single-valued constant membership function

14 views (last 30 days)
Maria
Maria on 13 Dec 2014
Commented: Truong Nhu on 31 Dec 2019
hi
how to define the output fuzzy set as fuzzy singletons - which mean's that the output membership function is single-valued constant as in picture below .
So in Matlab for example I want to set "PS" MF to 4 , Which type of MF should I choose and which value or Params[ ] should I set to it?
  8 Comments
Siddhant Gudhe
Siddhant Gudhe on 17 Aug 2015
Maria, use triangular membership function only( trimf), but instead of defining its params as [-20 0 20], make it as [20 20 20]. You will get an singleton membership function at 20.
Truong Nhu
Truong Nhu on 31 Dec 2019
Thank you so much Siddhant Gudhe, you're right !

Sign in to comment.

Answers (3)

Image Analyst
Image Analyst on 14 Dec 2014
Perhaps this:
function test2
clc;
for k = 1 : 20
x = randi(20, 1, 1); % Get random input value.
mf = SetMembership(x); % Find out which it's closest to.
fprintf('%d is closest to %d\n', x, mf);
end
function mf = SetMembership(x)
% Define values for NB through PB.
dividers = -12 : 4 : 12; % -12, -8, -4, 0, 4, 8, 12
% Find out which one x is closest to
distances = abs(dividers - x);
% Find the min distance and what index it occurs at
[minDistance, indexOfMin] = min(distances);
% return the value of the dividers at that index
mf = dividers(indexOfMin);
Output:
16 is closest to 12
14 is closest to 12
8 is closest to 8
20 is closest to 12
15 is closest to 12
13 is closest to 12
19 is closest to 12
13 is closest to 12
16 is closest to 12
18 is closest to 12
16 is closest to 12
5 is closest to 4
13 is closest to 12
14 is closest to 12
16 is closest to 12
12 is closest to 12
15 is closest to 12
8 is closest to 8
14 is closest to 12
7 is closest to 8

Liyana
Liyana on 27 Mar 2017
you could choose linear or constant type of mf

kohila ramkumar
kohila ramkumar on 5 Jun 2018
IF YOU ARE CREATE SUGENO FUZZY SYSTEM. IT IS HAVING CONSTANT MEMBERSHIP FUNCTION

Categories

Find more on Data Clustering in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!