Вы находитесь на странице: 1из 47

Advance Topics in Mathematical Methods ME7100

Fuzzy sets,System and


Modelling
Advance Topics in Mathematical Methods ME7100
Fuzzy sets were introduced by Zadeh in 1965 to represent/manipulate
data and information possessing nonstatistical uncertainties.
L.A.Zadeh, Fuzzy Sets, Information and Control, 8(1965) 338-353.


There are two main characteristics of fuzzy systems that give them
better performance for specific applications.

Fuzzy systems are suitable for uncertain or approximate reasoning,
especially for the system with a mathematical model that is difficult
to derive.
Fuzzy logic allows decision making with estimated values under
incomplete or uncertain information.
Advance Topics in Mathematical Methods ME7100
Classical sets or crisp set
A = {12, 24, 36, 48, }
Notation: A = {x | x = 12n, n is a natural number}

A = {cities adjoining Hyderabad}






A = {Mahbubnagar, Medak, Nalgonda, Rangareddy}

e
e
= X
A x
A x
x
A
if 0
if 1
) (
Advance Topics in Mathematical Methods ME7100
Classical sets vs. Fuzzy set
A = {cities near from Hyderabad}
A = {Hyderabad, Adilabad, Khammam, Karimnagar, Mahbubnagar,
Medak, Nalgonda, Nizamabad, Rangareddy, Warangal, Bidar,
Gulbarga,.., Mumbai, Pune, Bangalore,., Delhi,
Islamabad, Kabul, Kathmandu, Singapore. Rome, London,
Paris.}
Is the above information precise? : No it is Fuzzy not clear,
distinct, or precise; blurred

Definition of fuzzy logic : A form of knowledge representation suitable for
notions that cannot be defined precisely, but which depend upon their
contexts.

Advance Topics in Mathematical Methods ME7100
Fuzzy set : example 1
A = {cities near from Hyderabad}
What is near: A distance less the D kilometer
D = 100; if we are interested in adjoining cities
D = 200/300/400; if we are interested in cities in AP
D = 300/400/500..; if we are interested in cities in AP or Karnataka
D= 3000/4000/if we are interested in cities in India AP or Karnataka

Information is not precise and is context based




Advance Topics in Mathematical Methods ME7100
Fuzzy set :: example 1
A = {cities near from Hyderabad}
What is not near (far): A distance more the D kilometer

Say, D = 200; if we are interested in cities in AP
Lets make some rules:
1. If 0 D 100; very near ( depending upon road condition and traffic one can reach within 2 hr)
2. If 50< D 150; near ( depending upon road condition and traffic one can reach within 3hrs)
3. If 100< D 200; far ( depending upon road condition and traffic u can reach within 6hrs)
4. If D> 200; very far





Advance Topics in Mathematical Methods ME7100
Fuzzy set :: example 1
A = {cities near from Hyderabad}
What is not near (far): A distance more the D kilometer
Say, D = 200; if we are interested in cities in AP
Lets make some rules:
1. If 0 D 100; very near ( depending upon road condition and traffic one can reach within 2 hr)





100 200 300
0
1
Advance Topics in Mathematical Methods ME7100
Fuzzy set :: example 1
A = {cities near from Hyderabad}
What is not near (far): A distance more the D kilometer
Say, D = 200; if we are interested in cities in AP
Lets make some rules:
2. If 50< D 150; near ( depending upon road condition and traffic one can reach within 3hrs)





100 200 300
0
1
Advance Topics in Mathematical Methods ME7100
Fuzzy set :: example 1
A = {cities near from Hyderabad}
What is not near (far): A distance more the D kilometer
Say, D = 200; if we are interested in cities in AP
Lets make some rules:
3. If 100< D 200; far ( depending upon road condition and traffic u can reach within 6hrs)





100 200 300
0
1
Advance Topics in Mathematical Methods ME7100
Fuzzy set :: example 1
A = {cities near from Hyderabad}
What is not near (far): A distance more the D kilometer
Say, D = 200; if we are interested in cities in AP
Lets make some rules:
4. If D> 200; very far





100 200 300
0
1
Advance Topics in Mathematical Methods ME7100
Fuzzy set :: example 1
A = {cities near from Hyderabad}
What is not near (far): A distance more the D kilometer

