic Polynomial . . . . . . . 121
5.2.3 Optimal Performance Indices . . . . . . . . . . . . 123
5.3 Steady-State Error Improvement . . . . . . . . . . . . . . . 124
5.3.1 The Steady-State Error . . . . . . . . . . . . . . . . 124
5.3.2 System Error Constants . . . . . . . . . . . . . . . 125
5.3.3 Steady-State Error to Ramp Input . . . . . . . . . . 126
5.4 Disturbance Rejection . . . . . . . . . . . . . . . . . . . . . 128
5.5 Sensitivity and Robustness . . . . . . . . . . . . . . . . . . 130
6 Control System Design with Root Locus 133

6.1 The Root Locus . . . . . . . . . . . . . . . . . . . . . . . . 135
6.1.1 Roots of the Characteristic Polynomial . . . . . . . 135
6.1.2 Root Locus Rules . . . . . . . . . . . . . . . . . . . 136
6.1.3 Obtaining Root Locus Plot in MATLAB . . . . . . . 138
6.1.4 Stability from the Root Locus Plot . . . . . . . . . . 139
6.1.5 Analytic Root Locus Conditions . . . . . . . . . . . 141
6.2 Static Controller Design . . . . . . . . . . . . . . . . . . . 143
6.3 Dynamic Controller Design . . . . . . . . . . . . . . . . . . 144
6.3.1 Transient Response Improvement . . . . . . . . . . 145
6.3.2 Steady-State Error Improvement . . . . . . . . . . . 151
6.3.3 Lead–Lag and PID Designs . . . . . . . . . . . . . 152
6.3.4 Rate Feedback Compensation . . . . . . . . . . . . 156
6.3.5 Controller Designs Compared . . . . . . . . . . . . 161
6.4 Controller Realization . . . . . . . . . . . . . . . . . . . . . 163
6.4.1 Phase-Lead/Phase-Lag Controllers . . . . . . . . . . 164
6.4.2 PD, PI, PID Controllers . . . . . . . . . . . . . . . 164
7 Design of Sampled-Data Systems 167

7.1 Models of Sampled-Data Systems . . . . . . . . . . . . . . 169
7.1.1 Z-transform . . . . . . . . . . . . . . . . . . . . . . 169
viii Contents
7.1.2 Zero-Order Hold . . . . . . . . . . . . . . . . . . . 171

7.1.3 Pulse Transfer Function . . . . . . . . . . . . . . . 172
7.2 Sampled-Data System Response . . . . . . . . . . . . . . . 175
7.2.1 Difference Equation Solution by Iteration . . . . . . 175
7.2.2 Unit-Pulse Response . . . . . . . . . . . . . . . . . 176
7.2.3 Unit-Step Response . . . . . . . . . . . . . . . . . . 179
7.2.4 Response to Arbitrary Inputs . . . . . . . . . . . . . 183
7.3 Stability in the Case of Sampled-Data Systems . . . . . . . 184
7.3.1 Jury’s Stability Test . . . . . . . . . . . . . . . . . . 184
7.3.2 Stability Through Bilinear Transform . . . . . . . . 185
7.4 Closed-Loop Sampled-Data Systems . . . . . . . . . . . . . 186
7.4.1 Closed-Loop System Stability . . . . . . . . . . . . 186
7.4.2 Unit-Step Response . . . . . . . . . . . . . . . . . . 187
7.4.3 Steady-State Tracking Error . . . . . . . . . . . . . 190
7.5 Controllers for Sampled-Data Systems . . . . . . . . . . . . 192
7.5.1 Root Locus Design of Digital Controllers . . . . . . 193
7.5.2 Analog and Digital Controller Design Compared . . 196
7.5.3 Digital Controller Design by Emulation . . . . . . . 200
7.5.4 Emulation of Analog PID Controller . . . . . . . . . 203
8 Controller Design for State Variable Models 211

8.1 State Feedback Controller Design . . . . . . . . . . . . . . 212
8.1.1 Pole Placement with State Feedback . . . . . . . . . 213
8.1.2 Pole Placement in the Controller Form . . . . . . . . 215
8.1.3 Pole Placement using Bass–Gura Formula . . . . . . 217
8.1.4 Pole Placement using Ackermann’s Formula . . . . 218
8.1.5 Pole Placement using Sylvester’s Equation . . . . . 220
8.2 Tracking System Design . . . . . . . . . . . . . . . . . . . 222
8.2.1 Tracking System Design with Feedforward
Gain . . . . . . . . . . . . . . . . . . . . . . . . . . 222
8.2.2 Tracking PI Controller Design . . . . . . . . . . . . 225
8.3 State Variable Models of Sampled-Data Systems . . . . . . 230
8.3.1 Discretizing the State Equations . . . . . . . . . . . 230
8.3.2 Solution to the Discrete State Equations . . . . . . . 232
8.3.3 Pulse Transfer Function from State Equations . . . . 234
8.4 Controllers for Discrete State Variable Models . . . . . . . . 235
8.4.1 Emulating an Analog Controller . . . . . . . . . . . 235
8.4.2 Pole Placement Design of Digital Controller . . . . . 236
Contents ix
8.4.3 Deadbeat Controller Design . . . . . . . . . . . . . 238

8.4.4 Tracking PI Controller Design . . . . . . . . . . . . 241
9 Frequency Response Design of Compensators 247

9.1 Frequency Response Representation . . . . . . . . . . . . . 248
9.1.1 The Bode Plot . . . . . . . . . . . . . . . . . . . . 248
9.1.2 The Nyquist Plot . . . . . . . . . . . . . . . . . . . 250
9.2 Measures of Performance . . . . . . . . . . . . . . . . . . . 254
9.2.1 Relative Stability . . . . . . . . . . . . . . . . . . . 254
9.2.2 Phase Margin and the Transient Response . . . . . . 256
9.2.3 Error Constants and System Type . . . . . . . . . . 259
9.2.4 System Sensitivity . . . . . . . . . . . . . . . . . . 260
9.3 Frequency Response Design . . . . . . . . . . . . . . . . . 261
9.3.1 Gain Compensation . . . . . . . . . . . . . . . . . . 261
9.3.2 Phase-Lag Compensation . . . . . . . . . . . . . . . 262
9.3.3 Phase-Lead Compensation . . . . . . . . . . . . . . 264
9.3.4 Lead-Lag Compensation . . . . . . . . . . . . . . . 267
9.3.5 PI Compensator . . . . . . . . . . . . . . . . . . . . 269
9.3.6 PD Compensator . . . . . . . . . . . . . . . . . . . 271
9.3.7 PID Compensator . . . . . . . . . . . . . . . . . . . 273
9.3.8 Compensator Designs Compared . . . . . . . . . . . 275
9.4 Closed-Loop Frequency Response . . . . . . . . . . . . . . 276
Appendix 281
Index 285
About the Author 289

Foreword
Dr. Kamran Iqbal has assembled a very valuable and comprehensive book on
introductory and applied control. The book contains many tools and much
knowledge about the details of mankind’s and engineers’ everyday imple-
ments, and their control. The concepts, issues, challenges and alternatives are
lucidly and precisely presented.
The material is very well motivated and compactly and pragmatically
presented. Matlab programs and computer tools supplant the discussions and
analytical tools in order to ease the derivations and the conclusions.
A curious and motivated engineer or student of control will have no diffi-
culty following the material, the mathematical developments and derivations,
models of physical systems and interdisciplinary issues. Many examples and
exercises are added to enrich the imagination and creativity of the reader.
Subjects of Transfer Function, State Variables, Root Locus apparatus, Sample
Date systems, Pole Assignment are precisely described. A variety of assigned
problems and questions help the reader to become completely aware of
alternatives and be able to envision different solutions in his or her mind.
The book is equipped with ample explanations, guidance, tools and
encouragement. It will be a valuable tool in all countries of the world with
different educational, industrial and technical facilities and where engineers
are called upon to provide.
Hooshang Hemami
Professor emeritus of electrical and computer engineering
at Ohio State University
April 2020
xi
Preface
The aim of the second edition of A First Course in Control System Design,
similar to the first edition is to present model-based control system design
in a lucid, understandable and approachable manner. The book has been
written with the needs of undergraduates and beginning graduate students
in multiple engineering disciplines and practicing engineers in mind. The
second edition is organized into nine chapters; the first half of the book is
devoted to analysis and the second half to the design of control systems. The
book covers the design of controllers for analog and sampled-data systems
described by transfer function and state variable models. The coverage is
restricted to models of single-input single-output (SISO) systems. Examples
from diverse engineering disciplines are introduced in the first chapter and
carried forward in the later chapters. MATLAB and Control Systems Toolbox
are extensively used for design; occasionally, Symbolic Math Toolbox is also
used. MATLAB scripts for solutions to all book examples are provided.
Control systems, both natural and man made, are pervasive in our lives.
Our homes have environmental controls. The appliances we use at home, such
as the washing machine, microwave, etc. have embedded controllers. We fly
in airplanes and drive automobiles, which make extensive use of control sys-
tems. Our body regulates essential functions like blood pressure, heart-beat,
breathing, and insulin levels in blood, manifesting biological control systems.
The cells in the body regulate our metabolism and energy production using
nutrient levels and electrolytes. The postural stability of the body depends on
regulating body’s center of mass (CoM) over the base of support. Fine motor
control enables the manipulation and locomotion tasks we undertake as part
of daily living. We essentially perform the control function as we walk or
drive a car, the control objective in both cases being to follow a desired course
at a preferred speed.
The industrial revolution in the eighteenth century ushered in the age of
machines that needed automatic controls. As a result, ingenious solutions to
the control problems were developed. An early example involved the use of
centrifugal flyball governor for throttle adjustment to regulate the speed of
xiii
xiv Preface
Figure 1 A generic control system block diagram that includes the controller, process to be
controlled, the actuator and the sensor. The output is fed back and compared with the reference
signal in the comparator.
the steam engine that was essential to industrial progress. Though control
technology quickly developed to solve practical problems, the theoretical
understanding of the control systems and its underlying design process was
developed later in mid-twentieth century. The post WWII era launched the
space age that focused on the optimal design of control systems, and their
implementation via the computing machines. The quest for boosting the
industrial output through factory automation has enabled advancement in the
industrial process control, and in the industrial robots that makes extensive
use of feedback control systems. The growing automation in the past few
decades has increased our reliance on control systems.
A control system aims at realizing a desired behavior at the output of a
device or system (the plant) by manipulating its input through a controller.
Feedback based on observation of the process output via sensing elements
plays and important role in automatic control systems (Figure 1). In the
feedback control systems, the controller monitors the difference between the
desired and actual values of output variables, and adjusts the system inputs
accordingly by employing various control schemes. The control objective,
often, is to reduce the error to zero at a sufficiently fast rate and maintain it
there. The desired output may be expressed as a set point, that is, a constant
value that the controller will try to maintain at the output. Alternatively, in
tracking systems, the objective is to track a time-varying reference input. An
example of the latter is the control system used to make a drone-mounted
camera follow a moving object of interest.
The control system design is invariably undertaken to achieve multiple
objectives. The first and the foremost among them is the stability of the
closed-loop system, as the system outputs affect the inputs in real-time.
The next objective is the dynamic stability or the ability of the controller
to damp out the output oscillations, characterized by the damping ratio of
the dominant response modes. Further, the controller aims to improve the
speed of response, that is reflected by the system bandwidth. The steady-state
response of the closed-loop system, ideally, has a unity transfer function,
Preface xv
i.e., the system operates with no steady-state errors. Next, the controller is
required to curtail the effect of disturbance and noise inputs on system output.
A final objective in the control design is to impart robustness, which implies
an ability to maintain performance levels in the presence of disturbance
inputs, as well as its ability to withstand parameter variations and certain
unmodeled dynamics.
The controller designed for stability and performance may be of static
or dynamic type. In certain cases, a static gain controller may be adequate
to achieve a desired level of performance. An example is the automobile in
cruise control, where the gas intake is adjusted to affect the selected speed.
As the performance demands increase, a simple gain control is no longer ade-
quate and a dynamic controller becomes necessary. A dynamic controller is
a dynamic system in its own right, which generates a time varying controller
output that translates into the plant input. For example, the variation in the gas
pedal while driving an automobile in cruise control in response to the climb
or descent condition represents a time-varying controller output. An alternate
understanding of the dynamic controller is a frequency-selective filter that
emphasizes certain frequency bands in preference to others.
The dynamic controllers are traditionally distinguished as of phase-lead
or phase-lag type. These provide, respectively, improvements to the transient
or the steady-state response of the system. The two designs can be combined
when needed. In the contemporary design methods, the controller combines
one or more of the three basic control modes: a proportional (P), an integral
(I), and a derivative (D) mode. The resulting proportional-integral-derivative
(PID) controller is a general-purpose controller that has the ability to meet
many of the control objectives as defined above. The PID controller is
robust against variations in plant parameters, and is popular in industrial
control systems. In the traditional design, the controller is implemented using
analog circuits build with operational amplifiers and resistive-capacitive (RC)
networks. In contemporary control systems, the controller is implemented
as a software routine on a computer, a microcontroller, a DSP chip, or a
programmable logic controller (PLC).
A digital controller appropriate for computer implementation may be
obtained via emulation of an existing analog controller design. At high
enough sampling rates, digital approximation of the analog controller pro-
vides comparable performance to the original analog controller. Alternatively,
the design of the digital controller can be based on the pulse transfer function
of the plant, i.e., a transfer function obtained via z-transform that is valid at
the sampling intervals. Computer implementation of a controller invariably
xvi Preface
Figure 2 Block diagram of a digital control system that additionally includes an analog-to-
digital converter (ADC) and a digital-to-analog converter (DAC).
adds phase lag to the feedback loop that compromises the stability margins.
Hence, a more conservative controller design may be necessary if digital
implementation of the controller is intended. This book addresses the con-
troller design and their implementation for the analog systems as well as the
sampled-data systems (Figure 2).
This book covers the control system design as applicable to single-input
single-output (SISO) systems. The emphasis in this book is on understanding
and applying the techniques that enable the design of effective control sys-
tems. The controller design is based on the mathematical model of the plant
(the device or process to be controlled). System models are described in the
frequency-domain using the transfer functions, or in the time domain using
ordinary differential equations (ODEs). The state variable models describe
the system in terms of time derivatives of a set of state variables. Control
system design can be performed in either time or frequency-domain; essential
design techniques for both are covered in this book. A limited number of skill
assessment exercises are provided at the end of each chapter. Additional exer-
cises can be found in standard control systems textbooks (listed as references
at the end of the book).
The control systems concepts covered in this book are applicable to
the various engineering disciplines. These concepts are typically covered at
junior or senior level in the engineering curriculum. State variable models
are covered in more depth in a beginning graduate course. Students in
the scientific disciplines can also benefit from the control systems design
concepts. A typical audience of this book includes inquisitive readers with
interest in science, technology, engineering, and mathematics (STEM) fields.
The mathematical background required for understanding, and hence bene-
fitting from the material in this book, includes knowledge of linear algebra,
complex numbers, and elementary differential equations. Additionally, some
familiarity with Laplace and z-transforms is desired; the transform methods
are reviewed in the Appendix.
Preface xvii
The organization of this monograph is as follows: Chapter 1 discusses

