7.11.
2022 09:36 Virtual Labs
Theory
Defuzzification:
Defuzzification is the process of producing a quantifiable result in Crisp logic, given fuzzy sets and corresponding membership degrees. It is the process that maps a fuzzy
set to a crisp set. It is typically needed in fuzzy control systems. These will have a number of rules that transform a number of variables into a fuzzy result, that is, the result
is described in terms of membership in fuzzy sets. Defuzzification is the conversion of a fuzzy quantity to a precise quantity, just as fuzzification is the conversion of a
precise quantity to a fuzzy quantity. µ
For example, Fig (a) shows the first part of the Fuzzy output and Fig (b) shows the second part of the Fuzzy output.
Then Fig (c) shows the union of the two parts (a) and (b).
A fuzzy output process may involve many output parts, and the membership function representing each part of the output can have any shape. The membership function
of the fuzzy output need not be normal always.
Defuzzification Methods:
1. Max-Membership Principle
This method is also known as height method and is limited to peak output functions. This method is given by the algebraic expression:
µ(z*) >= µ(z) for all z ∊ Z.
2. Centroid Method
This method is also known as the centre of mass, centre of area or centre of gravity. It is the most commonly used defuzzification method. The defuzzified output z* is given
by:
z* = ∫µ(z).zdz / ∫µ(z)dz
[Link]/vlabs-dev/labs/machine_learning/labs/exp9/[Link] 1/4
7.11.2022 09:36 Virtual Labs
3. Weighted Average Method
This method is valid for symmetrical output membership functions only. Each membership function is weighted by its maximum membership value. The output in the case
is given by
z* = ∑µ(z').z' / ∑µ(z') ; where z' is the maximum value of the membership function.
4. Mean-Max Membership
This method is also known as the middle of the maxima. This is closely related to the max-membership method, except that the locations of the maximum membership can
be nonunique. The output here is given by:
z* = ∑z' / n ; where z' is the maximum value of the membership function.
5. Centre of Sums
This method employs the algebraic sum of the individual fuzzy subsets instead of their union. The calculations here are very fast, but the main drawback is that the
intersecting areas are added twice. The defuzzified value z* is given by
z* = ∫ z*∑µ(z).zdz / ∫ ∑µ(z)dz
6. Centre of Largest Area
This method can be adopted when the output of at least two convex fuzzy subsets which are not overlapping. The output, in this case, is biased towards a side of one
membership function. When output fuzzy st has at least two convex regions, then the centre of gravity of the convex fuzzy subregion having the largest are is used to obtain
the defuzzified value z*. The value is given by
z* = ∫ µc(z).zdz / ∫ ∑µc(z)dz
[Link]/vlabs-dev/labs/machine_learning/labs/exp9/[Link] 2/4
7.11.2022 09:36 Virtual Labs
Let us take an example and see how the defuzzified value is calculated by all the above mentioned methods for better understanding.
Let there be 3 different fuzzy sets as shown in the figures below:-
A B C
Hence the union of all the three sets can be represented by the following figure:-
D=AUBUC
Now we shall calculate the defuzzified value using all the methods one by one.
1. Max-Membership Method
From D we can say that the maximum membership value is 1. Hence the scalar value corresponding to the maximum membership value. Therefore,
x* = 6 , 7
2. Centroid Method
In this method, we have to find the area bounded by the union of the 3 sets and find its centroid which will be the defuzzified value.
x*=∑Area × x̄ ÷ ∑Area
x*=24.989 ÷ 3.715
x*=6.72
3. Weighted Average Method
In this case for each fuzzy set the centre is calculated individually by multiplying the mean with its membership value and then the average of all sets is calculated.
Membership value of A at 2.5 = 0.3
Membership value of B at 0.5 = 0.5
Membership value of C at 6.5 = 1
x* = (2.5 × 0.3 + 0.5 × 1 + 6.5 × 1) ÷ (0.3 + 0.5 + 1)
x* = 9.75 ÷ 1.8
x* = 5.146
4. Mean-Max Membership
Here we consider the mean of the range with maximum membership value. Here the maximum membership value is 1 which is between 6 and 7. Therefore,
x* = (6 + 7) ÷ 2
x* = 13 ÷ 2
6.5
[Link]/vlabs-dev/labs/machine_learning/labs/exp9/[Link] 3/4
7.11.2022 09:36 Virtual Labs
5. Centre of Sums
In this method, we calculate the sum of the areas of individual fuzzy sets and then find the centre of this area.
Area(A) = [(5 + 3) × 0.3] ÷ 2 = 1.2
Area(B) = [(4 + 2) × 0.5] ÷ 2 = 1.5
Area(C) = [(3 + 1) × 1] ÷ 2 = 2
x* = [(1.2 × 2.5) + (1.5 × 5) + (2 × 6.5)] ÷ (1.2 + 1.5 + 2)
x* = 5
6. Centre of Largest Area
In this method, the defuzzified value is the mean value of the fuzzy set having the maximum or the largest area
From method 5 we can see that C has the largest area. Therefore,
x* = (6 + 7) ÷ 2
x* = 6.5
Lab contributed by K. J. Somaiya College of Engineering
[Link]/vlabs-dev/labs/machine_learning/labs/exp9/[Link] 4/4