Say, D = 200; if we are interested in cities in AP
Lets make some rules:
1. If 0 D 100; very near ( depending upon road condition and traffic one can reach within 2 hr)
2. If 50< D 150; near ( depending upon road condition and traffic one can reach within 3hrs)
3. If 100< D 200; far ( depending upon road condition and traffic u can reach within 6hrs)
4. If D> 200; very far





100 200 300
0
1
100 200 300
0
1
100 200 300
0
1
100 200 300
0
1
Advance Topics in Mathematical Methods ME7100
Fuzzy set :
100 200 300
0
1
100 200 300
0
1
100 200 300
0
1
100 200 300
0
1
Distance (x)
M
e
m
b
e
r
s
h
i
p

(
x
)

Advance Topics in Mathematical Methods ME7100
Fuzzy set :
100 200 300
0
1
100 200 300
0
1
100 200 300
0
1
100 200 300
0
1
Distance (x)
M
e
m
b
e
r
s
h
i
p

(
x
)

Advance Topics in Mathematical Methods ME7100
Fuzzy set :
100 200 300
0
1
100 200 300
0
1
100 200 300
0
1
100 200 300
0
1
Distance (x)
M
e
m
b
e
r
s
h
i
p

(
x
)

Imp. Note: + sign stands for the union of membership grades; /
stands for a marker and does not imply division.
Advance Topics in Mathematical Methods ME7100
Fuzzy set :
100 200 300
0.25
1
100 200 300
1
100 200 300
0.75
1
X=75
0
1
Distance (x)
M
e
m
b
e
r
s
h
i
p

(
x
)

A(x=75) = 0.25/very near + 0.75/near + 0.0 far + 0.0/very far
A(x=300) = 0.00/very near + 0.00/near + 0.0 far + 1.0/very far
Advance Topics in Mathematical Methods ME7100
Fuzzy set : Example 2
150 210 170 180 190 200 160
Height, cm
Degreeof
Membership
150 210 180 190 200
1.0
0.0
0.2
0.4
0.6
0.8
160
Degreeof
Membership
Short Average Tall
170
1.0
0.0
0.2
0.4
0.6
0.8
Fuzzy Sets
CrispSets
Short Average Tall
Negnevitsky, Pearson Education, 2005
Advance Topics in Mathematical Methods ME7100
Fuzzy set :
A fuzzy set A can be denoted as

If x is discrete


If x is continuous



Example 2. The membership function of the fuzzy set of real numbers close
to 1, is can be defined as

A x x
A
x X
i i
i
=
e

( ) /
A x x
A
X
=
}
( ) /
Advance Topics in Mathematical Methods ME7100
Triangular Fuzzy Number :
Advance Topics in Mathematical Methods ME7100
Trapezoidal Fuzzy Number :
Advance Topics in Mathematical Methods ME7100
Operations on crisp sets

Intersection Union
Complement
Not A
A
Containment
AA
B
B A AA B
Negnevitsky, Pearson Education, 2005
Advance Topics in Mathematical Methods ME7100
Operations on fuzzy sets

Intersection: The intersection of A and B is defined
as
Advance Topics in Mathematical Methods ME7100
Operations on fuzzy sets

Intersection: The union of A and B is defined as
Advance Topics in Mathematical Methods ME7100
Operations on fuzzy sets

Negnevitsky, Pearson Education, 2005
Complement
0
x
1
( x)
0
x
1
Containment
0
x
1
0
x
1
AB
Not A
A
Intersection
0
x
1
0
x
AB
Union
0
1
AB
AB
0
x
1
0
x
1
B
A
B
A
( x)
( x) ( x)
Advance Topics in Mathematical Methods ME7100
Operations on fuzzy sets

Negnevitsky, Pearson Education, 2005
Complement
0
x
1
( x)
0
x
1
Containment
0
x
1
0
x
1
AB
Not A
A
Intersection
0
x
1
0
x
AB
Union
0
1
AB
AB
0
x
1
0
x
1
B
A
B
A
( x)
( x) ( x)
Advance Topics in Mathematical Methods ME7100
Operations on fuzzy sets

Important: Operations on fuzzy sets are also fuzzy, i.e., union
of intersection may be treated different ( different operators)

Eg. min-max operator

Union ( fuzzy OR)


