Chapter 11 Fuzzy Control Theory

You might also like

Download as ppt, pdf, or txt
Download as ppt, pdf, or txt
You are on page 1of 63

Chapter 11

Fuzzy Control
Theory
• 1 Reference
• 2 Introduction
• 3 Fuzzy sets
• 4 Universe
• 5 Membership function
• 6 Singletons
• 7 Linguistic variables
• 8 Tank level example
• 9 Operations on fuzzy sets
• 10 Fuzzy logic
Chapter 1.1

References
1. Zi-Xing Cai. Intelligent Control: Principles,Techniques and Applica
tions. World Scientific, Singapore: 1997
2. Panos J. Antsaklis and Kevin M. Passino. An Introduction to Intelli
gent and Autonomous Control. Available:
http://www.nd.edu/~pantsakl/book1/Download/downloadpage.htm
3. Neural Networks, Available:
http://www.willamette.edu/~gorr/classes/cs449/intro.html
4. Fuzzy Logic, Available: http://www.control.hut.fi/Kurssit/AS-74.11
5/Material/
5. Martin T. Haguan 等著,戴葵等译 . 神经网络设计 . 机械工业出
版社,北京: 2002.9
6. 何玉彬 李新忠著 . 神经网络控制技术及其应用 . 科学出版社,
北京: 2000.11
7. 孙增圻等 . 智能控制理论与技术 . 清华大学出版社 , 广西科学
技术出版社 , 北京: 1997.4
Chapter 1.2

Introduction
Why Use Fuzzy Logic?
• Fuzzy logic is conceptually easy to understand.
• Fuzzy logic is flexible.
• Fuzzy logic is tolerant of imprecise data.
• Fuzzy logic can model nonlinear functions of arb
itrary complexity.
• Fuzzy logic can be built on top of the experience
of experts.
• Fuzzy logic can be blended with conventional co
ntrol techniques. ( PID )
• Fuzzy logic is based on natural language.
About PID
• PI?PD?PID?
• Model(Multi-Parameter?) basalt( 玄武
岩)
• BW , Orders , Linear , non-linear
(G(s)?) , impedance(Topology)
• =>Topology , Modulation(PWM , P
FM , PAM...),Control
In order to increase control performance, the
complexity of the control systems has to b
e increased too. As a result, the reliability
of the control would be decreased.
为了克服传统控制理论的局限性,产生了模
拟人类思维和活动的智能控制。
Advancement
Intelligent c
ontrol
Learning co
ntrol
Adaptive and r
obust control
Stochastic con
trol
Optimal contr
ol
Feedback con
trol
Open-loop con
trol Sophistication in Control
• The objective is to cont
rol the valve V1, such t
hat the tank is refilled w
hen the level is as low
as LL, and stop the refil
ling when the level is a
s high as LH.
The sensor LL is 1 when the level is above the mark, and 0
when the level is below; likewise with the sensor LH. The val
ve opens when V1 is set to 1, and it closes when V1 is set to
0. In two-valued (Boolean) logic the controller can be describ
ed:

1 if LL switches from 1 to 0 Suitabl


V1   e for P

0 if LH switches from 0 to 1 LC

An operator, whose responsibility is to open and close the val


ve, would perhaps describe the control strategy as:

If the Level is low then open V1 Suitable for a


fuzzy logic co
If the Level is high then close V1 ntroller
Fuzzy Set: 1965 … Fuzzy Logic: 1973 E.H.M
AMDANI … BISC: 1990 Takagi… Human-Ma
chine Perception: 2000
http://www.cs.berkeley.edu/~zadeh/

Lotfi Zadeh, the father of fuzzy logic, clai


med that many sets in the world that surr
ounds us are defined by a non-distinct bo
undary. Indeed, the set of high mountains,
or, the set of low level of measurements

Zadeh [0,1]
{0,1}

introducing a gradual transition from falsehood to truth.


Chapter 1.3

Fuzzy sets
The terms set, collection and class are synonyms.

• Conventional Sets

A={0, 1, 2, 3} a finite set


x>10 a infinite set
• Fuzzy Sets
Example 1:
The set of young people
Example 2:
A weather report regarding high temperatures, strong
winds, or nice days.
标准
In other cases a criterion appears nonfuzzy, but is pe
rceived as fuzzy: a speed limit of 60 kilometers per
hour, a check-out time at 12 noon in a hotel, a 50 y
ears old man.
Zadeh proposed a grade of membership, such that th
e transition from membership to non-membership i
s gradual rather than abrupt.
脆的 , 易碎的 进退两难的局面 , 困难的选择
Chapter 1.4

Universe
• Elements of a fuzzy set are taken from a universe
of discourse , or universe for short. The universe
contains all elements that can come into consider
ation. Even the universe depends on the context,
as the next example shows.
Chapter 1.5

Membership functio
n
Every element in the universe of discourse is a memb
er of the fuzzy set to some grade, maybe even zero.
The set of elements that have a non-zero membershi
p is called the support of the fuzzy set. The function t
hat ties a number to each element x of the universe is
called the membership function μ(x).

Continuous and Discrete Representation


Difference of two sigmoid; Two-sided Gaussian curve; Gaussian curve; Generalized bell
curve ;Pi-shaped curve membership function. Product of two sigmoidal ; smfS-shaped c
urve . Sigmoid curve. Trapezoidal; Triangular; Z-shaped curve
Calculation S, Z
• function [s,z]=cal_sz(xl,xr)
• x=0:1:100;
• s=x;
• z=x;
• pai=x;

• for i=0:100,
• if i<=xl
• s(i+1)=0;
• z(i+1)=1;
• elseif i>=xr
• s(i+1)=1;
• z(i+1)=0;
• else
• s(i+1)=0.5+0.5*cos((i-xr)*pi/(xr-xl));
• z(i+1)=0.5+0.5*cos((i-xl)*pi/(xr-xl));
• end;
• end;
Display S, Z, 
• x=0:1:100;
• temp=x;
• [s,temp]=cal_sz(10,90);
• [temp,z]=cal_sz(10,50);

• for i=0:100,
• pai(i+1)=min(s(i+1),z(i+1));
• end;

• plot(x,s,'*',x,z,'H',x,pai,'.')
S(10,90,x)
S(10,90,x), Z(100,100,x), (10,90,1
00,100,x)
S(10,90,x), Z(10,50,x), (10,90,10,
100,x)
Normalization: A fuzzy set is normalized if its largest membership
value equals 1. You normalize by dividing each membership value
by the largest membership in the set, a/max(a)
Chapter 1.6

Singletons
Chapter 1.7

Linguistic variables
Chapter 1.8

Tank level example


Chapter 1.9

Operations on fuz
zy sets
Set operations
Modifiers

或多或少
涉及 Relations Between Sets

笛卡尔
正交
Multiplication->intersection; summation -> union
*Set

0.8*0.5+0.9*0.6 (0.8∩0.5)∪(0.9∩0.6)

0.5∪0.6 0.6
Chapter 1.10

Fuzzy logic
命题
推理
Connectives

You might also like