!"#$%&'$# !)*%%+ ,'$-'.

/0*1-.2'3 043'$ #2' !56789!76 :$%;'<#= >-#2 #2?4+. #% @A,AB C'*$0?$D EFGH


"#$%&' (#)*'+,$
-./+$ (0&$$&1
m
x
F
k b
G
H
y
u
u
operator

-
+
!"#$%&'$# !)*%%+ ,'$-'.
/0*1-.2'3 043'$ #2' !56789!76 :$%;'<#= >-#2 #2?4+. #% @A,AB C'*$0?$D EFGH

!"#$%&'%() +(,-./01%-(

23& */%3.0 40*%&5/11# *,6).71&84&$ $/99.0% /)8&0 %3& :;"; <)).=*%+.) +)
;&0.$9*,& <)$%0/,%+.) <)+%+*%+=&> :;"; ?0*)% :.@ ::ABC;DEF?@ 2.)# "90+)4&0 +$
%3& 2&,3)+,*1 G.)+%.0@

23+$ =&0$+.) +$ 8*%&8 D&H0/*0# IJ> KBIL@ M.9#0+43% &N,&9% 73&0& +)8+,*%&8> +$ 3&18
H# :*0*#*)*) G@ O.'&0*%3 *)8 -./+$ (0&$$&1@ P1&*$& ,.)%*,% 6.'&0*%3Q4*%&,3@&8/
5.0 +)5.0'*%+.) *)8 9&0'+$$+.) %. ,.9#@



2%&3$0%/4.

Any opinions, findings, and conclusions or recommendations expressed in this material
are those of the author(s) and do not necessarily reflect the views of the National
Aeronautics and Space Administration.

Georgia Institute of Technology
System Dynamics
Louis Dressel
Contents
1 Math Review 3
1.1 Complex Number Review . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.2 Zeros/Poles of a Function . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.3 Linearization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
2 The Laplace Transform 5
2.1 Properties of the Laplace Transform . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
3 Systems and System Types 7
3.1 Denition of Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.2 Input Range . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.3 Linearity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.4 Time Variance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
4 Mechanical Systems 8
4.1 Elements of Mechanical Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
4.2 Basic Mechanical Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
4.3 Degrees of Freedom . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
5 Transfer Functions 11
5.1 Transfer Function Denition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.2 Block Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
5.3 Convolution Integral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
6 State-Space Models 12
7 Time-Domain Analysis of Dynamic Systems 13
7.1 Dynamic System Responses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
7.2 First-Order Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
7.3 Second-Order Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
7.4 Goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
8 Frequency-Domain Analysis of Dynamic Systems 17
9 Basic Control Theory 18
9.1 Closed-Loop System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
9.2 Open-Loop System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
9.3 Block Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
9.4 SAS vs CAS . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
1
CONTENTS 2
10 Control Systems 20
10.1 P-Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
10.2 PI-Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
10.3 PD-Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
10.4 PID-Controller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
Index 21
Chapter 1
Math Review
1.1 Complex Number Review
Polar Form
z = |z|e
i
(1.1)
n
th
Root of a Complex Number
z
1/n
= |z|
1/n
e
i(+2k)/n
(1.2)
1.2 Zeros/Poles of a Function
The zeros of a function f(x) are all values of x such that f(x) = 0. These values of x make the
numerator equal to zero.
The poles of a function f(x) are all the values of x for which f(x) approaches innite. These
values of x make the denominator equal to zero.
1.3 Linearization
Linear functions are much easier to work with than nonlinear functions. Unfortunately, not every-
thing worth modeling acts in a linear fashion.
For example, consider the equation for the drag force on an object.
D =
1
2
v
2
SC
D
Assuming a constant density, reference area, and drag coecient, drag only varies as a function of
the square of velocity. This is not linear relationship.
The solution to this is to linearize the function in question. This requires taking a line tangent
to the function at some point and using the line to represent the function. This is shown in Fig-
ure 1.1. In this example, the function y is a function of x. The value of x at which the tangent line
is generated is denoted by x. The linear approximation is only valid for a limited range of x-values
around x. If the function is highly non-linear, this range could be very small.
3
CHAPTER 1. MATH REVIEW 4
x
y
Figure 1.1: Example of linearization.
The process of linearizing a function f(x) about some value x involves nding the function value
at x, and nding the slope of the tangent line at this point (with calculus). The values of points
near x can be determined from this. The process is shown below:
f(x) f( x) +
df
dx

x= x
(x x)
A multivariable function z = f(x, y) can be linearized about x and y:
z f( x, y) +
f
x

x= x
y= y
(x x) +
f
y

x= x
y= y
(y y)
Chapter 2
The Laplace Transform
The Laplace transform is a type of transform function. Transform functions maps a function f(t)
in one domain into another function F(s) in another domain. The general equation of a transform
function is shown in Equation 2.1:
F(s) =

k(t, s)f(t)dt (2.1)