)] x ( ), x ( [ max ) x (
B A B A
=
+
Advance Topics in Mathematical Methods ME7100
Operations on fuzzy sets

Important: Operations on fuzzy sets are also fuzzy, i.e., union
of intersection may be treated different ( different operators)

Eg. Min and max operator

Intersection ( fuzzy AND)


)] x ( ), x ( [ min ) x (
B A B A
=
-
Advance Topics in Mathematical Methods ME7100
Operations on fuzzy sets

Important: Operations on fuzzy sets are also fuzzy, i.e., union
of intersection may be treated different ( different operators)

Eg. min-max operator

Min-max : intersection distributive over union


) ( ) (
) ( ) ( ) (
x x
C A B A C B A + - + + -
=
min[ A, max(B,C) ]=min[ max(A,B), max(A,C) ]
Advance Topics in Mathematical Methods ME7100
Operations on fuzzy sets

Important: Operations on fuzzy sets are also fuzzy, i.e., union
of intersection may be treated different ( different operators)

Eg. Min, max and min-max operator

Min-max : union distributive over intersection

) ( ) (
) ( ) ( ) (
x x
C A B A C B A - + - - +
=
max[ A,min(B,C) ]= max[ min(A,B), min(A,C) ]
Advance Topics in Mathematical Methods ME7100
Operations on fuzzy sets

Important: Operations on fuzzy sets are also fuzzy, i.e., union
of intersection may be treated different ( different operators)

Eg. sum-product operator


) ( ) ( ) (
A
x x x
B B A
=
-
] 1 ), ( ) ( [ min ) (
A
x x x
B B A
+ =
+
Advance Topics in Mathematical Methods ME7100
Fuzzy System
Input
Fuzzifier
Inference
Engine
Defuzzifier
Output
Fuzzy Knowledge base
a) Fuzzification information
b) Fuzzy Rules
c) Inference method
d) Defuzzification method

Advance Topics in Mathematical Methods ME7100
Fuzzy System
Input
Fuzzifier
Inference
Engine
Defuzzifier
Output
Fuzzy
Knowledge base
Fuzzier : Crisp input to a linguistic variable using
the membership functions
Eg: If 0 D 100; very near
Advance Topics in Mathematical Methods ME7100
Fuzzy System
Input Fuzzifier
Inference
Engine
Defuzzifier Output
Fuzzy
Knowledge base
Fuzzy Rule: There are many Fuzzy rules

e.g. Zadeh Mamdani rule (( commonly known as Mamdani Rule)
If x is A and y is B then z


if D is near and T is moderate then customers = more
if D is very near and T is high then customers = reasonable
if D is near and T is high then customers = reasonable
if D is V near and T is high then customers = less
if D is far or very far then customers = less
If D =75 and T =21 how many customers will be there ???????

Advance Topics in Mathematical Methods ME7100
Fuzzy System
Input Fuzzifier
Inference
Engine
Defuzzifier Output
Fuzzy
Knowledge base
Fuzzy Rule: There are many Fuzzy rules


e.g. Takagi, Sugeno & Kang( commonly known as Sugeno Rule)
If x
i
is A
i
and y
i
is B
i
then z
i
=f(x
i
,y
i
)
(f(x
i
,y
i
) is a polynomial of x and y)

if D is near and T is moderate z = 3D+2T +1
if X is very small and Y is very large then z = xy
Advance Topics in Mathematical Methods ME7100
Fuzzy System
Input Fuzzifier
Inference
Engine
Defuzzifier Output
Fuzzy
Knowledge base
Fuzzy Rule: Sugeno Rule

if D is near and T is moderate z = 3D+2T +1
if X is very small and Y is very large then z = xy
Advance Topics in Mathematical Methods ME7100
Fuzzy System
Input Fuzzifier
Inference
Engine
Defuzzifier Output
Fuzzy
Knowledge base
Inference: converts the fuzzy input to the fuzzy output, i.e., computation
of output fuzzy membership grade and their aggregation
Advance Topics in Mathematical Methods ME7100
Fuzzy System
Input Fuzzifier
Inference
Engine
Defuzzifier Output
Fuzzy
Knowledge base
Defuzzifier: converts the fuzzy output to crisp output