the modeling of physical systems. The dynamic character of such a system is
typically captured using ODE models. Application of the Laplace transform
converts the ODEs to algebraic equations in the Laplace transform variable
‘s’, that are manipulated to obtain the input-output system description in
the form of a transfer function (TF). The examples in this chapter include
electrical, mechanical, electromechanical, thermal, and fluid systems. The
chapter culminates with a discussion about linearizing the nonlinear dynamic
system models.
Chapter 2 addresses the methods used to analyze the transfer function
models. These models are characterized in terms of their poles and zeros.
The poles effectively determine the modes of system natural response. The
stability characterization requires the poles to be located in the open left
half of the complex plane (OLHP). The system response to arbitrary inputs
comprises natural and forced response components. The natural response of
passive systems is of transitory nature and generally dies out with time. The
forced response signifies the presence of a persistent input and is observable
in the steady-state. System frequency response characterizes its response to
sinusoidal inputs, that manifests as a sinusoid at the input frequency.
Chapter 3 addresses the methods of analysis for the state-variable models.
State variables are often the natural variables, like inductor current and capac-
itor voltage in the electrical systems, or position and velocity of the inertial
mass in the mechanical systems. The state variable description typically
includes a set of matrices, that is, the system, the input, and the output
matrices. A solution to the first-order state equations includes a convolution
integral involving the state-transition matrix of the system. The choice of
the state-variables for a given system is not unique, giving rise to several
equivalent system descriptions, some of which may be preferred over others.
The popular descriptions are the controller form, the observer form, and the
modal form descriptions.
Chapter 4 introduces controller structures used with transfer function
models. The static controller includes a scalar gain that multiplies the error
signal generated by a comparator. The dynamic controllers include phase-
lead, phase-lag and lead-lag types. Alternate description of the dynamic
controllers includes the PD, PI, and PID controllers. Rate feedback con-
trollers additionally make use of the rate signal and are similar to PD and
PID controllers.
Chapter 5 discusses the control system design objectives. These include
closed-loop stability, transient response improvement, steady-state error
xviii Preface
improvement, and improvement to the sensitivity robustness. The chapter

discusses ways to characterize and achieve these objectives.
Chapter 6 discusses the root locus technique for designing cascade com-
pensators for transfer function models. The root locus (RL) is the loci of roots
of the closed-loop characteristic polynomial with variation in the controller
gain. The RL technique primarily addresses the design of static controllers,
but is easily extended to the design of first-order dynamic controllers. The
chapter includes a discussion of the rate feedback design that may be pre-
ferred over cascade controller design. The chapter ends with a discussion
of the controller realization methods with analog circuits using operational
amplifiers and resister-capacitor networks.
Chapter 7 discusses the techniques to analyze and design sampled-data
systems, that is, systems that include a clock-driven device, such as a micro-
processor in the loop. The sampled-data systems are characterized by their
pulse transfer function, that is obtained with the application of z-transform,
and describes system input-output behavior at the sampling intervals. The
time-domain description of sampled-data systems involves difference equa-
tions that can be solved by iteration. The stability of the sampled-data control
system requires the poles of the closed-loop pulse characteristic polynomial
to be located inside the unit circle in the complex z-plane. For high enough
sampling rates, an approximate digital controller for sampled-data system can
be obtained by emulating the analog controller designed for the continuous-
time system. Alternatively, the pulse transfer function can be used to design
a cascade digital controller by using the root locus technique.
Chapter 8 discusses the controller design for state variable models of
analog and sampled-data systems. The state feedback design allows arbitrary
placement of the roots of the closed-loop characteristic polynomial. The
design is facilitated by first transforming the state variable model into the
controller form. The pole placement method extends to the design of tracking
systems that require placing an integrator in the feedback loop to achieve
zero steady-state error. A continuous-time state variable model is converted
to discrete-time by assuming a zero-order-hold at the input. The discrete-time
state equations can be solved by iteration. A digital controller for the discrete
state variable model can be similarly designed using the pole placement
technique. Placing all poles of the closed-loop pulse transfer function at the
origin results in a deadbeat design that guarantees the sampled-data system
response to settle in n time periods.
Chapter 9 discusses compensator design by frequency response modifica-
tion. Frequency-domain methods, which predate the time-domain methods,
Preface xix
utilize gain and phase margins to characterize the relative stability of the
closed-loop system. The frequency-domain performance measures comple-
ment those used in time-domain analysis. A phase-lag compensator improves
the DC gain and/or the phase margin of the system, resulting in steady-state
or transient response improvement. A phase-lead compensator improves the
system bandwidth hence its transient response. The two can be combined
when both transient and steady-state response improvements are desired. The
closed-loop frequency response can be visualized on the Nichol’s Nyquist
plot or the chart and reveals the presence of resonance in the system.
Throughout this book, symbols in regular font represent scalar variables,
symbols in boldface letters represent arrays of variables, lower case letters
represent vectors, and upper case letters represent matrices. Control systems
designs presented in the book were performed in MATLAB (Mathworks,
Inc.) using the Control Systems Toolbox; additionally Symbolic Math
Toolbox and Simulink we used when needed. The MATLAB script used
for system design and simulation is provided with the examples. Web links
to solutions to the end-of-chapter problems are also provided. The figures
representing control systems were mostly drawn in the Simulink GUI. Circuit
models were drawn with TINA circuit simulator (Texas Instruments).
Acknowledgement
I would like to extend my gratitude to my teachers for instilling the love of

Control Systems in me.
I would like to thank my employer, University of Arkansas Little Rock,
department of Systems Engineering for facilitating my work.
I would like to thank the editor and the publisher for helping me with this
project.
I would like to thank my family and friends for their love and support.
I would like to thank my daughter Eeman for helping me draw the figures
in Chapter 1 and with editing the manuscript.
xxi
List of Figures
Figure 1.1 An RC circuit. . . . . . . . . . . . . . . . . . . . . 5

Figure 1.2 An RL circuit. . . . . . . . . . . . . . . . . . . . . 5
Figure 1.3 Motion of an inertial mass with surface friction. . . 6
Figure 1.4 Room heating with heatflow through walls. . . . . 7
Figure 1.5 Fluid reservoir with constricted outflow. . . . . . . 7
Figure 1.6 Output of a first-order system model to a unit-step
input. . . . . . . . . . . . . . . . . . . . . . . . . 10
Figure 1.7 A series RLC circuit. . . . . . . . . . . . . . . . . 11
Figure 1.8 Motion of an inertial mass under gravity. . . . . . . 11
Figure 1.9 A mass–spring–damper system. . . . . . . . . . . . 12
Figure 1.10 Time response of second-order system models:
characteristic equation with real roots (a); with
complex roots (b). . . . . . . . . . . . . . . . . . . 14
Figure 1.11 A bandpass RLC network. . . . . . . . . . . . . . 16
Figure 1.12 An armature-controlled DC motor. . . . . . . . . . 17
Figure 1.13 DC motor response to unit-step input: second-order
motor model (a); first-order motor model (b). . . . 19
Figure 1.14 A block diagram of the car cruise control system. . 26
Figure 2.1 Second-order transfer function pole locations in the
complex plane. . . . . . . . . . . . . . . . . . . . 34
Figure 2.2 Impulse response of a transfer function model:
(a) first-order system; (b) second-order system with
a pole at the origin; (c) second-order system with
real poles; and (d) second-order system with com-
plex poles. . . . . . . . . . . . . . . . . . . . . . . 39
Figure 2.3 Step response of second-order systems: (a) DC
motor model with real poles; (b) mass–spring–
damper system with complex poles. . . . . . . . . 42
Figure 2.4 Step response of an industrial process model with
deadtime. . . . . . . . . . . . . . . . . . . . . . . 43
Figure 2.5 System step response with quality metrics. . . . . . 47
xxiii
xxiv List of Figures
Figure 2.6 Sinusoidal response of DC motor. . . . . . . . . . 52