0
e
st
f(t)dt (2.2)
2.1 Properties of the Laplace Transform
The Laplace transform is a linear operator. This means it comforms to the two dening properties
of linearity: homogeneity and additivity.
Homogeneity is also called the scalar rule. Let the constant a be some real or complex number:
L(af(t)) = aL(f(t))
Additivity is also called the superposition principle. The Laplace transform of the sum of two
functions f
1
and f
2
is equal to the sum of the Laplace transforms of the individual functions.
L(f
1
(t) + f
2
(t)) = L(f
1
(t)) + L(f
2
(t))
Combined, these properties mean the Laplace transform is a linear operator.
L(a
1
f
1
(t) + a
2
f
2
(t)) = a
1
L(f
1
(t) + a
2
L(f
2
(t))
Dierentiation
L[f

(t)] = sL(f(t)) f(0)

(t)] = s
2
L(f(t)) sf(0) f

(0)
L(f
(n)
(t)) = s
n
L(f(t)) s
n1
f(0) ... sf
(n2)
(0) f
(n1)
(0) (2.3)
5
CHAPTER 2. THE LAPLACE TRANSFORM 6
Integration
Final Value Theorem
If lim
t
f(t) exisits, then
lim
t
f(t) = lim
s0
sF(s)
Initial Value Theorem
If the Laplace transform of f(t) and
d
dt
f(t) exists, and if lim
s
sF(s) exists, then
f(0
+
) = lim
s
sF(s)
It is important to note that f(0) = lim
s
sF(s) is usually true too. However, it is not true for
impulse functions.
Chapter 3
Systems and System Types
3.1 Denition of Systems
3.2 Input Range
A static system is a system in which the output is a function of the current input.
A dynamic system is a system in which the output is a function of the current input and its
history.
A causal system is a system in which the output does not depend on future input. Most sys-
tems are causal.
3.3 Linearity
In a linear system, the governing equation is linear.
In a nonlinear system, the governing equation is nonlinear.
3.4 Time Variance
7
Chapter 4
Mechanical Systems
4.1 Elements of Mechanical Systems
There are three basic elements of mechanical systems that we will consider. Each element exerts a
force to resist a dierent kinematic quantity. Spring elements resist displacement, damper elements
resist velocity, and mass elements resisit acceleration. The general equation for the force exerted by
a basic mechanical element is:
Force = (constant)(kinematic quantity)
The negative sign is present because the force always resists the kinematic quantity, and thus works
against it.
Spring Element
Spring elements resist displacement. The force F exerted by a spring is given by Hookes law:
F = kx
where k is known as the spring constant, and x is the displacement.
When springs are in parallel, the total spring constant can be found by adding the two individual
spring constants.
When springs are in in series, their spring constant can be found by adding the reciprocals of
individual spring constants, and taking the reciprocal of this value.
Damper Element
Damper elements resist velocity. If an object connected to a damper is in motion, a damper will
exert a force to stop the motion. The damping force exerted in response to velocity can be found
with the following relationship:
F = b x
As shown in the above equation, we denote the damper constant with b, and x is the velocity.
Mass Element
Mass elements resisit acceleration.
F = m x
8
CHAPTER 4. MECHANICAL SYSTEMS 9
The negative sign makes this dierent from the F = ma you might already be familiar with. However,
the F in that equation is referring to the total force exerted on the object, which acts in the direction
of acceleration. In our equation, we are discussing the inertial resistance to acceleration, which acts
opposite to the acceleration. This is the reason for the negative sign.
This dierence can also be explained with dAlemberts Principle. DAlemberts principle is simply
an alternative form of Newtons second law. While Newtons second law can be expressed

F = ma
DAlemberts principle expresses it as

F ma = 0
This allows a problem in dynamics to be reduced to a problem in statics. This latter format suggests
an equilibrium between the force

