Adama Science andTechnology

University

EPCE6202 Intelligent Control

and Its Applications
(Introduction)

ASTU
School of EEC

1. Lecturer
Name: Prof. Gang Gyoo Jin, PhD
Affiliation: Dept. of EPCE
Academic background:
1976 Korea Maritime Univ. (BE)
1987 Florida Inst. of Tech (MSc)
2006 Cardiff University (PhD)
Major:
- Intelligent control(Fuzzy, Neural network)
- Optimization
- Applications of genetic algorithms and PSO
- Fractal technique
Office: Building No 605, Room #1

1
 2. Syllabus
Course title: EPCE6202 Intelligent Control and Its
Applications
Course hours: Lecture(2hr), Lab(2hr)
Lecture room and time: International
Time
 Lecture Monday 08:00~09:50AM
 PC Lab Wednesday 14:00~ 15:50PM
Announcement:
 Telegram: IntelligentControl
Email and phone:
 Email: gjin30@gmail.com
 Phone: 093 212 1300

3. Course Objective
Learning expectations include:
• To introduce fuzzy set, fuzzy logic, approximate
reasoning and to learn how to code the program for
fuzzy modules
• To learn how to design fuzzy logic controllers and to
get fuzzy model of dynamic systems
• To learn how to design and train NNs and Adaptive
Neuro-Fuzzy Inference System (ANFIS)
• To learn how to apply NNs and ANFIS to
classification, modeling and control
• To introduce GA and PSO and apply them to tuning
of controllers

2
 4. Weekly schedule
The course will be sectioned into parts:
• Section 01- Introduction to intelligent control
• Section 02- Fuzzy set theory
• Section 03- Fuzzy relation
• Section 04- Fuzzy logic(compound and implication)
• Section 05- Approximate reasoning(Mamdani and
Sugeno type)
• Section 06- Applications to fuzzy modeling
• Section 07- Applications to Fuzzy PID controller
design

Contd
• Section 08- Application of Fuzzy Toolbox to Simulink
model
• Section 09- Introduction to GA and PSO
• Section 10- BP neural networks(NNs) design
• Section 11- BP NNs learning
• Section 12- Applications of NNs to classification
• Section 13- Applications of NNs to data fitting

3
 Contd
• Section 14- Applications of ANFIS system to control
systems
• Section 15- Applications of ANFIS system to dynamic
system modeling

5. Text book and References

Text book and references
• Jinkun Liu, Intelligent Control Design and
MATLAB Simulation, Tsinghua University Press,
Beijing and Springer Nature Singapore Pte Ltd,
2018.
• Guanrong Chen and Trung T. Pham, Introduction to
Fuzzy Sets, Fuzzy Logic, and Fuzzy Control
Systems, CRC Press, Inc., N.Y., 2001.
• Clarence W. de Silva, Intelligent Control: Fuzzy
Logic Applications, CRC Press, Inc., N.Y., 1995.

4
 Contd
Text book and references
• Math Works, Neural Network Toolbox For Use
with MATLAB, Version 4, 2004.
• J.-S. Roger Jang and Ned Gulley, Fuzzy Logic
Toolbox For Use with MATLAB, Version 1, Math
Works, 1997.
• C. C. Lee, “Fuzzy logic in control systems: fuzzy
logic controller I,” IEEE Transactions on Systems,
Man, and Cybernetics, vol. 20, no. 2, 1990.
• C. C. Lee, “Fuzzy logic in control systems: fuzzy
logic controller II,” IEEE Transactions on Systems,
Man, and Cybernetics, vol. 20, no. 2, 1990.

6. Evaluation
The course will be graded on the basis of

• Midterm exam 40%

• Final exam 40%
• Quiz / project 10%
• Class attendance 10%

Total 100%

5
 7. Intelligent Approach
Definition of Intelligence
• When we say a person, animal or thing is
intelligent, what does intelligence mean ?
• In fact, everyone has a different concept based on
their own experiences and views. This can easily
change – what is intelligent at one time and it may
not be later or elsewhere.
• Intelligence is defined as “The exercise of
understanding; intellectual power; acquired
knowledge; quickness of intellect” (New English
Dictionary, 1932)

Contd
• Intelligence is the ability to reason and to profit by
experience. (Macmillan Encyclopedia, 1995)
• In order to think, someone or something has to have
a brain or computing devices, sensors and actuators
(e.g. muscles) that enable them to learn and
understand things, to solve problems and to make
decisions. So we can define intelligence as the
ability:
 to learn and understand,
 to solve problems, and
 to make appropriate decisions.