Figure 2.7 The Bode magnitude and phase plots for: (a) a first-
order system; (b) second-order system with a pole
at the origin. . . . . . . . . . . . . . . . . . . . . . 54
Figure 2.8 The Bode magnitude and phase plots for: (a) a
second-order system with real poles; (b) a second-
order system with complex poles. . . . . . . . . . . 55
Figure 2.9 Second-order system with complex poles: Bode
magnitude and phase plots for selected damping
ratios (a); accompanying unit-step response (b). . . 57
Figure 3.1 DC motor: state variable response to initial
conditions. . . . . . . . . . . . . . . . . . . . . . . 76
Figure 3.2 DC motor: impulse response (a) and step
response (b). . . . . . . . . . . . . . . . . . . . . . 77
Figure 3.3 Simulation diagram for realization of second-order
transfer function: serial realization (a); parallel
realization (b). . . . . . . . . . . . . . . . . . . . . 79
Figure 3.4 Simulation diagram for controller form realization
of the transfer function model. . . . . . . . . . . . 81
Figure 4.1 Feedback control system with static gain controller
(a); with dynamic controller K(s), plant G(s), and
sensor H(s) (b). . . . . . . . . . . . . . . . . . . . 94
Figure 4.2 Phase contribution from first-order controllers;
phase-lead (a); phase-lag (b). . . . . . . . . . . . . 98
Figure 4.3 Three basic control modes represented in the PID
controller. . . . . . . . . . . . . . . . . . . . . . . 99
Figure 4.4 Step response of the DC motor with cascade PID
controller. . . . . . . . . . . . . . . . . . . . . . . 103
Figure 4.5 The rate feedback controller: basic rate feedback
configuration (a); equivalent PD compensator in the
feedback loop (b). . . . . . . . . . . . . . . . . . . 103
Figure 4.6 The rate feedback with cascade PI controller: basic
controller configuration (a); equivalent PID com-
pensator in the feedback loop (b). . . . . . . . . . . 105
Figure 4.7 Step responde of the mass–spring–damper sys-
tem: rate feedback (a); rate feedback with cascade
PI (b). . . . . . . . . . . . . . . . . . . . . . . . . 106
List of Figures xxv
Figure 5.1 Feedback control system with plant G(s), sensor

H(s), and static gain controller represented
by K(s). . . . . . . . . . . . . . . . . . . . . . . . 112
Figure 5.2 Determination of relative stability margins on
the Bode plot of the loop transfer function;
1500
(a) KGH (s) = s2 +110s+1025 ; (b) KGH (s) =
1500
s(s2 +110s+1025)
. . . . . . . . . . . . . . . . . . . . 118
Figure 5.3 Step response of the closed-loop simplified DC
motor model (Example 5.6). . . . . . . . . . . . . 120
Figure 5.4 Pole locations for the desired characteristic polyno-
mial (Example 5.7). . . . . . . . . . . . . . . . . . 122
Figure 5.5 Steady-state error to a ramp input (Example 5.13). . 127
Figure 5.6 A feedback control system with reference and
disturbance inputs. . . . . . . . . . . . . . . . . . 128
Figure 6.1 Feedback control system with plant G(s), sensor
H(s), and a static gain controller represented
by K. . . . . . . . . . . . . . . . . . . . . . . . . 134
1
Figure 6.2 The root locus plot for G(s) = s+1 , H(s) = 1 (a);
1
the root locus plot for G(s) = s(s+2) ,
H(s) = 1 (b). . . . . . . . . . . . . . . . . . . . . 137
(s+3)
Figure 6.3 The root locus plot for G(s) = s(s+2) (a); the root
1
locus plot for G(s) = s(s+1)(s+2) (b). . . . . . . . 140
Figure 6.4 The impulse response of the third-order system for
K = 6. . . . . . . . . . . . . . . . . . . . . . . . 141
Figure 6.5 The desired region for closed-loop pole placement
for 0.7 < ζ < 1 and ts < 2s. . . . . . . . . . . . . 143
Figure 6.6 Application of the angle criteria to choose the PD
controller zero at the desired root location s1 : θz −
θp1 − θp2 = ±180◦ . . . . . . . . . . . . . . . . . 147
Figure 6.7 The root locus design with: PD controller (a); phase-
lead controller (b). . . . . . . . . . . . . . . . . . . 148
Figure 6.8 RL improvement with phase-lag design (a); steady-
state velocity error constant improvement with
phase-lag design (b). . . . . . . . . . . . . . . . . 150
Figure 6.9 Root locus improvement with lead–lag compensa-
tion (a); RL improvement with PID
compensation (b). . . . . . . . . . . . . . . . . . . 153
xxvi List of Figures
Figure 6.10 Step response of phase-lead plus PI compensated

system (Example 6.10). . . . . . . . . . . . . . . . 155
Figure 6.11 Rate feedback design for Example 6.12: minor loop
design (a); outer loop design (b). . . . . . . . . . . 158
Figure 6.12 Rate feedback design for Example 6.13: minor loop
design (a); outer loop design (b). . . . . . . . . . . 160
Figure 6.13 A comparison of the step responses for the various
compensator designs. . . . . . . . . . . . . . . . . 162
Figure 6.14 Operational amplifier in inverting configuration. . . 163
Figure 7.1 A closed-loop sampled-data system model. . . . . . 168
Figure 7.2 Impulse response of the ZOH. . . . . . . . . . . . 171
Figure 7.3 The unit-pulse response of a sampled-data system:
First-order system (a); Second-order system (b). . . 179
Figure 7.4 The unit-step response of a sampled-data system:
Figure 7.5 Sinusoidal response of a second-order sampled-data
system preceded by a ZOH. . . . . . . . . . . . . . 183
Figure 7.6 The unit-step response of a sampled-data system:
Figure 7.7 Step response of a second-order sampled-data sys-
tem: T = 0.5s (a); T = 1s (b). . . . . . . . . . . . 191
Figure 7.8 Root locus design in the z-plane (a); step response
of the closed-loop system (b). . . . . . . . . . . . . 195
Figure 7.9 Root locus design of the DC motor model: analog
system design (a); sampled-data system
design (b). . . . . . . . . . . . . . . . . . . . . . . 197
Figure 7.10 Step response of the DC motor for the analog and
discrete controller design. . . . . . . . . . . . . . . 199
Figure 7.11 Step response of the DC motor for the analog and
discrete PID controller design. . . . . . . . . . . . 200
Figure 7.12 Step response comparison for controller emula-
tion methods: MATLAB simulation using the pulse
transfer function (a); Simulink simulation with the
analog plant (b). . . . . . . . . . . . . . . . . . . . 203
Figure 7.13 Comparison of the step response using analog and
discrete PID controllers. . . . . . . . . . . . . . . . 205
Figure 8.1 The step response of the DC motor with pole place-
ment controller: armature current (top); motor speed
(bottom). . . . . . . . . . . . . . . . . . . . . . . 215
List of Figures xxvii
Figure 8.2 The step respone of the DC motor with pole place-
ment controller and feedforward compensation:
armature current (top); motor speed (bottom). . . . 224
Figure 8.3 The block diagram for tracking PI control of a state
variable model. . . . . . . . . . . . . . . . . . . . 225
Figure 8.4 Step response of the mass–spring–damper system
with integrator in the loop. . . . . . . . . . . . . . 228
Figure 8.5 The root locus for the DC motor model with
integrator in the loop. . . . . . . . . . . . . . . . . 229
Figure 8.6 The step response of the DC motor with controller
emulation: armature current (top); motor speed
(bottom). . . . . . . . . . . . . . . . . . . . . . . 236
Figure 8.7 Unit-step response of the DC motor model: analog
system; controller emulation; digital controller. . . 238
Figure 8.8 The step response of a small DC motor model:
analog system; deadbeat controller. . . . . . . . . . 240
Figure 8.9 Tracking PI control of the DC motor model: analog
system design; discrete system design. . . . . . . . 242
1
Figure 9.1 Bode and Nyquist plots for G(s) = (s+1) . . . . . . 251
1
Figure 9.2 Bode and Nyquist plots for G(s) = s(s+1) . . . . . . 252
1
Figure 9.3 Bode and Nyquist plots for G(s) = s(s+1)(s+2) . . . 254
Figure 9.4 The relative stability from the Bode plot: a second-
order transfer function (a); a third-order transfer
function (b). . . . . . . . . . . . . . . . . . . . . . 257
Figure 9.5 The DC motor model with PI controller: Bode plot
of the loop transfer function (a); step response of the
compensated system (b). . . . . . . . . . . . . . . 259
Figure 9.6 Gain compensation for the desired phase margin
improvement: uncompensated system (a); compen-
sated system (b). . . . . . . . . . . . . . . . . . . . 263
Figure 9.7 Phase-lag compensation: magnitude compensated
plot (a); phase-lag compensated plot (b). . . . . . . 265
Figure 9.8 Phase-lead compensation: uncompensated plot (a);
phase-lead compensated (b). . . . . . . . . . . . . 268
Figure 9.9 Lag-lead compensation: phase lag compensated
system (a); lag-lead compensated system (b). . . . 270
Figure 9.10 PI compensation (a); PD compensation (b). . . . . 272
xxviii List of Figures
Figure 9.11 PID compensator design: PI compensation (a),

followed by PD compensation (b). . . . . . . . . . 274
Figure 9.12 A comparison of the step responses for the various
controller designs. . . . . . . . . . . . . . . . . . . 275
Figure 9.13 Closed-loop frequency response on the Nyquist plot
(a), and the Nichol’s chart (b). . . . . . . . . . . . 279
List of Tables
Table 2.1 Quality metrics used to measure second-order system

response . . . . . . . . . . . . . . . . . . . . . . . . 46
Table 6.1 A comparison of the unit step response of the
compensated system for six different compensator
1
designs for G(s) = s(s+2)(s+5) . . . . . . . . . . . 161
Table 9.1 A comparison of the unit step response of the
compensated system for six different controller
2
designs for G(s) = s(s+1)(s+2) . . . . . . . . . . . 276
xxix
List of Abbreviations
AC Alternating current
BIBO Bounded-input bounded-output
CL Closed-loop
CLCP Closed-loop characteristic polynomial
dB Decibel
DC Direct current
MIMO Multi-input multi-output
OL Open loop
OLHP Open left half-plane
PD Proportional-derivative
PFE Partial fraction expansion
PI Proportional-integral
PID Proportional-integral-derivative
RL Root locus
RLC Resistor, inductor, and capacitor
SISO Single-input single-output
ZOH Zero-order hold
xxxi
xxxii List of Abbreviations
Common Symbols used in the book