F exerted on the object and the inertial force ma that opposes
the object.
4.2 Basic Mechanical Systems
Mass-Spring System
This system includes a mass element and a spring element. This system is shown in Figure 4.1.
m
x
F
k
Figure 4.1: Mass-spring system experiencing a force.
The coordinate system denes positive x as downward.
m
F
mx
kx
Figure 4.2: Force diagram of mass-spring system.
CHAPTER 4. MECHANICAL SYSTEMS 10
The mathematical model of the system is shown below:
m x + kx = F
Mass-Spring-Damper System
This system includes a mass element, a spring element, and a damper element.
m
x
F
k b
Figure 4.3: Mass-spring-damper system.
The mathematical model of the mass-spring-damper system is shown below.
m x + b x + kx = F
4.3 Degrees of Freedom
Chapter 5
Transfer Functions
5.1 Transfer Function Denition
The transfer function of a dynamic system is the ratio between the Laplace transform of the
output over the Laplace transform of the input with all initial conditions set to zero.
Transfer Function =
L(output)
L(input)
We denote the transfer function with G, the Laplace transform of the output with Y , and the Laplace
transform of the input with U or X:
G(s) =
Y (s)
U(s)
(5.1)
Because Laplace transforms convert dierential equations into algebraic equations, the transfer func-
tion is always a ratio of polynomials.
Example:
5.2 Block Diagrams
5.3 Convolution Integral
Given g(t), the convolution integral can be used to nd the response to some input u(t).
y(t) =

t
0
g()u(t )d (5.2)
This is especially useful in nding the unit step response.
11
Chapter 6
State-Space Models
x = Ax + Bu
y = Cx + Du
Matrix A is the state matirx.
Matrix B is the input matrix.
Matrix C is the output matrix.
Matrix D is the direct transmission matrix.
12
Chapter 7
Time-Domain Analysis of Dynamic
Systems
7.1 Dynamic System Responses
The response, or solution, of a system refers to
For example, recall the mathematical model of a mass-spring-damper system:
m x + b x + kx = f(t) (7.1)
The solution, or response, to this system would simply be the position as a function of time, x(t).
x(t) = x
c
(t) + x
p
(t)
Complementary Solution
The complementary solution assumes no input. In equation 7.1, this would be achieved if f(t)
were simply set to zero. Because it only relies on the initial conditions, and not the input, the
complementary solution is often called the initial condition response. It is also called the natural
response.
Particular Solution
The particular solution is often called the forced response.
Transient Response
7.2 First-Order Systems
The general equation of a rst order system is:
x + ax = bu
Where a and b are simply constants. The transfer function G(s) can be found using the Laplace
transform:
G =
X
U
=
b
s + a
It becomes useful to dene the transfer function in terms of a time constant and a gain k.
x + x = ku (7.2)
13
CHAPTER 7. TIME-DOMAIN ANALYSIS OF DYNAMIC SYSTEMS 14
Initial Condition Response
The complementary, or initial condition, response, is the response of the system to specic initial
conditions and no input. Thus, in solving for it, the right-hand side of Equation 7.2 is set to zero.
x + x = 0
Taking the Laplace transform of this function yields
sX x
0
+ X = 0
Rearranging
X =
x
0
s + 1
Taking the inverse Laplace yields
x(t) = x
0
e
t/
(7.3)
Unit Impulse Response
If u = (t), what is x(t)? Because the Laplace transform of the unit impulse function is one, the
output of the system x(t) is equal to the inverse Laplace of the transfer function.
x(t) = g(t) = L
1
(G(s))
=
k

e
t/
(7.4)
Unit Step Response
If u = 1(t), what is x(t)? The convolution integral can be used for this problem.
x(t) =

t
0
g(t
1
)u(t t
1
)dt
1
=

t
0
k

e
t
1
/
dt
1
=
k

[e
t
1
/
]

t
0
= k[1 e
t/
] (7.5)
7.3 Second-Order Systems
A typical second-order system is the mass-spring-damper system that has already been explored.
m x + b x + kx = u
Imagine that there is no damping in this system (b = 0). If displaced from its equilibrium position,
the mass would oscillate indenitely. This would not happen in reality, because no real system is
completely devoid of damping. But an ideal second-order system with no damping would oscillate
forever. This is shown in Figure 7.1. The amplitude of the response is constant because no damper
removes energy from the system.
CHAPTER 7. TIME-DOMAIN ANALYSIS OF DYNAMIC SYSTEMS 15
t
x
Figure 7.1: Initial condition response of a second-order system with no damping.
The frequency at which the undamped system oscillates is known as the natural frequency. For
any second-order system, damped or not, the natural frequency is the square root of the spring
constant divided by the mass. Equation 7.6 shows how the natural frequency,
n
, is calculated. If
the units are consistent, this frequency is given in rad/s.