Centroid of area (COA)
Bisector of area (BOA)
Mean of maximum (MOM)
Smallest of maximum (SOM)
Largest of maximum (LOM)

Advance Topics in Mathematical Methods ME7100
Fuzzy System
Input Fuzzifier
Inference
Engine
Defuzzifier Output
Fuzzy
Knowledge base
Defuzzifier: converts the fuzzy output to crisp output

Centroid of area (COA)
Bisector of area (BOA)
Mean of maximum (MOM)
Smallest of maximum (SOM)
Largest of maximum (LOM)

( )
,
( )
A
Z
COA
A
Z
z zdz
z
z dz

=
}
}
*
,
{ ; ( ) }
Z
MOM
Z
A
zdz
z
dz
where Z z z
'
'
=
' = =
}
}
( ) ( ) ,
BOA
BOA
z
A A
z
z dz z dz
|
o
=
} }
Used in Mamdani
FIS
Advance Topics in Mathematical Methods ME7100
Fuzzy System
Input Fuzzifier
Inference
Engine
Defuzzifier Output
Fuzzy
Knowledge base
Defuzzifier: converts the fuzzy output to crisp output

Weighted sum
Weighted Average

Used in Mamdani
Sugeno FIS
Advance Topics in Mathematical Methods ME7100
Fuzzy System
Mamdani FIS with Max-min inference
R1:If x =A1, y=B1 then Z=C1
R2: If x =A2, y=B2 then Z=C2
Advance Topics in Mathematical Methods ME7100
Fuzzy System
Mamdani FIS with Max-Product inference
R1:If x =A1, y=B1 then Z=C1
R2: If x =A2, y=B2 then Z=C2
Advance Topics in Mathematical Methods ME7100

Advantages of the Sugeno Method

It is computationally efficient.

It works well with linear techniques (e.g., PID control).

It works well with optimization and adaptive techniques.

It has guaranteed continuity of the output surface.

It is well suited to mathematical analysis.


Advantages of the Mamdani Method


It is intuitive.

It has widespread acceptance.

It is well suited to human input.
Advance Topics in Mathematical Methods ME7100
Fuzzy system example (Mamdani)
(1) fuzzification
Fuzzy Distance Fuzzy temperature Fuzzy Visitor
If D =75 and temperature =25 how many customers are accepted?
Advance Topics in Mathematical Methods ME7100
Fuzzy system example (Mamdani)
(2) Rule
R1 : If Distance is near and temperature is pleasant then sufficient visitors
R2 : If Distance is very far or temperature is harsh then no visitors
R3 : If Distance is near and temperature is bad then less visitors
R4 : If Distance is far and temperature is reasonable then less visitors
R5 : If Distance is very near and temperature is pleasant then more visitors




R1



R2






R2
Advance Topics in Mathematical Methods ME7100
Fuzzy system example (Mamdani)
(2) Rule
R1 : If Distance is near and temperature is pleasant then sufficient visitors
R2 : If Distance is very far or temperature is harsh then no visitors
R3 : If Distance is near and temperature is bad then less visitors
R4 : If Distance is far and temperature is reasonable then less visitors
R5 : If Distance is very near and temperature is pleasant then more visitors




R3



R4




R5
Advance Topics in Mathematical Methods ME7100
Fuzzy system example (Mamdani)
(2) Rule
Max- min operator




R1



R2






R2
min


Advance Topics in Mathematical Methods ME7100
Fuzzy system example (Mamdani)
(2) Rule
R1 : If Distance is near and temperature is pleasant then sufficient visitors
R2 : If Distance is very far or temperature is harsh then no visitors
R3 : If Distance is near and temperature is bad then less visitors
R4 : If Distance is far and temperature is reasonable then less visitors
R5 : If Distance is very near and temperature is pleasant then more visitors




R3



R4




R5
Advance Topics in Mathematical Methods ME7100
Fuzzy system example (Mamdani)
(2) Rule
Max- min operator




R1









min


( )
,
( )
A
Z
COA
A
Z
z zdz
z
z dz

=
}
}
( ) ( ) ,
BOA
BOA
z
A A
z
z dz z dz
|
o
=
} }
*
,
{ ; ( ) }
Z
MOM
Z
A
zdz
z
dz
where Z z z
'
'
=
' = =
}
}

Вам также может понравиться