s Laplace transform variable
z Z transform variable
G(s), G(z) Plant transfer function
K(s), K(z) Controller/compensator transfer function
H(s) Sensor transfer function
L(s) Loop gain
T (s), T (z) Closed-loop transfer function
r(t), rk Reference input
u(t), uk Plant input
y(t), yk Plant output
e(t), ek Error signal
d(t) Disturbance input
K DC gain of the plant/controller
τ Time constant
ωn Natural frequency
ωd Damped natural frequency
ζ Damping ratio
x(t) Vector of state variables
x0 Vector of initial conditions
A, Ad System matrix in the state variable representation
b Input distributions vector
c Output contributions vector
k Feedback gains vector
eAt State transition matrix
1
Mathematical Models
of Physical Systems
Learning Objectives
1. Obtain a differential equation model of a low order dynamic system.
2. Obtain system transfer function from the differential equation descrip-
tion.
3. Obtain a physical system model in the state variable form.
4. Linearize a nonlinear dynamic system model about an operating point.
In this chapter we discuss the process of obtaining mathematical description
of a dynamic system, i.e., a system whose behavior changes over time. The
models of continuous-time systems are primarily described in terms of linear
or nonlinear differential equations.
Physical systems of interest to engineers include electrical, mechanical,
electromechanical, thermal, and fluid systems, among others. Using the
lumped parameter assumption, their behavior is mathematically described by
ordinary differential equation (ODE) models. These equations are, in general,
nonlinear, but can be linearized about an operating point for analysis and
design purposes.
To model physical systems with interconnected components, individual
component models can be assembled to obtain a system model. In the case
of electrical circuits and networks, these elements include resistors, capaci-
tors, and inductors. For mechanical systems, these include inertia (masses),
springs, and dampers (or friction elements). For thermal systems, these
include thermal capacitance and thermal resistance. For hydraulic and fluid
systems, these include reservoir capacity and flow resistance.
The individual components or devices that form a physical system can
store, exchange, or dissipate energy, which, gives rise to the time-varying
or dynamic behavior of the system. The dynamic behavior is captured by
the differential equation model. The natural variables associated with those
1
2 Mathematical Models of Physical Systems
elements, e.g., positions and velocities in mechanical systems, or currents

and voltages in electrical systems, form the set of state variables commonly
used to describe system behavior.
In physical systems, properties (or entities) may flow in and out of a
system boundary, for example, heat, mass, or volume flowing in or out of
a reservoir. To model such systems, conservation laws or balance equations
maybe used to describe system dynamics in terms of rate of change of an
accumulated property. Specifically, let Q represent the accumulated property,
and let qin and qout represent the inflow and outflow rates, then the relevant
dynamic equation is described as:
dQ
= qin − qout + g − c
dt
where g and c denote the internal generation and consumption.
The Laplace transform is commonly used to convert a set of linear
differential equations into algebraic equations that can then be manipulated
to obtain an input–output description, that is, a transfer function. The transfer
function that describes the system behavior is a rational function of a complex
frequency variable (s).
The transfer function forms the basis for analysis and design of control
systems in the frequency domain. In contrast, the modern control theory
is established on time-domain techniques involving the state equations that
describe system behavior as time derivatives of a set of state variables. These
are defined as any set of minimum number of variables that can capture
system behavior under the various operating conditions.
Linearization of nonlinear models is accomplished by using Taylor series
expansion about a critical point. The resulting linear model is only effective in
the neighborhood of the critical or stationary point. The linear systems theory
is well estabished and is relied upon for controller design, that is aimed to
modify the system behavior.
In this chapter we discuss simples models of electrical, mechanical,
electromechanical, thermal, and fluid systems that form the basis for system
ananlysis and design techniques that are covered in the later chapters.
1.1 Modeling of Physical Systems

The mathematical modeling of a physical system is enabled by the choice
of variables associated with the physical characteristics of its components.
1.1 Modeling of Physical Systems 3
These variables naturally divide into flow and across variables. The relation-
ship between these two variables determines the element type, as discussed
below.
1.1.1 Model Variables and Element Types

Modeling of a physical system involves two kinds of variables: flow variables
that “flow” through the system components and across variables that are
measured across those components. In electrical circuits, voltage or potential
is measured across the circuit nodes, whereas current or electrical charge
flows through the circuit branches.
In mechanical linkage systems, displacement and velocity are measured
across the connecting nodes, whereas force or effort “flows” through the
linkages. In the case of thermal and fluid systems, heat and mass serve as the
flow variables, while temperature and pressure constitute the across variables.
The relationship between flow and across variables associated with an
element in the system defines the type of physical element being modeled.
The three basic types are the resistive, inductive, and the capacitive elements.
The terminology that is specific to electrical circuits extends to many other
types of physical systems.
To proceed further, let q(t) denote a flow variable and x(t) denote an
across variable associated with a physical element; then, the element type is
defined by their mutual relationships, as described hereunder.
The resistive element. The resistive element is described by the relation
x(t) = k q(t)
i.e., the flow and across variables for the element vary in proportion to each
other.
As an example, the voltage and current relationship through a resistor is
described by Ohm’s law, V (t) = Ri(t), that states that the voltage across a
resistor varies in proportion to the current flowing through the resistor.
Similarly, the force–velocity relationship through a mechanical damper is
described as v(t) = 1b f (t) , i.e., the velocity increases in proportion to the
applied force.
The capacitive element. The capacitive element is described by the relation
Z
x(t) = k q(t)dt + x0
i.e., the across variable varies in proportion to the accumulated amount of the
flow variable. Alternatively, the flow variable varies proportionally with the
rate of change of the across variable as q(t) = k1 dx(t)
dt .
As an example, the voltage and current relationship through a capacitor
is given as V (t) = C1 i(t)dt + V0 , i.e., the voltage across the capaci-
R
tor is proportional to the integral of current through it, where the current
integral represents the accumulation of electrical charge; hence, Q = CV .
The inverse relationship is described as i(t) = C dV dt .
Similarly, the force–velocityRrelationship that governs an inertial mass
1
element is given as v(t) = m f (t)dt + v0 . The inverse relation is the
familiar Newton’s second law of motion, f (t) = m dv(t) dt = ma(t), where
a(t) denotes acceleration.
The inductive element. The inductive element is described by the relation
dq(t)
x(t) = k
dt
i.e., the flow variable is obtained by differentiating the across variable. Alter-
natively, the across variable varies in proportion to the accumulation of the
flow variable.
As an example, the voltage–current relationship through an inductive coil
in an electric circuit is Rgiven as V (t) = L di(t)
dt . The inverse relationship is
1
described as i(t) = L V (t)dt + i0 . Similarly, the force–velocity relation-
1 df (t)
ship through R a linear spring is given as v(t) = K dt . The inverse relation is
f (t) = K v(t)dt + f0 .
While the resistive element dissipates energy, both the capacitive and
inductive elements store energy in some form. For example, a capacitor
stores electrical energy and a moving mass stores kinetic energy. The energy
storage accords memory to the element that accounts for the dynamic system
behavior modeled by an ODE.
Next, we discuss system models obtained from the individual component
descriptions. These models are described in increasing order of complexity,
starting from the most basic ones.
1.1.2 First-Order ODE Models

Electrical, mechanical, thermal, and fluid systems that contain a single energy
storage element are described by the first-order ODE models, where the
mathematical model is conveniently described in terms of the output of the
energy storage element. This is illustrated in the following examples.
Figure 1.1 An RC circuit.
Figure 1.2 An RL circuit.
Example 1.1: A series RC network.

We consider a series RC network connected across a constant voltage
source, Vs (Figure 1.1). Kirchhoff’s voltage law (KVL) is used to model
the circuit behavior as vR + vC = Vs , where the capital letters are used to
represent constant values and small letters represent time-varying quantities.
By substituting vC = v0 , the ciruit output, and vR = iR = RC dv 0
dt , we
obtain the first-order ODE model that describes the circuit behavior as
dv0 (t)
RC + v0 (t) = Vs .
dt
Example 1.2: A parallel RL network.

We similarly consider a parallel RL network connected across a constant
current source, Is (Figure 1.2). The circuit is modeled by the first-order ODE,
where the variable of interest is the inductor current, iL , and Kirchhoff’s
current law (KCL) is applied at either of the nodes, to obtain iR + iL = Is .
L diL
Then, by substituting iR = Rv = R dt we obtain the ODE dscription of
the RL circuit as
L diL (t)
+ iL (t) = Is
R dt
We may note that the constant multiplier appearing with the derivative
term in both RC and RL circuits defines the time constant of the circuit, i.e.,
the time when the system output in response to a constant input rises to 63.2%
of its final value. The time constant is denoted by τ and is measured in [sec].
In particular, τ = RC for the RC circuit, and τ = L/R for the RL circuit.
( )
( )
( )
Figure 1.3 Motion of an inertial mass with surface friction.
Example 1.3: Inertial mass acted upon by a force.

The motion of an inertial mass, m, acted by a force, f (t), in the presence
of kinetic friction represented by b is governed by Newton’s second law of
motion. The friction opposing motion is represented as bv, where v is the
velocity variable and b is the friction constant. The resultant force on the
mass element is f − bv. The resulting first-order ODE model for the system
is given as
dv(t)
m + bv(t) = f (t).
dt
The time constant for the mechanical model is τ = m b , and desribes the
rate at which the velocity builds up in response to a constant force input.
A Generic first-order ODE model. The first-order ODE models in the above
examples can be generalized. Accordingly, let u(t) denote a generic input,
y(t) denote a generic output, and τ denote a time constant; then, the generic
first-order ODE model is expressed as
dy(t)
τ + y(t) = u(t).
dt
Further examples of the first-order models include the thermal and fluid
systems. These systems maybe modeled using balance equations applied to a
control volume as illustrated hereunder.
Example 1.4: A model for room heating.
In order to model the room heating process, we assume that the heat flow
into the room is denoted as qi , the thermal capacity of the room is Cr , the
temperature of the room is θr , the ambient temperature is θa , and the wall
insulation is represented by a thermal resistance, Rw . Then, from the heat
energy balance, we can write
dθr θ r − θa
Cr + = qi .
dt Rw
Figure 1.4 Room heating with heatflow through walls.
Figure 1.5 Fluid reservoir with constricted outflow.
In terms of the temperature differential, ∆θ = θr − θa , the govering

differential equation is
d∆θ
Rw Cr + ∆θ = Rw qi .
dt
The temperature is measured in [◦ C], heat flow is measured in [W ],
thermal capacitance is measured in [ ◦JC ], and thermal resistance is measured
◦
in [ WC ].
We may note the similarity of the room heating model with the general
first-order system model, τ dy
dt + y = u, where the thermal time constant is
given by τ = Rw Cr .
Example 1.5: A model of hydraulic reservoir.
We consider a cylindrical reservoir filled with an incompressible fluid
with a controlled exit at the bottom (Figure 1.5).
To proceed further, let P denote the hydraulic pressure, Patm denote
the atmospheric pressure, A denote the area of the reservoir, h denote the
height, V denote the volume, ρ denote the mass density, Rl denote the valve
resistance to the fluid flow; qin , qout denote the volumetric flow rates, and g
denote the gravitational constant. Then, the base pressure in the reservoir is
obtained as
ρg
P = Patm + ρgh = Patm + V.
A
Using reservoir capacitance, defined as: Ch = dV A
dP = ρg , the governing
equation of the hydraulic flow through the reservoir is given as
dP P − Patm
Ch = qin −
dt Rl
In terms of the pressure difference, the equation is written as
d∆P
Rl Ch + ∆P = Rl qin .
dt
Then, the above equation matches the standard first-order system model
A
with τ = Rl Ch . Further, using ∆P = ρgh and Ch = ρg , we can equivalently
express the governing equation in terms of the liquid height, h(t), in the
reservoir as
dh
ARl + ρgh = Rl qin
dt
In the above, the hydraulic pressure is measured in [ mN2 ], volumetric flow
3 5
is measured in [ ms ], hydraulic capacitance is measured in [ mN ], and flow
Ns
resistance is measured in [ m5 ].
1.1.3 Solving First-Order ODE Models with Step Input

