Scribd
Upload
28 views

2nintro FuzzyLogic PDF

The document introduces fuzzy logic and its applications. It discusses (1) using fuzzy logic to control an inverted pendulum similar to how humans estimate movement, (2) representing values as fuzzy sets rather than binary true/false, (3) fuzzy rule-based systems and the inference process, and (4) provides an example of balancing an inverted pendulum using fuzzy logic.

Uploaded by

pkrsuresh2013
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
28 views

2nintro FuzzyLogic PDF

The document introduces fuzzy logic and its applications. It discusses (1) using fuzzy logic to control an inverted pendulum similar to how humans estimate movement, (2) representing values as fuzzy sets rather than binary true/false, (3) fuzzy rule-based systems and the inference process, and (4) provides an example of balancing an inverted pendulum using fuzzy logic.

Uploaded by

pkrsuresh2013
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 21

Fuzzy Logic

Introduction
Fuzzy Sets
Logic
Example
References and . . .

Peter Molnar
c
2003
JJ

II

Page 1 of 21

First Prev Next Last Go Back Full Screen Close Quit

1. Introduction
1.1. Motivation: The classical engineering problem of balancing an inverted pendulum
A solid pendulum is hinged at
its base to a platform which
can move in opposite directions.

Introduction
Fuzzy Sets
Logic

Pendulum can move in the


same plane as the platform.

Example
References and . . .

JJ

II

Inverted Pendulum

The objective is to keep the


pendulum upright by compensating for the tilt of the pendulum by corresponding movements of the platform.

Page 2 of 21

First Prev Next Last Go Back Full Screen Close Quit

1.2. Practical Application

Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 3 of 21

First Prev Next Last Go Back Full Screen Close Quit

1.3. Classical Controller


Input Measurements:
the angle the pendulum makes with the upright,
the angular velocity of the pendulum must be measured,
the rate of change of this velocity.
Introduction

Output:

Fuzzy Sets
Logic

the direction, velocity, and

Example
References and . . .

change in velocity of the platform.


Find a suitable relation between these variables. Can become very
complicated (Newtonian mechanics)
Demand a lot of computing power.

JJ

II

Page 4 of 21

First Prev Next Last Go Back Full Screen Close Quit

1.4. Human Controller


When the pendulum tilts we measure the nature of this movement.
How much the broomstick has moved, what direction it is moving in
and how quickly it is moving?
We automatically make the corresponding movement to compensate.
We dont actually quantify these factors, rather we make quick, instinctive estimations.

Introduction
Fuzzy Sets
Logic
Example
References and . . .

If we were to examine the thought pattern in such a situation it


might read like this:
The pendulum is tilting, to the right, a little , so I must move
my hand, to the right, a little.
The pendulum is tilting, quickly, forward, a lot, so I must move
my hand, quickly, forwards, a lot.

JJ

II

Page 5 of 21

First Prev Next Last Go Back Full Screen Close Quit

1.5. Sufficient Accurate Estimates


And so on. The key thing here is our ability to make sufficiently accurate estimations about the nature of the situation, build a number
of rules with them and act on these rules accordingly.
We find it easy to do this. We make use of abstract concepts such
as a little, a lot, quickly, slowly and apply them to the variables of
pendulum tilt and hand movement.

Introduction
Fuzzy Sets

Fuzzy logic provides a means by which computers can imitate the


way we humans make these kinds of estimations

Logic
Example
References and . . .

rather than A or NOT A,


we can say MOSTLY A or SLIGHTLY A.

JJ

II

Page 6 of 21

First Prev Next Last Go Back Full Screen Close Quit

2. Fuzzy Sets

Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Fuzzy Sets - a matter of degree

Page 7 of 21

First Prev Next Last Go Back Full Screen Close Quit

Introduction
Fuzzy Sets
Logic
Example
References and . . .

Fuzzy truth, T , likely hood of a predicate to be true, given a crisp


input value.
Degree of membership, (x), of a given crisp input value to a fuzzy
set.
JJ

II

Page 8 of 21

First Prev Next Last Go Back Full Screen Close Quit

2.1. Representation of Knowledge: Sets


Fuzzy Sets: define attributes
Height: TALL, AVERAGE HIGHT, SHORT
Build: FAT, MEDIUM, SLIM
Weight, HEAVY, MEDIUM, LIGHT
Introduction

Membership

Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 9 of 21

First Prev Next Last Go Back Full Screen Close Quit

2.2. Representation of Knowledge: Rules


Linguistic rules:
if a man is tall and fat, then he will be heavy in weight.
if a man is tall and slim, then he will be average in weight.
if a man is tall and of medium build, then he will be heavy in
weight.

Introduction
Fuzzy Sets

if a man is short and fat, then he will be average in weight. if


a man is short and slim, then he will be light in weight.

Logic
Example
References and . . .

if a man is short and of medium build, then he will be light in


weight.
if a man is of average height and fat, then he will be heavy in
weight.
if a man is of average height and slim, then he will be light in
weight.

JJ

II

if a man is of average height and of medium build, then he will


be medium in weight.

Page 10 of 21

First Prev Next Last Go Back Full Screen Close Quit

3. Logic
3.1. Fuzzy Rule Based Systems

Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 11 of 21

First Prev Next Last Go Back Full Screen Close Quit

3.2. Fuzzification
Process by which crisp, non-fuzzy, input values are converted into
their fuzzy representations.
Example: a crisp input value of 63 for height, fuzzification of this
input entails applying this value to the fuzzy set average.
Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

The fuzzification process will take place for all inputs in all corresponding fuzzy sets, yielding fuzzified membership functions for use
in each rule.

Page 12 of 21

First Prev Next Last Go Back Full Screen Close Quit

3.3. Inference in Fuzzy Logic


Min-Max inference (Lotfi Zadeh)
if A is X and B is Y, then C is min(X,Y)
if A is X or B is Y, then C is max(X,Y)
Example X=0.75, Y=0.25
Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 13 of 21

First Prev Next Last Go Back Full Screen Close Quit

Logic Operators

Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 14 of 21

First Prev Next Last Go Back Full Screen Close Quit

Compute all rules that apply, i.e. for any set that any input value is
a member of.
Combine results: unify sets.
Defuzzify result: calculating a crisp output value.
Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 15 of 21

First Prev Next Last Go Back Full Screen Close Quit

3.4. Defuzzification
The opposite of fuzzification; entails the rationalization of a fuzzified
output to obtain a crisp value for the output.
Several methods can be used, one of them: the Center of Gravity
method.
Finding the center of gravity of the fuzzified output membership
function and returning the crisp value that corresponds to this point.

Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 16 of 21

First Prev Next Last Go Back Full Screen Close Quit

4. Example
Inverted Pendulum

Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 17 of 21

First Prev Next Last Go Back Full Screen Close Quit

Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 18 of 21

First Prev Next Last Go Back Full Screen Close Quit

Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 19 of 21

First Prev Next Last Go Back Full Screen Close Quit

5. References and Further Reading

Introduction
Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 20 of 21

First Prev Next Last Go Back Full Screen Close Quit

Disclaimer
This presentation material has been created for class presentation
only.
This presentation is intended solely for students at Clark Atlanta
University.
Introduction

All rights reserved, 20002003.

Fuzzy Sets
Logic
Example
References and . . .

JJ

II

Page 21 of 21

First Prev Next Last Go Back Full Screen Close Quit

You might also like