Problem with fuzzy logic membership function assignment in code

I have a problem with Optimization of fuzzy controller in code before run simulink file with the updated mfs values

fuzzy3 is fis file name with 3 inputs and 1 output

first command works

z=[Fuzzy3.input(2).mf.params] 

but the seconds not work

[Fuzzy3.input(2).mf.params]= zed

it returns ... Insufficient number of outputs from right hand side of equal sign to satisfy assignment.

the first output z is 1*23 double and also the second zed ... same size

could anyone help me solve this issue.

Answer 1

[Walter Roberson]

zc = num2cell(z);
[Fuzzy3.input(2).mf.params]= zc{:};

[Ahmed shams]

Thanks for responding

i try to complete the project ,but i have another errors

   zc=num2cell(z);
    [Fuzzy3.input(2).mf.params]=  zc{:};
     writefis(Fuzzy3,'FIS4');
    sim('Fuzzytuning_ABC')

the error

writefis(Fuzzy3,'FIS4');

Error using Experiment_Fuzzy (line 58)

Error in 'Fuzzytuning_ABC/Fuzzy

control ver2/Fuzzy Logic Controller':

Initialization commands cannot be

evaluated.

Caused by:

Error using coder.internal.assert

(line 33)

Trapezoidal membership function

must have four parameters.

[Ahmed shams]

or another solution to convert all FIS in matlab to a code

is this possible and how?

[Walter Roberson]

z=[Fuzzy3.input(2).mf.params] 

That will expand all of the struct entries Fuzzy3.input(2).mf(:) taking their params field and concatenating them all together into one single row. If there are any internal divisions, such as one of them having 3 parameters and another having 4 parameters, then this will not preserve any internal boundaries.

[Fuzzy3.input(2).mf.params]= zed

If zed is a vector of values concatenated together, then it has no internal structure to say which values are to go into which mf() entry. You would need to do something like

plens = cellfun(@length, {Fuzzy3.input(2).mf.params});
zedsplit = mat2cell(zed, 1, plens);
[Fuzzy3.input(2).mf.params] = zedsplit{:};

[Ahmed shams]

yes , i have 7 mfs , 5 with three (Triang)and 2 with 4 parameters (trap)

could you give more on code to change them?

[Walter Roberson]

I already gave you the code needed. The code with plens will figure out how many parameters are needed for each entry, and will split the new vector as needed to match those lengths.

[Walter Roberson]

Example:

%t is a structure array with fields of different sizes
for K = 1 : 5; t(K).params = 1:randi(7); end
%look at it
for K = 1 : 5; disp(t(K).params); end

     1     2

     1     2     3     4     5     6     7

     1     2

     1     2     3     4

     1     2     3     4

plens = cellfun(@length, {t.params});

%replacement data
zed = 1 : sum(plens);

%do the replacement
zedsplit = mat2cell(zed, 1, plens);
[t.params] = zedsplit{:};

%look at the result
for K = 1 : 5; disp(t(K).params); end

     1     2

     3     4     5     6     7     8     9

    10    11

    12    13    14    15

    16    17    18    19

Yup, it works.

Answer 2

[Ahmed shams]

Thanks for help

the last code works for fuzzy MF with cellfun and i will insert the code here