Scribd
Upload
3 views

Unit-4 Application of Soft Computing

The document discusses the application of soft computing, focusing on fuzzy logic controllers (FLC) and their components, including membership functions, fuzzification, and defuzzification processes. It outlines various membership functions, fuzzy partitions, and the structure of fuzzy if-then rules used in FLCs. Additionally, it describes different defuzzification methods such as the center of gravity, mean of maxima, and center of sums, highlighting their significance in converting fuzzy outputs to crisp values.

Uploaded by

masterblaster4595
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
3 views

Unit-4 Application of Soft Computing

The document discusses the application of soft computing, focusing on fuzzy logic controllers (FLC) and their components, including membership functions, fuzzification, and defuzzification processes. It outlines various membership functions, fuzzy partitions, and the structure of fuzzy if-then rules used in FLCs. Additionally, it describes different defuzzification methods such as the center of gravity, mean of maxima, and center of sums, highlighting their significance in converting fuzzy outputs to crisp values.

Uploaded by

masterblaster4595
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 48

Application of Soft Computing

Unit 4
Membership Functions (MF’s)
• A fuzzy set is completely characterized by a membership
function.
– a possibility measure.
– not a probability measure.

“tall” in Asia
Membership

1
“tall” in USA
value

“tall” in NBA
0
5’10” height
Fuzzy Partitions
• Fuzzy partitions are formed by the linguistic values “young”,
“middle aged”, and “old”.
Commonly used membership functions
• Triangular MF

• Trapezoidal MF

• Gaussian MF

• Generalized bell MF

• Sigmoid MF
Modules of FLC
• The basic configuration of fuzzy logic controller (FLC) is
shown in Figure.

Input Output

Fuzzification Defuzzification
Inference Engine
Inte rface Inte rface

• Knowle dge Base

Data base Rule base

5
Fuzzy If-Then Rules
AB  If x is A then y is B.
antecedent consequence
or or
premise conclusion

Linguistic Linguistic Linguistic


Linguistic
Variable Value Value
Variable

if temperature is cold and oil is cheap


then heating is high
FUZZIFICATION
• Fuzzification converts input data into suitable linguistic values which may
be viewed as labels of fuzzy sets.

• It consists of two main components:


– Membership functions, and
– Labels

• The function that ties a number to each element x of the universe is


called the membership function μ(x).

• Thus a membership function μA(x) is associated with a fuzzy set A such


that the function maps every element of the universe of discourse X to
the interval [0, 1].

27
Knowledge Base and Fuzzy Inference

• Data base and rule base are the two components of knowledge base
of an FLC.

• Database provides necessary definitions used for linguistic control


rules and fuzzy data.

• If-then rule statements are used to formulate the conditional


statements. A rule base is the collection of all these statements. The
rules are a set of linguistic statements based on expert knowledge,
including experience and heuristics, instead of detailed mathematical
model.

28
Contd……

• In the fuzzy inference engine, fuzzy logic principles are used to


combine fuzzy rules into a mapping from fuzzy input sets to fuzzy
output sets.

• The process of fuzzy inference involves: membership functions, fuzzy


logic operators, and if-then rules.

• Mathematically, fuzzy rule-based inference can be viewed as an


interpolation scheme because it enables the fusion of multiple fuzzy
rules when their conditions are all satisfied to a degree.

29
Defuzzification

• During defuzzification, the controller converts the fuzzy output into a


crisp data. Hence, defuzzification is a process of conversion of fuzzy
set to a single crisp value. It is the reverse process of fuzzification.

• There are several defuzzification methods. The following


defuzzification methods have been predominant in the fuzzy control:
– Centre of gravity method,

– Mean of maxima method, and

– Centre of sums

30
Centre of gravity method

• The center of gravity defuzzification method is also referred as centroid


method or centre of area method.

• It is most commonly used defuzzification method as it provides an accurate


result based on the weighted values of several output membership functions.
The defuzzified value is the weighted average of the elements in the support
set.

• The crisp output value, say x*, for a continuous membership function is
expressed as:

x* 
 μ(x).xdx
 μ(x)dx
31
Contd…..

32
Cond….
• On the other hand, for a discrete membership function, the integration is
replaced by summation.

n
 μ(xi ).xi
x*  i 1
n
 μ(xi )
i 1

• where, n represents number of samples, xi’s are the elements and μ(xi) is
its membership value.

33
Mean of maxima method

• An intuitive approach is to choose the point with the strongest possibility,


i.e., maximal membership. The maximum value defuzzification method
form the basis of final fuzzy output value on the rule output with the
highest membership function grade.

• It may happen that several such points exit and a common practice is to
take the mean of maxima (MOM). This method is mainly used with discrete
output membership functions. The defuzzified output in this case is given
as:
•
n
 xi
x M
x*M={
• where,  ixi | μ(xi) is equal to the height of fuzzy set}
M

34
Mean of maxima method

35
Centre of sums method
• The overlapping area in centroid method is considered only once, whereas
in centre of sums (COS) method the overlapping area is counted twice. The
defuzzified output x* in this case is given as:
N n
 xi  μAk (xi )
x*  i 1 k 1
N n
  μAk (xi )
i 1 k 1
• where, n is the number of fuzzy sets and N is the number of fuzzy variables.

• This method is also quite popular due to its easy implementation and fast
inference cycles.

36
37

You might also like