n
=

k
m
(7.6)
Now imagine that there is light damping, so that the damping constant is now greater than zero.
Because the damping component reduces the energy of the system, the amplitude of the response
continually diminishes until it is zero. The oscillations no longer occur at the natural frequency, but
occur at a lower frequency. If the damping is increased, the amplitudes diminish more quickly, and
the frequency of the oscillation becomes smaller.
Figure 7.2: Eect of slight damping on the response of a second-order system.
If the damping is increased to a certain value, the system no longer oscillates, and converges to the
steady state value. At this value of the damping constant, the system is said to be critically damped.
For damping constants less than this value, the system is under-damped. For damping constants
larger than this value, the system is over-damped. The ratio of the damping constant to the critical
damping constant is known as the damping ratio, denoted by . Equation 7.7 shows the how the
damping ratio is calculated from the mass, spring constant, and damping constant.
=
b
2

km
(7.7)
It has been noted that damping causes a system to oscillate at a frequency other than its natural
frequency. The frequency at which the oscillations occur in a damped system is the damped
frequency,
d
. If the damping is increased on the system, the damped frequency decreases.

d
=
n

1
2
(7.8)
Now that the concepts of natural frequency and damping ratio have been introduced, it becomes
convenient to rewrite the mathematical model of the second-order system in these terms:
x +
b
m
x +
k
m
x =
1
m
u.
CHAPTER 7. TIME-DOMAIN ANALYSIS OF DYNAMIC SYSTEMS 16
x + 2
n
x +
2
n
x =
1
m
u (7.9)
7.4 Goals
1. What are the complementary and particular solutions of a dynamic system?
2. What eect does damping have on the response of a system?
3. Dene the natural frequency, damped frequency, and damping ratio of a second-order system.
Chapter 8
Frequency-Domain Analysis of Dynamic
Systems
17
Chapter 9
Basic Control Theory
Until this point, we have only explored the responses of systems to specic inputs. However, it is
useful to modify a system such that it responds to specic inputs in a particular manner. In order
to control an existing system, or plant, a controller is added. This controller modies the system so
that it will respond to inputs in a manner that we designate. A successfully controlled system will
produce correct outputs based on the input of a user.
Control systems are divided into two main groups depending on their use of feedback. Feed-
back is the comparison of system output to the desired input. Systems that use feedback are known
as closed-loop systems. Systems that do not use feedback are known as open-loop systems.
9.1 Closed-Loop System
Closed-loop systems are systems that employ the use of feedback. They are commonly called feedback
control systems.
9.2 Open-Loop System
9.3 Block Diagrams
Including Feedback
Masons Rule for Block Diagram Reduction
G =
Sum of forward path gains
1 + sum of loop path gains
(9.1)
9.4 SAS vs CAS
As shown in the gure above, the controller can be in the forward path or in the feedback path.
Positioning the controller in the feedback path yields a stability augmentation system (SAS), and
positioning the controller in the forward path yields a command augmentation system (CAS).
18
CHAPTER 9. BASIC CONTROL THEORY 19
Stability Augmentation System
G
H
y
u
u
operator

-
+
Figure 9.1: Stability Augmentation System.
Command Augmentation System
H G
y
u
u
operator

-
+
Figure 9.2: Command Augmentation System.
Chapter 10
Control Systems
10.1 P-Controller
C = Kp (10.1)
10.2 PI-Controller
C = K
p
+
K
I
s
(10.2)
C =
K
I
(
Kp
K
I
s + 1)
s
(10.3)
10.3 PD-Controller
C = K
p
+ sK
d
(10.4)
10.4 PID-Controller
20
Index
Closed-loop, 18
Complementary solution, 13
Damped frequency, 15
Damper, 8
Damping ratio, 15
Linearization, 3
Mass-spring-damper system, 10
Natural frequency, 15
Open-loop, 18
Particular solution, 13
Poles, 3
Spring, 8
Transfer function, 11
Zeros, 3
21