The solution to a first-order ODE in the presence of a step forcing function
and with given initial conditions can be obtained with the help of the Laplace
transform (see Appendix). Accordingly, we consider the generic first-order
ODE model,
dy(t)
τ + y(t) = u(t).
dt
We apply Laplace transform to the above equation assuming an initial
condition, y(0) = y0 , to obtain:
τ (sy(s) − y0 ) + y(s) = u(s)
1
Next, assuming a unit step input u(t), where u(s) = s, the output is
solved as:
1 τ y0
y(s) = +
s(τ s + 1) τ s + 1
We may use partial fraction expansion (PFE) to express the output as:
1 τ τ y0
y(s) = − +
s τs + 1 τs + 1
After applying the inverse Laplace transform, we obtain the time-domain
solution to the first-order ODE as:
y(t) = [1 + (y0 − 1)e−t/τ ]u(t)
The u(t) in the above expression represents a unit step function that is
used to represent causality, i.e., the output is valid for t ≥ 0.
Steady-state output. The steady-state value of the system response is
denoted as y∞ = limt→∞ y(t). In terms of the steady-state output, the step
response of the first-order ODE model is given as:
y(t) = [y∞ + (y0 − y∞ )e−t/τ ]u(t)
Assuming zero initial conditions, i.e., y0 = 0, the output is expressed as:
y(t) = (1 − e−t/τ )u(t)
We may evaluate the output at t = kτ, k = 0, 1, . . . to compile the

following table:
Time Output Value

0 y(0) = 0
1τ 1 − e−1 ∼
= 0.632
2τ 1 − e−2 ∼
= 0.865
3τ 1 − e−3 ∼
= 0.950
4τ 1 − e−4 ∼
= 0.982
5τ 1 − e−5 ∼
= 0.993
By convention, the model output is assumed to have reached the steady

state when the output attains 98% of its final value. Hence, the settling time
of the system is given as ts = 4τ .
The output of the first-order ODE model to a unit-step input with zero
initial conditions for τ = 1 sec is plotted in Figure 1.6.
Figure 1.6 Output of a first-order system model to a unit-step input.
Further, the time variable in any first-order ODE model can be scaled
by dividing it with the time constant. The response of a first-order ODE
with nondimensionalized time variable (t/τ ) is similar to that represented
in Figure 1.6.
1.1.4 Second-Order ODE Models

A physical system that contains two energy storage elements is described
by a second-order system model, where each energy-storing component
contibutes a first-order term to the model. Examples of the second-order
systems include RLC networks and the motion of an inertial mass with
position output.
The modeling of a second-order system is illustrated by the following
examples.
Example 1.6: Series RLC circuit.
A series RLC circuit with voltage input Vs (t) and current output i(t) has
the following governing relationship obtained by applying KVL to the mesh:
Z
di(t) 1
L + Ri(t) + i(t)dt = Vs (t)
dt C
The integro-differential equation can by converted into a second-order
ODE by expressing it in terms of the electric charge, q(t), as
d2 q(t) dq(t) 1
L 2
+R + q(t) = Vs (t)
dt dt C
Alternatively, the series RLC circuit behavior can be described in terms
of the two first-order ODEs involving dual variables, the current, i(t), and the
Figure 1.7 A series RLC circuit.
( )
( )
Figure 1.8 Motion of an inertial mass under gravity.
capacitor voltage, Vc (t), as
di(t)
L + Ri(t) + Vc (t) = Vs (t),
dt
dVc
C = i(t)
dt
The above formulation, where i(t) and Vc (t) serve as variables of inter-
est, is known as the state variable formulation and is further discussed in
Section 1.2.
Example 1.7: Inertial mass with position output.
An inertial mass moving in a constant gravitational field has both kinetic
and potential energies and is modeled by a second-order ODE. For example,
the vertical motion of a mass element of weight, mg, that is pulled upward
by a force, f (t), is described using position output, y(t), by a second-order
ODE:
d2 y(t)
m + mg = f (t).
dt2
Figure 1.9 A mass–spring–damper system.
Example 1.8: A mass–spring–damper system.

A mass–spring–damper system includes a mass affected by an applied
force, f (t), when its motion is restrained by a combination of a spring and
a damper (Figure 1.9). Let x(t) denote the displacement of the mass from a
fixed reference; then, the dynamic equation of the system obtained by using
Newton’s second law of motion takes a familiar form, given as:
d2 x(t) dx(t)
m 2
+b + kx(t) = f (t).
dt dt
The left-hand side in the above equation represents the sum of applied
(inertial, damping, and spring) forces. In compact notation, we may express
the ODE as:
mẍ + bẋ + kx = f
dx(t)
where the dots above the variable represent time derivative, i.e., ẋ(t) = dt
d2 x(t)
and ẍ(t) = dt2 .
In the absence of damping, the dynamic equation of the mass-spring
system reduces to
d2 x(t)
m +kx(t)=f (t).
dt2
We may recognize that this equation models simple harmonic motion. Let
2
ω0 = k/m; then, it can be verified by substitution that the general solution to
the equation is given as:
x(t) = A cos ω0 t + B sin ω0 t.
1.1.5 Solving Second-Order ODE Models