6
 Contd
Traditional Approaches
• Many traditional approaches have been developed for
mathematically well-defined problems with accurate
models.
• Conventional control uses theories and the methods,
that have developed in the past decades, of which is
primarily described by differential equations.
• PID control, Optimal Control, Adaptive Control, etc.

Contd
Intelligent Approaches

• Currently systems are more and more complex and

uncertain and could not be solved by traditional
methods.
• Conventional control may lack in
autonomy and decisionmaking
ability and hence cannot provide
adequate solutions under
uncertain and fuzzy
environments.

7
 Contd
• Advances in artificial intelligence (AI) and related
scientific fields have brought new opportunities and
challenges for complex problems with uncertainties.
• Intelligent systems defined with such attributes as
high degree of autonomy, reasoning under
uncertainty, higher performance in a goal seeking
manner, data fusion from a multitude of sensors,
learning and adaptation in an environment, etc
represent a new approach to addressing those
complex problems.

Contd
• In the real world, there are many systems that are
too complex or uncertain to represent them in
complete mathematical models. But we need to
design, control, or optimize the behavior of such
systems. For examples,
Human
• Humans can learn information
through their cognitive process,
make decisions, and even adapt
to the environment without
using mathematical models.

8
 Contd
NASA Rover
• The distance from Earth to Mars can be anywhere
from 54.6 million km to 401 million km.
• The shortest spacecraft travel will take about 7 months
(about 39,600 km/hr). The travel of a radio signal to
Mars will take about 5 to 20min.

Contd
• Development of a Mars rover
requires autonomy to solve its
own problems without NASA’s
help for some time due to
communication delays,
• The rover must be able to land and move across the
Mar’s surface safely. The navigation system takes
pictures of the nearby terrain using stereo camera and
generate 3D terrain maps. Traversability and safety is
then determined from the height and density of rocks
or steps, excessive tilts and roughness of the terrain.

9
 Contd
Driverless Car
• Development of a driverless car requires millions of
images and thousands of hours of video from many
sensors.
• In order to deal with large amounts of data,
substantial computing power and High-performance
GPUs are required.

Contd
Conventional vs Intelligent Control
• Intelligent control attempts to enhance conventional
control to solve new challenging control problems.
• It uses conventional control methods to solve
"lower level" control problems. That is,
conventional control is included in the area of
intelligent control.
• Intelligent control means some form of control
using fuzzy, neural network and/or optimization
techniques. However Intelligent control does not
restrict only to those techniques.

10
 Contd
There are three fundamental characteristics in
intelligent control systems.

Adaptation and Learning

• The ability to adapt to changing conditions is

necessary in an intelligent system.
• The ability to learn, for systems to be able to adapt
to a variety of unexpected changes is essential.

Contd
Autonomy and Intelligence
• Autonomy in achieving goals is an important
characteristic of intelligent control systems. When a
system has the ability to act appropriately in an
uncertain environment for periods of time without
external intervention, it is considered to be highly
autonomous. An adaptive controller is higher
autonomous than a fixed controller.
• Although for low autonomy no intelligence (or
"low" intelligence) is necessary, for high degrees of
autonomy, intelligence in the system is essential.

11
 Contd
Structures and Hierarchies
• In order to cope with complexity, an intelligent
system must have an appropriate functional
structure for effcient analysis and evaluation of
control strategies.
• Hierarchies (that may be localized or combined in
heterarchies) that may serve for such structures to
cope with complexity.

Contd
Development of Intelligent Systems
• There are three major components for developing
intelligent systems: knowledge acquisition and
representation, inferencing, and decision-making
processes.
• Three possible types of knowledge or information
representation describing the input–output or cause–
effect relationships of nonlinear systems are
quantitative analytical models, empirical data, and
heuristic rules as

12
 Contd
Three different knowledge representation

Input Nonlinear Systems Output

or Problems

Heuristic rules
Input Output
Analytical models
Domain Domain
Empirical data

13