SCE 507

Linear Control Systems

Dr. NM Alyazidi
Term: 241
 Course Information

OFFICE: 233
PHONE: 860 1484
EMAIL: [email protected]
OFFICE HRS: Mon, Wed.: 11:00-1230
14:00-15:00
or by appointment.
GRADING: H/W & Simulation 10%
Midterm 30%
Final (exam2) 30%
Term Project 20%
Quiz (IBL) 10
TEXTBOOK Linear System Theory and Design, 3rd
Edition, C. T. Chen

References
• Linear System, Panos Antsaklis and Anthony Michel
• Linear System, Thomas Kailath
• Class Notes

EXAMINATIONS:
Exam I:
Exam II:
Course Objective:
This course provides a basic understanding of
linear multivariable systems through their modeling
and analysis. Both continuous-time and discrete-
time systems will be discussed in the course. After
taking this course, the student will be in a position
to move on to more advanced courses and topics
in systems, control, communications and signal
processing.
 TENTATIVE COURSE OUTLINE

•Overview
•Mathematical Description of Systems
– Input-Output Description
– State-Variable Description
•State Space Solutions and Realization
•Stability of Linear Systems
•Controllability and Observability
•Canonical Decompostion
•Minimal Realizations
•State Feedback and State Estimators
•Other Topics as Time Allows
 TENTATIVE COURSE OUTLINE

• Introductory material:
Review of linear systems and matrix algebra.
• State-Space Solutions and Realizations:
Solution of LTI systems, State equations,
Realizations
• Structural Properties:
Stability, Controllability and Observability
• Minimal Realizations:
Coprime Fractions, Balanced Realizations
• State Feedback and State Estimators/Observers:
Design methods
• 6. Project
 State Space Solutions and
Realization
• State Space Representation: A mathematical model
that represents a physical system in terms of state
variables, inputs, outputs, and system dynamics. It is
used to describe multi-input multi-output (MIMO)
systems in a compact form.

• Realization: Refers to finding a state-space model that

matches a given input-output behavior. Different
realizations can represent the same system, but with
varying numbers of state variables.
 Stability of Linear Systems
• Stability: Indicates whether a system will return to its
equilibrium state after a disturbance. For linear systems,
stability is analyzed using eigenvalues of the system
matrix. If all eigenvalues have negative real parts
(continuous systems) or lie within the unit circle (discrete
systems), the system is stable.
 Controllability and Observability
• Controllability: A system is controllable if it is possible
to move the system from any initial state to any desired
state using an appropriate control input. This property is
crucial for designing effective controllers.

• Observability: A system is observable if, by observing

the outputs over time, it is possible to determine the
internal state of the system. This is important for state
estimation and for ensuring that the system behaves as
expected under different conditions.
 Minimal Realizations
• Minimal Realization: The simplest (in
terms of the number of state variables)
state-space representation that still
accurately represents the input-output
behavior of the system. Minimal
realizations are desirable because they
reduce computational complexity and
improve system performance.
 State Feedback and State
Estimators
• State Feedback: A control strategy where the control
input is designed based on the current state of the
system. By feeding back the state into the control law,
one can achieve desired system behavior, such as
stability or optimal performance.

• State Estimators (Observers): When not all states can

be measured directly, state estimators are used to
reconstruct the unmeasurable states based on the
available measurements and the system model. The
most common type of state estimator is the Kalman filter.
 The Study of Systems
Systems: It is a medium that relates a cause to an
effect, or an input to an output.
The study and design of physical systems often
consists of:
1. Performance specifications
2. Modeling
3. Simulations
4. Analysis
5. Optimization
6. Physical Realization
 Example of a non-physical
system:
• Finance

• Social Networks

• Economic Models
 Control Design
• The classical control design may fail for complex systems
such as those with multiple inputs and multiple outputs, or
systems with nonlinearities, or fast change of initial
conditions example:
– High-speed flight control.
– Flow control of high-speed communication networks or multi-phase
gases.
• So, usefully, we start with a simple mathematical model of
the system to be controlled.

• In control theory, people focus on math, we need a

mathematical representation of the physical system
Modeling
Modeling is the representation of a system and
all its components in a mathematical form.

Depending on the questions asked, or

depending on the operating ranges, a physical
system may have different models

Example: An Automobile may be modeled as a

single particle if we are studying traffic flow but
may be modeled as a spring-mass-damper
system if we are interested in the vibration of the
occupants
 System’s Models
• Input-output Model: (identification)
– Blackbox model( Neural networks,
Hammerstein Models)
– Here, we don’t care about the inner dynamics
(working) of the system.
• State variable model: (our focus)
• Here, we model everything
• Mathematical Modeling: Obtain a set of
equations that describes the behavior of
the system.
• A model is a simplified representation or
abstraction of reality.
• It is not necessary to take the system
apart, to model every screw, valve, and
axle. We use a simple model that gives a
reasonably accurate description of the
system behavior.
• Once a model is selected for a physical
system, the next step is to develop
mathematical equations to describe the
system from the fundamental physical
principles such as:

• Newton’s law for mechanical systems

• Kirchhoff’s voltage and current laws in
electrical systems
• Laws of thermodynamics and transport
phenomena in fluid and thermal systems
• Modeling of Translational Mechanical Systems
– Basic laws
– Finding Input-Output ODEs
– Solution of ODEs for Mechanical Systems
• Modeling of Electrical Systems
– Review Basic laws
– Finding Input-Output ODEs
– Solution of ODEs for Electrical Systems
– Transfer Function
• After the mathematical equations for the
model have been obtained the next step in
the study of systems involves both

• Quantitative
– System responses to specified inputs

• Qualitative
– Stability
– Controllability
– Observability
– If the response of the system is found to be
unsatisfactory then an engineering design phase
termed improvement or optimization is initiated

– In some cases a system parameter may be adjusted

to improve the response but in other cases
compensation devices must be injected into the
system

– Finally in the realization phase the proposed system

must be built using actual physical hardware
 Complex System
• In fact, there is no physical system that is truly linear.
– It isn’t easy to model accurately.
– Unexpected behavior.
– Needs complex solutions.
• We may apply linearization methods.
• Why do we use them? The reasons that
• Control theory is effective in practice even when this
complex assumption fails because physical can be
approximated by linear models.
• Control systems can be designed to be robust with
respect to uncertainties in the system model.
 Multivariable Computer-controlled Systems
is divided into three parts:
• Preliminary algebraic material describing the established
fundamentals of polynomial and rational matrices
necessary for the understanding of later chapters.
• Control problems, important in their own right but which
also have a substantial bearing on what follows (eigenvalue
assignment and the use of z- and zeta-transforms in
discrete systems).

• Frequency methods for the investigation of MIMO

sampled-data systems (parametric discrete-time models;
stability; stochastic methods; H2 optimization and L2
design).
SYSTEMS AND CLASSIFICATION
OF SYSTEMS

KFUPM-CIE Department-SCE
507
 What is a system?
• A system is formally defined as an entity
that manipulates one or more signals to
accomplish a function in order to yield new
signals.

KFUPM-CIE Department-SCE 507

 System Classification

System Classes

Distributed Lumped

Parameter

Stochastic Deterministic

Continuous Discrete
Time Time

Nonlinear Linear Nonlinear Linear

Time Varying Time Invariant

Linear/Nonlinear

A system which is not linear is said to be nonlinear

KFUPM-CIE Department-SCE
507
 Memoryless/With Memory

A system is considered to be memoryless if the value of the output signal at every

moment in time is determined only by the value of the input at the same point in
time. In other words, if x is the input and y is the output, y(t) totally depends on
x(t) for every t.

KFUPM-CIE Department-SCE
507
Examples

KFUPM-CIE Department-SCE
507
Causal/Noncausal

KFUPM-CIE Department-SCE
507
Causal/Noncausal

KFUPM-CIE Department-SCE
507
Causal/Noncausal

KFUPM-CIE Department-SCE
507
 Determine whether the
systems

a. y(n)=0.5x(n)+2.5x(n−2), for n≥0

b.
y(n)=0.25x(n−1)+0.5x(n+1)−0.4y(n−1), for n≥0

KFUPM-CIE Department-SCE
507
Time Invariant/Time Varying

KFUPM-CIE Department-SCE
507
 Invertible and non-invertible
systems
• Let a system 𝑆 produce an output 𝑦(𝑡) with input 𝑥(𝑡). If there exists another system 𝑆𝑖 ,
which produces 𝑥(𝑡) from 𝑦(𝑡), then 𝑆 is an invertible system.
• It is essential that there is a one-to-one mapping between input and output so that a system
is invertible.
• For example if 𝑆 is an amplifier with gain 𝐺, it is invertible and 𝑆𝑖 is an attenuator with gain
1/𝐺.
• Apply 𝑆𝑖 following 𝑆 gives a combined which has no impact in the input (i.e., the input 𝑥(𝑡) is
not changed).

KFUPM-CIE Department-SCE
507
Stable and unstable systems
• Externally stable systems are the ones in
which a bounded input results in a
bounded output (the system is said to be
stable in the BIBO sense).
• Stability of a system will be discussed after
introducing Fourier and Laplace transforms.
• More detailed analysis of stability is
covered in the second year Control course.

KFUPM-CIE Department-SCE
507
Controllability and Observability
• Controllability and Observability are two
fundamental concepts in control theory
that determine the ability to control and
observe the internal states of a system.
Both concepts are essential for designing
effective control systems and for
understanding how a system behaves.
Controllability
Observability
State-Space Design
A General State-Space Model
 Internal variable
Current through R, C, L
Voltage through R, L, C

𝑥1 𝑥2

States are a minimal set of information, so we have

Enough information in x1, x2,..xn to expressed everything
 State of a system
• The initial state of a systems 𝑥0 at time 𝑡0
is information at time 𝑡0 that together with
a control signal ( u(t): t≥ 𝑡0 ) and the
system model determine uniquely the
system behavior for t≥ 𝑡0 .
 𝑢(𝑡)

𝑡0 , 𝑥0 model
 𝑦(𝑡)

𝑡0

The initial State completely summarizes the past.

From State-Space to Transfer
Function