A second-order ODE model can be similarly solved by applying the Laplace
transform to both sides of the differential equation. We consider a general
Another random document with
no related content on Scribd:
“I took him home and fixed him up. It turned out that he was the
leader of a notorious gang. I’ve never known a finer chap than he.
I’ve found out in this life that if you scratch deep enough you’ll
always find true worth.
“About a week after that night he came around and took me to a
notorious saloon. He took me into the back room and introduced me
to the bunch. Several of them have gone to the chair since, but they
were good fellows.[12] I’ve gone into that saloon without a nickel in
my pocket, and looking it. I’ve had one of the gang say to me: ‘Stony
up against it, kid? Will a fiver help?’ and before I could know what
happened the gang would have taken up a collection of $25 and
given it to me.[13]
“My wife was living with her family at that time, but often she would
come to bring me baskets with chicken and all sorts of delightful little
delicacies. The neighborhood was a terrible one in those days, and I
was afraid at first to have her come there. I told some of the boys
about it. They told me never to worry again. They arranged that an
unseen bodyguard should follow her from the street car and escort
her to my room and back again when she was ready to leave.... She
believed in me even then when it meant more to me than anything in
the world. People don’t know it, but I am naturally a timid man. She
gave me confidence in myself, and with it came the ability to
succeed.”
The room at Thirty-eighth Street and Tenth Avenue was a studio,
“dismal, rat-haunted,” where a job as a painter of theatrical scenery
compelled him to spend a great many of his nights and days. In
intervals of scene-painting he began The Broad Highway. “I met O.
Henry several times in the offices of Ainslee’s Magazine. I think it
was Will Irwin who introduced us. O. Henry was unusually taciturn
for an American, and I—well, I am an Englishman. So though we
saw each other frequently, never more than ‘How d’ ye do’ passed
between us.
“The pleasantest recollections I have of those old days was the
time I spent in dabbling in painting and theatricals at the old Astor
Theatre. One day a down-and-out young man got past the
doorkeeper and strolled on the stage. ‘I’ve got a fortune here in my
pocket,’ he said. ‘We all have that,’ I replied.
“The young fellow said he had been a cub reporter in Chicago, but
now he was hungry and looking for a job. Finally he got the attention
of the producer at the theatre. He pulled out a manuscript and began
reading. The producer at first paid no attention, but gradually
became more and more interested. When the first act had been read
the producer said, ‘All right, I’ll take it.’ The starving dramatist was
Eugene Walter and the manuscript was that of ‘Paid in Full.’”[14]
Farnol wrote in the studio and also at Mr. Hawley’s home, in
Englewood, New Jersey. When The Broad Highway was the best
seller, Mr. Hawley rounded out the picture of the New York period.[15]
“Farnol,” he said, “is a dreamer and a bookworm, and has just about
as much practical idea of time and money as that type is popularly
supposed to have. He kept right on writing, and night was the time
he had to do it. Many a time when I’ve been detained late with a
press of work I’d get home at midnight or thereabouts to find a light
supper waiting for me and Jeffery up working, only waiting to be
called to entertain me while I ate. For he is the most entertaining
talker I’ve ever known and loves to talk. His natural speech is the
phraseology in The Broad Highway.[16] It has become natural to him
through many years of living with the characters in the books of that
period he loves so well. And he is a born storyteller. He always kept
us sitting overtime at meals, just as he used to keep me sitting up till
2 or 3 o’clock in the morning on the occasion of the midnight
suppers. When he gets to spinning a yarn, whether telling it or
writing it, he loses all knowledge of the flight of time. Often when I’ve
come down to breakfast before catching my early train to New York
I’ve found him just finishing his night’s work, fresh and enthusiastic.
Even when his days are full of leisure he likes best to work at night.
Then, he says, his brain is clearer, and there are no interruptions.
His power of concentration and absorption is the most marvellous
thing I’ve ever encountered. I remember once taking him to the
Players Club with me for luncheon. After luncheon he wandered into
the library and was delighted to see the work of Aphra Behn, an
early writer I’d never heard of, but belonging to his favorite period
and well known to him. I left him there renewing her acquaintance
with delight. I forgot all about him, but chancing to go back for dinner,
on entering the library to my amazement I saw him sitting there in
exactly the same posture in which I’d left him hours before. He didn’t
know whether ten minutes or as many hours had elapsed.”
Farnol succeeded in selling a number of short stories and had
some work as an illustrator. He wrote two light romances, My Lady
Caprice and The Money Moon, which magazines bought. For two
years he put all his spare time on The Broad Highway, the history of
which is among the curiosities of book publishing.
iv
Like Main Street a decade later, The Broad Highway was possibly
conceived and certainly executed in a spirit of revolt. Such rebellions
are common, and the only wisdom that can be uttered in respect of
them is embodied in that proverb which says that one man’s meat is
another man’s poison. Editors and publishers endeavor to give the
public what the public wants. The public, very naturally, never knows
what it wants until it tastes it. The public is like a husband sitting
down to his wife’s dinner. He may like everything or nothing; he may
enormously relish the unexpected, placed before him with inward
perturbation and in a spirit of desperate doubt. He may pounce with
appetite upon, and sing loudly the praise of, some dish denounced
by him and refused by his palate the week before. If a writer attempt
to please editors or publishers, who, in turn, are attempting to please
their publics, he will be successful with the entrepreneurs and
possibly with the audiences. And his temperament may make such a
course the very best thing. If his temperament is otherwise, sooner
or later he will please himself; and if he can then get published a
large public may just possibly discover that he greatly pleases them.
The completed manuscript of The Broad Highway was submitted
to the Century Company and Charles Scribner’s Sons in New York,
both of whom failed to come to terms and returned it shortly. It was
then submitted to Dodd, Mead & Company, who indicated a
conditional acceptance and asked the author to come in and discuss
possible changes. The firm’s readers offered their suggestions and
Farnol took notes. The principal result was that he cut 20,000 words
out of the book, which still remained of 200,000-word length, or twice
the length of the usual “full-length” novel. The alterations were not
enough to give the publisher the necessary confidence; the year was
1907, the year of the money panic; and the manuscript was finally
returned to Farnol with a definite declination. The reasons were
sound: There had been a bad slump in Wall Street, the book was
formidably long, the author was unknown, the interest of the tale
might be almost wholly for English readers. But there was another
reason in the nature of the novel to which a few words should be
devoted.
Robert Louis Stevenson had died in 1894. His work spawned a
school of historical fiction, much of it pseudo-historical, which had
dominated the American book market for years. The public taste did
not discriminate during that decade between the good and the bad;
To Have and To Hold and When Knighthood Was in Flower were
equally hailed as masterpieces and alike elevated to the top of the
heap. From that day, indeed, dates the name and the peculiarity of
the “best seller.” The term remains, but it has only in very recent
years begun to undergo a transformation of meaning, the idea of
relativity having crept in. With a truer perception and a better sense
of proportion, we now tend to speak of a book as a best seller in its
class, or in relation to the literary merit of the work or the record of
the author or generally with an eye to what sale could be expected in
the circumstances. The fact of a sale in so many figures remains; but
the estimation in which the fact is held is quite different. A sale of
20,000 copies that would have passed unregarded twenty years ago
is now likely to be accounted as of the greatest significance.
What was fundamentally the trouble in 1907 was not to be stated
with vigor until 1914, when Frank Swinnerton’s critique of
Stevenson[17] was to appear with such concluding sentences as
these: “Stevenson ... created a school which has brought romance to
be the sweepings of an old costume-chest.... If romance rests upon
no better base than this, if romance is to be conventional in a double
sense, if it spring not from a personal vision of life, but is only a
tedious virtuosity, a pretence, a conscious toy, romance as an art is
dead. And if it is dead, Stevenson killed it.” Such, even in 1907, was
in various quarters uneasily felt to be the fact. In 1907, it is true,
George Meredith was spending his declining years in poetry, and
Thomas Hardy was at work on The Dynast; but The Way of All Flesh
had been published four years earlier, Shaw’s plays were being
staged, the dead George Gissing was at last coming into attention,
Mr. Galsworthy had just given us The Man of Property, Mr. Wells
was brewing Tono-Bungay, and Mr. Bennett was at work on The Old
Wives’ Tale. If the lid of the costume-chest was still raised, it had
every appearance of being propped most insecurely. All cogent and
immediate reasons aside, the publisher of books had every
psychological and intuitive reason for doubting the appeal of a
volume of 500 closely printed pages, much of it in dialect and all of it
concerned with Kentish scenes of a hundred years earlier.
To return to The Broad Highway: An actor with whom the author
had become acquainted at the Astor Theatre was about to play an
engagement in Boston, and offered to show the manuscript to
friends in the office of Little, Brown and Company. Farnol waited for
some word in vain; after several months he learned that the actor
had returned to New York, and sought him out. The actor had visited
the publishing house but had completely forgotten the manuscript....
It was taken from the bottom of his trunk, where it had lain all the
while, and Farnol was minded, first to sell it, with all rights, for $500.
Mr. Hawley said Farnol would do no such thing, “if I have to buy it
myself.” Farnol’s next impulse was to burn the cumbersome bundle.
He finally gave it to his wife, and Mrs. Farnol sent it to her husband’s
mother in England. Shirley Byron Jevons, at that time editor of The
Sportsman, was the next to see it. He took it to the publishing house
of Sampson Low, Marston & Company, introducing it with: “Here is
another Lorna Doone”—Blackmore’s novel having been the firm’s
greatest fiction success. The publishing house accepted the book
and had drawn an agreement with Mr. Jevons as Mr. Farnol’s agent
when the author appeared unexpectedly. In fact, Farnol, discouraged
by his fortunes in America, had simply got on the boat with his wife
and little daughter. A new agreement was drawn with him direct, and
signed. Then, but some time before the book was set up in type, the
publisher showed it to Clement K. Shorter, editor of The Sphere,
whose devotion to the work of George Borrow was well known. Mr.
Shorter’s account of the incident embodies an interesting estimate of
the book:[18]
“I read The Broad Highway with avidity, and recognized at once—
as who would not have done?—that here was a striking addition to
picaresque romances, that the author had not read Don Quixote, Gil
Blas and the best stories by Defoe and Fielding for nothing, nor had
he walked along the broad highways of England without observation
and profit any more than had the creator of Lavengro and Romany
Rye. For the vast multitude of readers of each epoch the dictum of
Emerson stands: ‘Every age must write its own books.’ It is of no use
for the pedantic critic to affirm, with pontifical fervor, that Cervantes
and Le Sage and Defoe are masters of literature and that our
contemporaries are but pigmies in comparison. The great reading
public of any age will not be bullied into reading the authors who
have reached the dignity of classics. The writer who can catch some
element of the spirit of the ‘masters’ and modernize it is destined to
win the favor of the crowd. And thus Mr. Jeffery Farnol has entered
into his kingdom.... The Broad Highway sold in hundreds of
thousands. It is a breezy, healthy book, as unpretentious as it is
sincere. Neither its author nor his friends need to worry themselves
as to whether it is a masterpiece of literature. For our day, at least, it
has added to the stock of harmless pleasures. To the critic who
complains that ‘it is but an exercise in archæology,’ and that the
author ‘has never felt what he has written but has gathered it up from
books, one can reply in the language of Goldsmith’s Mr. Burchell,
‘Fudge.’ It is still possible in England, in spite of its railway trains and
its mechanical development, to feel the impulse which inspired
Charles Dickens, George Borrow and all the masters of the
picaresque romance, who have in days gone by traveled with delight
through the countryside, seeking adventures and finding them. ‘I felt
some desire,’ says Lavengro, ‘to meet with one of those adventures
which, upon the roads of England, are as plentiful as blackberries in
autumn.’ Mr. Farnol has a talent for recreating such adventures, and
he is perfectly frank with his readers, anticipating a certain type of
criticism. ‘Whereas the writing of books was once a painful art,’ he
makes Peter Vibart say in The Broad Highway, ‘it has of late become
a trick very easy of accomplishment, requiring no regard for
probability and little thought, so long as it is packed sufficiently full of
impossible incidents through which a ridiculous heroine and a more
absurd hero duly sigh their appointed way to the last chapters.
Whereas books were once a power, they are of late degenerated
into things of amusement, with which to kill an idle hour, and be
promptly forgotten the next.’”
The letter killeth, but the spirit giveth life. On a famous occasion
the late Maurice Hewlett tore to shreds the historicity of the work of
James Branch Cabell, and Mr. Cabell completely lost his temper. Mr.
Farnol’s hero in Beltane the Smith “finds himself in an England which
from the internal evidence of friars, bowmen, arms and armor we
might vaguely describe as Edwardian (Edward I., II., III.)—the pikes
he appears to have borrowed from a later period. And yet it is not
Edwardian either; for there is no hint of a king in it all, and never,
never was there such an anarchical England, save in the reign of
Stephen of Blois.”[19] Mr. Farnol’s Latin, says J. P. Collins, “gives one
the shivers. He mixes his thee’s and his ye’s, and precisians may
murmur at his forms of archaic diction. But ... if Farnol makes a slip
in the way of detail, or lapses into excess, he preserves the most
important thing, and that is atmosphere.”[20] Everyone will recall
Scott’s inaccuracies in Ivanhoe, of which the most serious was the
depiction of a state of feeling between the Saxons and Normans
existing a century earlier than the time of the novel.
Mr. Shorter has made us longer; it remains to say that The Broad
Highway, accepted in England, was offered by the English publisher
in America, in one instance to Dodd, Mead & Company, who again
declined it. Little, Brown and Company were the acceptors, learning
for the first time of the actor’s delinquency a few years earlier. The
book was published on both sides of the ocean and sprang into
instant success. In the midst of the smother of applause, appeals,
money and golden prospects Mr. Farnol had a moment. He
ejaculated: “Just think! I’ve lost four years of my life!”[21]
v
He was 33—ten years older than it has been done, and thirty
years younger than it has been done, also.[22] Mr. Henry Sydnor
Harrison achieved Queed at 31, Sinclair Lewis was 35 when Main
Street appeared. On the whole, the four years seem not an
excessive price to have paid for a coup, nor thirty-three years long to
have found oneself. The point was neither in the success nor the
time taken to reach it; it was in Mr. Farnol’s ability to keep his head
on his shoulders. This he proceeded to do; although he sold, while
yet unwritten, the serial rights to his next work, pressure upon him
did not prevent his destroying ten unsatisfactory chapters, as has
been related, and although as he said at the time, “I really cannot
blame the magazine people,” he was emphatic in saying, “I never
wish to undergo such an experience again.”[23] Two years elapsed
after the publication of The Broad Highway before the appearance of
The Amateur Gentleman; and except for the publication of a piece of
work written before The Broad Highway[24] and his effort to help in
the war (Great Britain at War), he has had only nine books brought
out in the dozen years since he raised the curtain. And of these one,
The Geste of Duke Jocelyn, a romance in prose, blank verse and
rhyme, is a novelty written for his daughter, Gillian, published
because what had entertained one girl might very conceivably
entertain others. Mr. Farnol’s method of keeping his head on his
shoulders has been to practise industry without becoming industrial.
Although homesick at first on his return to England, he settled in
Kent, at Lee, with a den at the top of the house where he could work
from midnight to breakfast. Old English books lined two walls of this
refuge; another wall was given up to a collection of old pistols and
sabres; and on the desk there usually lay a dictionary of slang dated
1812. More recently the Farnols have lived at the seashore at
Brighton, but the winters are generally spent at Ospedaletti, which is
on the Italian Riviera.
Except for a short visit to report for the London Daily Mail the fight
between Jack Dempsey and Georges Carpentier in 1921,[25] Mr.
Farnol had not revisited America until autumn, 1923. At that time
visitors met a shortish man, anything but a figure of romance, whose
outstanding trait was his genuine friendliness, “a friendliness which is
not an affectation with which he tries to put strangers at their ease,
but an actual part of him.”[26] To see him rehearse and enact, rather
than merely outline, his next novel was an exceptional experience,
for at such times he suits his voice to his characters and displays a
considerable range of dramatic skill. Interviewers developed
evidences of a struggle in the romancer’s mind between the type of
woman he writes about and the types more usual today; however,
chivalry, or perhaps the romantic vision, enabled him to come
through the ordeal by newsprint without dishonor. He denied the
possibility of platonic love and friendships. “After a certain point,
such friendships are bound to be no longer platonic. Mark me! I
know they wouldn’t be in my case, anyway.”[27] A subject he did not
tire of discussing was the wonder of America.[28] To several who
talked with him he expressed the intention of writing another novel
about New York City.[29] “New York should be called ‘The City of
Great Adventures,’” he said, with characteristic enthusiasm,
“because anything might happen in New York.”[30]
More expressive of the man is the story of how his slightest novel
came to be published. He was discussing with his mother the
advisability of bringing out work written before The Broad Highway.
“Look here,” he said to her, “why not rout out Mr. Tawnish? You have
been very good to me, and I can never properly repay you, but if you
can do anything with Mr. Tawnish you shall have it.” The tale—one
that reminds most readers of Tarkington’s Monsieur Beaucaire—was
taken out of a drawer, touched up and added to, and accepted for
book publication. The advance royalties, constituting a generous gift,
were handed over to Mrs. Henry Farnol. This was in the autumn of
1913.
vi
In an article appearing at the height of Mr. Farnol’s first success,
Henry Keats wrote: “The Broad Highway has seemed to the critics to
invite comparison with so many different masters of the English
novel—George Borrow, Blackmore, Le Sage, Dickens, Stevenson,
Thackeray, to mention a few—that I asked him about his ‘foster-
father.’ Mr. Farnol smiled. ‘I would not know my own literary parent if
I met him out here in the broad highway of Kent,’ he exclaimed.
Judging from his subsequent confessions, the creator of Peter Vibart
and Charmian is under greater indebtedness to Laurence Sterne
than to any of the immortals named above. And that was owing to
the friend of his ‘boyish ambitions,’ to whom The Broad Highway is
dedicated. Mr. Shirley Byron Jevons was the first, some years since,
to call Mr. Farnol’s attention to the supreme difficulty of writing a
book dealing with the abstract, citing, as a rare example of success
in that line, Sterne’s Tristram Shandy. A copy of that unusual book
was speedily procured by Mr. Farnol, and he recalls as though an
impression of yesterday the manner in which he was ‘enthralled’ by
its pages. ‘Then,’ he adds, ‘I went on to The Spectator and The
Tatler, the reading of which showed me how great is the loss of those
who are unacquainted with the Queen Anne essayists.’”[31] This
settles the matter of style.
Certain books by Mr. Farnol—The Honourable Mr. Tawnish, Great
Britain at War, and The Geste of Duke Jocelyn, each outside the true
succession of his work—have been incidentally characterized. A
descriptive note on his principal novels may perhaps fittingly
conclude this account.
The Broad Highway (1911) has probably already been sufficiently
described, as it must be familiar to many who may read these lines.
The Amateur Gentleman (1913) has for hero Barnabas, son of a
retired and famous boxing champion of England. Having come into a
legacy, the young man resolves to journey to London to become a
gentleman. The period is that of the Prince Regent. There is a rapid
series of adventures on the journey; Barnabas meets Cleone, the
heroine; he acquires a valet and establishes himself with the quality,
and the fashionable world loses him, for he returns home again.
Beltane the Smith (1915) concerns a golden-haired giant and
matchless swordsman whose odyssey of adventure is lived in a
much earlier England. The Definite Object (1917) is the story of a
young New Yorker whose wealth has taken from him all incentive to
action. For want of a definite object in life he is toying with the
thought of suicide when he surprises a youthful burglar in the act of
entering his rooms. Then, as “Mr. Geoffrey,” he takes up lodgings
with the housebreaker in the old Hell’s Kitchen district of New York—
which was the region between Thirty-fourth and Forty-second
Streets and west of Sixth Avenue. Our Admirable Betty (1918) is a
return to the dimension of The Broad Highway and The Amateur
Gentleman. Black Bartlemy’s Treasure (1920) and Martin Conisby’s
Vengeance (1921) are tales of piracy and the Spanish Main, the
second novel completing the first. Peregrine’s Progress (1922) more
closely than any other book approximates the scenes and action of
The Broad Highway; it is laid in Kent, it relates a boy’s adventurings
on the road and by the roadside, and it reintroduces the Tinker. Sir
John Dering (1923) keeps to the same period. A skilled swordsman
who has incurred the enmity of the Lady Herminia Barrasdaile is
forced to fight duel after duel which she has instigated in the hope
that he will meet his death.
The dictum of Mr. Shorter best fits the case of this friendly writer
and honest workman. We have already quoted the words: “The great
reading public of any age will not be bullied into reading the authors
who have reached the dignity of classics. The writer who can catch
some element of the spirit of the ‘masters’ and modernize it is
destined to win the favor of the crowd.” The love of a fairy tale,
delight in action, pleasure in such characterizations as Black
George, the Tinker, and the Ancient—picturesque; in outline broadly
simple—have been potent. Stevenson was dead; the good as well
as the bad of his legacy had been swallowed up in a flood wherein
the sound could no longer be distinguished from the meretricious;
what we loosely call realism was in the ascendant. Years were to go
by before “realism” could be seen to be the necessary clearing of
paths to an exploration of the romantic impulse more intelligent as
well as more subtle. In the meantime an age-old thirst found these
draughts to quench itself. On the porch of the “Bull” at Sissinghurst
the readers of Mr. Farnol have sat for many an afternoon, washing
the dust from their throats with a pleasant ale and enjoying the
surprising procession of knights, scholars, gipsies, gallants, pirates
and simple maids and ladies of fashion which has passed before
them, coming from and returning to a world without end, truly.
BOOKS BY JEFFERY FARNOL
1907 My Lady Caprice

In England: The Chronicles of the Imp
1911 The Broad Highway
The Money Moon. Earlier, in point of composition, than The
Broad Highway.
1913 The Amateur Gentleman
1913 The Honourable Mr. Tawnish. Earlier, in point of
composition, than The Broad Highway.
1915 Beltane the Smith
1917 The Definite Object
1918 Great Britain at War
In England: Some War Impressions
1918 Our Admirable Betty
1920 The Geste of Duke Jocelyn
1920 Black Bartlemy’s Treasure
1921 Martin Conisby’s Vengeance
1922 Peregrine’s Progress
1923 Sir John Dering
SOURCES ON JEFFERY FARNOL

The Novels of Jeffery Farnol. Booklet published by Little, Brown
& Company, Boston, 1923.
The Country of “The Broad Highway,” by Henry C. Shelley. The
Book News Monthly, October, 1912.
Jeffery Farnol’s Life and Career, by Herbert F. Jenkins. The Book
News Monthly, September, 1911.
How I Began, by Jeffery Farnol. Booklet published by Sampson
Low, Marston & Company, Ltd., London. The text is reprinted from
T. P.’s Weekly of 14 February 1913.
Love Still Ruling Motive in Life of the Modern Youth and Maiden.
Interview by Marguerite Dean in The Evening World, New York, 28
June 1921. Note especially: “As a matter of fact I knew a charming
fellow who did, in real life, just what my hero did in The Definite
Object—took a little girl from the New York slums, educated her,
loved and married her.”
Other references will be found in the footnotes to the text of this
chapter. They do not include, by any means, all the interviews in
newspapers.
5. Adults Please Skip[32]
i
Not that age has anything to do with it. A man is as young as he
feels and a woman is as young as her imagination keeps her. The
idea of never growing up is a mistake. Everyone wants to grow up,
but that’s no reason for not keeping youthful.
There’s something in fellows like Irvin S. Cobb and Owen Johnson
and Ralph Henry Barbour which is just as good at forty as at
fourteen—maybe better. And there’s something in books like Little
Women that you keep coming back to....
S’pose we’d better begin with the Bedtime Story Man. Half the
children in the United States of America are willing to call it a day
when Thornton W. Burgess says the word. Mr. Burgess owes his
success to the fact that he was born in a place called Sandwich in
the State of Massachusetts. It made him realize that something was
needed “between the dark and the daylight,” as Longfellow said.
Having splendid eyesight and some excellent connections, he was
able to enter the best animal circles, and early met Peter Rabbit,
Lightfoot the Deer, and loads of others. The way to meet them is by
all means under Mr. Burgess’s auspices, in The Burgess Animal
Book for Children. Sammy Jay, Bob White and the feathered
companions who have more or less dealings with Striped Chipmunk
and Johnny Chuck are introduced in The Burgess Bird Book for
Children. It is a point of honor with Mr. Burgess always to let his
animal friends tell their own stories. Louis Agassiz Fuertes illustrates
these books with pictures in full colors. For example, he shows fifty-
eight birds in all their glory. The Burgess Bird Book for Children
brought cries of joy from Dr. W. T. Hornaday, who is America’s
leading naturalist and who presides over more animals at the New
York Zoo than went into the ark with Noah. But wait! Here’s a third
volume to put with these two, The Burgess Flower Book for Children,
also illustrated in color and black and white and showing 103
flowers. You should see the color pictures, for instance, of the yellow
adder’s-tongue and the wild columbine! Let it be stressed: the books
by Burgess are the most popular and most successful published for
little children. Their interest and joy is communicable to the child of
four years—and they are read and re-read by boys and girls up to
twelve, and sometimes by their elders.
Rose Fyleman, with Fairies and Chimneys, The Fairy Flute, and
The Fairy Green, won some time ago chief honors as the children’s
poet; and now she seems to be on the path to distinguished honors
for her prose stories. The Rainbow Cat, whose color scheme
included orange hind legs and a red, red tail, gave the greatest
satisfaction, and so will Forty Good Night Tales, in which errant
fairies explain themselves. But probably the most ambitious book is
the new Rose Fyleman’s Fairy Book. Rose Fyleman for fairies, as
the advertisers would say.
Did you see Number One Joy Street, ever? At any rate, you will
see Number Two Joy Street, won’t you? Like the first book, it has a
jolly cover and endpapers, and plenty of illustrations in color and
otherwise. The collection of prose and verse for boys and girls in
Number Two Joy Street is from writers whose names will make even
older people prick up their ears—Gilbert K. Chesterton, Walter de la
Mare, A. A. Milne, Hilaire Belloc, Hugh Walpole, Laurence Housman
and Rose Fyleman are some of them. With such an array of
contributors, you will have very hard work to keep your copy of this
book for your very own. It will be necessary to speak nicely but firmly
to older people.
You will also need to explain that you must be left alone with Edna
Geister’s new book, What Shall We Play. Grownups are almost
certain to think that they ought to stick around and tell you how to go
about the fifty best games in this book. They are wrong. Miss Geister
herself says so. She says she took the very best out of her hundreds
of inspirations for play and took pains to explain them so that
children can play them without help or direction, and at sight. Let’s
Play is another one of her game books written especially for boys
and girls. It Is To Laugh is a little more grown-up (not a great deal).
The Velveteen Rabbit, by Margery Williams, has remarkable
pictures. As you look at William Nicholson’s drawings (they were
made on stone, which gives them their peculiar texture) you can
really see the sawdust hero come to life and leap for joy! We will
come in a moment to other books with glorious pictures, but first let
us see if we have one or two more books for four and five and six
years. Yes. Here is Mother Hubbard’s Wonderful Cupboard, by
Maude Radford Warren and Eve Davenport, who also wrote Tales
Told by the Gander and Adventures in the Old Woman’s Shoe. The
scheme in each of these books is the delightful one of continuing the
Mother Goose stories. Mother Goose, with her unfinished tales, is
extremely tantalizing. Probably the good woman told all she knew,
but it is by no means enough. For instance, she appeared to know
nothing of the circumstances in which Mother Hubbard acquired her
dog. They were highly interesting. You see, she needed some one to
work for and be interested in, and as no child was available she took
Diccon, who was a trained, performing dog attached to a circus, but
so ill that his owner thought he wouldn’t live long anyway. C. A.
Federer has made the many illustrations, some of them in color.
The Wiggly Weasel and Other Stories, by Mabel Marlowe, is
another book with many pictures that is full of the fun of clever
animals, brownies, and their kin.
Perhaps you take part in plays? Then, if you are young enough, I
think you will be enthusiastic over the seven gay masques in The
Magic Sea Shell and Other Plays, by John Farrar. There is probably
a place in the garden that looks as though a play were about to
begin there, or a spot down the meadow or a roomy chimney-corner
in the house. Home-made music and costumes, please!
For seriousness, and in moments when you want to know more
about the world you live in, and how men came to live in it, anyway,
the two most helpful books are likely to be Frederic Arnold Kummer’s
The First Days of Man and The First Days of Knowledge. These
relate the true fairy story of Creation and man’s coming-to-know.
Moreover, it is told in such a way that any boy and most girls can
make for themselves the simple tools that man first began by
making.
ii
I spoke of books with pictures. If you are so lucky as to have the
Fairy Tales by Hans Andersen, illustrated by Kay Nielsen, or those
other books for which he made the illustrations, The Twelve Dancing
Princesses and East of the Sun and West of the Moon, I hope you
will say a word to your older friends about this artist. They are likely
to be as ecstatically happy as yourself, in the contemplation of the
pictures, but not to know what it is they admire. Then you must tell
them that Kay Nielsen is a Dane, the son of an actor and a famous
actress, who was brought up in a home where the rich furnishings
and beautiful colors came from Constantinople and the East. He
went to London and saw drawings by Aubrey Beardsley in which all
the lines combined elegance, suavity (or great smoothness), power
(or sureness and ease), and a certain austerity (bareness,
simplicity). And so, by what he had seen and by the nature received
from his parents, he became a great artist who could do fiery work
with an occasional effect of grim strength; but in these pictures you
know he is riotously playful with his lines and his colors alike.
Edmund Dulac and Arthur Rackham are other great artists who
have done much of their finest work in illustrating children’s books.
You may have Hawthorne’s A Wonder Book, illustrated by Rackham,
or Edmund Dulac’s Fairy Book, or Stories From the Arabian Nights,
The Sleeping Beauty and Other Fairy Tales, or Stories from Hans
Andersen, each with Dulac’s pictures. Perhaps the most wonderful of
all Dulac’s books is the edition of Shakespeare’s comedy of The
Tempest.
iii
You are old enough to be thinking about going away to school. You
can’t get too many school stories. In particular, you are keen for a
new book by Ralph Henry Barbour. Nobody writes better school
sport books! The Fighting Scrub, Mr. Barbour’s latest, is a picture of
life at a famous New England school; and the fellows and the
incidents of the tale are just as actual as the setting. Clif Bingham
and Tom Kemble are boys everyone can recognize among his
friends, and while Loring Dean, a cripple confined to a wheel-chair, is
a new character in a story of boys, his splendid head-work in
planning a forward pass play that makes the winning touchdown for
the school is proof that a boy need not be an athlete to count.
There are some very pretty points about The Fighting Scrub. It has
been usual to write only about a fellow who “made the team.” The
scrub team has been an unsung, unhonored aggregation on which
the first team sharpened its teeth. Mr. Barbour’s hero is only a scrub;
but even a scrub has been known to play in the big game and with
crucial results. Here’s another thing: people have begun to recognize
the fact that we are in danger of losing sight of football and other
games as sport, and of thinking only of winning. Nothing could show
better than the history of Clif Bingham and Tom Kemble in The
Fighting Scrub that the real joy of football lies in the spirit in which
you play. Every fellow can see himself in Clif or Tom or Loring Dean.
Again, there are thousands of boys who will be able to see
themselves in Joe Kenton, the hero of Mr. Barbour’s Follow the Ball.
Joe is a fellow who is far from having things all his own way, but he
is a sticker. He has to earn the money to get through school, and that
never made it easier to make a record in athletics. But he shows up
well, and Follow the Ball has baseball, skating, hockey and camping
in its pages as well as football.
The proved classics in the way of school stories are assuredly
Owen Johnson’s. It is sixteen years since the first publication of his
first book of Lawrenceville stories, but Hickey Hicks, Dink Stover,
Doc Macnooder, Hungry Smeed, the Gutter Pup, the Tennessee
Shad and Lovely Mead are as “generally and specifically bully” as
when Booth Tarkington hailed them. Mr. Johnson’s success in The
Prodigious Hickey, The Varmint and The Tennessee Shad is as
great as Mr. Tarkington’s own in Penrod; immeasurably greater than
Kipling’s effort in Stalky & Co. It is true to say that the Lawrenceville
stories blend speed, surprise, mischief and humor with a
smoothness and a perfection untouched by anything else of their
sort. They avoid the utter priggishness and complacency of Tom
Brown’s School Days, while having the same positive value of a real
school, under its own name and with its own tradition, as their
background. “The only real prep school story ever written,” said
George Ade, crisply, after reading The Varmint. Why? No doubt the
fact that the Lawrenceville stories are semi-autobiographical has
much to do with it. For Johnson was a Lawrenceville boy in the
1890s; there is extant a picture showing him with the original (but
somewhat older) Brian de Boru Finnegan, Turkey Reiter, the Old
Roman, and the Prodigious Hickey. Johnson himself it was who held
the skeleton while Hickey attached it to a rope hung from a ventilator.
Johnson sat on the roof when Old Ironsides—afterward a New York
real estate broker—slid off and got filled with gravel. It was Johnson
who experienced the agony of muffing a ball and being attacked by
the whole baseball team, which he has described as the Varmint’s
first discouraging experience with Lawrenceville athletics.
And after a dozen years, Mr. Johnson recently returned to the
Lawrenceville scene in Skippy Bedelle, which tells how Skippy
planned to invent a foot regulator for bathtubs and of certain
deplorable experiments which were to produce mosquito-proof
socks. Skippy Bedelle is largely the story of a sentimental
progression and includes the first dress suit and Skippy’s first girl.
In the days when professional ballplayers still had mustaches and
you could give people a thrill by riding down the post-office steps on
a high-wheel bicycle, Irvin S. Cobb was goin’ on fourteen. And in
Goin’ on Fourteen, his new book with pictures by Worth Brehm, the
artist for Tarkington’s Penrod stories, Cobb has cut a few cross-
sections out of a year in the life of an average boy. Now without in
any respect being literal reminiscences, these chapters accurately
and joyously reflect a scene and a period and a boy most
unmistakably American. For Johnny Custer, otherwise John C.
Calhoun Custer, Jr., is neither Tom Sawyer nor Penrod Schofield—
though perhaps more like Tom and Huck Finn than Penrod—but he
is as instantly recognizable and as entirely “boy” as either. And
Johnny Custer was his own trained investigator; he did not depend
upon others to tell him what would happen in untried circumstances,
no, sir! The account of how he and Mr. Simons short-circuited the
fowls of a chicken fancier should be read with caution; it is likely to
leave you in the same condition of happy helplessness in which it left
Johnny.
Albert Payson Terhune, like John Taintor Foote,[33] writes a capital
dog story, and Mr. Terhune’s stories of collies, Buff and Lad and the
others, are known wherever the dog has his due. In The Heart of a
Dog, Terhune’s new book, there are one or two tales in which a collie
is not the hero, but Lad and Buff and Treve and Lochinvar Bobby are
familiar friends of the breed which Mr. Terhune himself raises and
takes prizes with. Marguerite Kirmse has made the eight pictures in
color and others in black and white. This is another book that grown-
ups will borrow and neglect to return, if you don’t watch out.
Every boy and girl knows how hard it is to find a good, readable
history. The difficulty was pointed out to Sidney Dark, who set to
work at once to do something about it. And so far he has done
magnificently, producing, in The Book of Scotland for Young People,
The Book of England for Young People, and The Book of France for
Young People, three histories more clear and interesting to boys and
girls of ten to sixteen than any similar accounts. (I do not even
except Charles Dickens’s A Child’s History of England, which is one-
sided in spots.) Each of Mr. Dark’s books has sixteen illustrations
from famous paintings of historic scenes.
iv
There is an Everyman’s Library—why not an Everychild’s Library?
Well, one has been begun. It is called The Beacon Hill Bookshelf
and already eight volumes are to be had. One of them, The Boy
Whaleman, has never been published before; the others are all
established favorites stamped with the approval of librarians and

Ebook A First Course in Control System Design 2Nd Edition Kamran Iqbal Online PDF All Chapter

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Ebook A First Course in Control System Design 2Nd Edition Kamran Iqbal Online PDF All Chapter

Uploaded by

Copyright:

Available Formats

A First Course in Control System

Design 2nd Edition Kamran Iqbal

Robust Control System Design 3rd Edition Chia-Chi Tsui

A First Course in Machine Learning 2nd Edition

Abstract Algebra: A First Course 2nd Edition Stephen

Nonlinear Functional Analysis - A First Course 2nd

System Reliability and Security: Techniques and

Advances in Power System Modelling, Control and

A First Course in Spectral Theory 1st Edition Milivoje

A First Course in General Relativity 3rd Edition

A First Course in Control System Design

QUAN MIN ZHU

The “River Publishers Series in Automation, Control and Robotics” is a series of

• Robots and Intelligent Machines

For a list of other books in this series, visit www.riverpublishers.com

ISBN: 978-87-7022-152-8 (Hardback)

©2020 River Publishers

All rights reserved. No part of this publication may be reproduced, stored in

List of Figures xxiii

List of Tables xxix

List of Abbreviations xxxi

1 Mathematical Models of Physical Systems 1

2 Analysis of Transfer Function Models 31

2.2 System Response to Inputs . . . . . . . . . . . . . . . . . . 36

3 Analysis of State Variable Models 63

4 Feedback Control Systems 93

5 Control System Design Objectives 111

6 Control System Design with Root Locus 133

7 Design of Sampled-Data Systems 167

7.1.2 Zero-Order Hold . . . . . . . . . . . . . . . . . . . 171

8 Controller Design for State Variable Models 211

8.4.3 Deadbeat Controller Design . . . . . . . . . . . . . 238

9 Frequency Response Design of Compensators 247

About the Author 289

The organization of this monograph is as follows: Chapter 1 discusses

improvement, and improvement to the sensitivity robustness. The chapter

I would like to extend my gratitude to my teachers for instilling the love of

Figure 1.1 An RC circuit. . . . . . . . . . . . . . . . . . . . . 5

Figure 2.6 Sinusoidal response of DC motor. . . . . . . . . . 52

Figure 5.1 Feedback control system with plant G(s), sensor

Figure 6.10 Step response of phase-lead plus PI compensated

Figure 9.11 PID compensator design: PI compensation (a),

Table 2.1 Quality metrics used to measure second-order system

Common Symbols used in the book

elements, e.g., positions and velocities in mechanical systems, or currents

1.1 Modeling of Physical Systems

1.1.1 Model Variables and Element Types

1.1.2 First-Order ODE Models

Figure 1.1 An RC circuit.

Figure 1.2 An RL circuit.

Example 1.1: A series RC network.

Example 1.2: A parallel RL network.

Example 1.3: Inertial mass acted upon by a force.

Figure 1.4 Room heating with heatflow through walls.

Figure 1.5 Fluid reservoir with constricted outflow.

In terms of the temperature differential, ∆θ = θr − θa , the govering

1.1.3 Solving First-Order ODE Models with Step Input

y(t) = [1 + (y0 − 1)e−t/τ ]u(t)

y(t) = [y∞ + (y0 − y∞ )e−t/τ ]u(t)

Assuming zero initial conditions, i.e., y0 = 0, the output is expressed as:

y(t) = (1 − e−t/τ )u(t)

We may evaluate the output at t = kτ, k = 0, 1, . . . to compile the

Time Output Value