Operator '*' is not supported for operands of type 'struct'.

Dear Sir

Nice Greeting

I get this Error

Operator '*' is not supported for operands of type 'struct'.

with this objective function

of =2.6773*x(1)+3.1007*x(2)+3.2178*x(3)+ 3.8021*x(4)+3.6727*x(5)+3.0442*x(6)+1.6362*x(7)+3.0479*x(8)+3.3917*x(9)+3.7468*x(10)+3.8989*x(11)+3.1073*x(12)+3.3207*x(13)+ 1.6382*x(14);

How can I solve this Error

Thanks in advance

[Walter Roberson]

what is class(x) when you stop in the function? How is the function being invoked?

[Andrea Strappato]

Hi,

The Matlab struct type is defined as "struct_name = struct('element_name', Array)" where Array is the set of x(i) values. So to extract the x(i) value try to write "struct_name(i).x".

[tahseen alshmary]

Hi

excuse me, can you writ the objective function to me by formula x(i)

[Andrea Strappato]

x(i) is the i-values/element from your x struct data type. So , for example, to get the i-th element from x try this: "struct_name(i).x".

Can you show your struct x data type?

Answer 1

[Walter Roberson]

The code itself did not have the problem you reported, but you need a replacement PSO.m file.

% Problem preparation 
problem.nVar = 14;
problem.ub = 1.1 * ones(1, 14);
problem.lb = 0.05 * ones(1, 14);
problem.fobj = @ofun;
%--------------------------------------------------------------------------
noP=30;                 %number of population
maxIter=1000;   %number of iteration
func=@ofun;
visFlag = 1; % set this to 0 if you do not want visualization
RunNo  = 30; 
BestSolutions_PSO = zeros(1 , RunNo);
%%
[GBEST,cgcurve ] = PSO( noP , maxIter, problem , visFlag ) ;

The modified PSO.m is attached.

Also, I optimized your ofun slightly;

function f = ofun(x)   % objective function (minimization) 
of =2.6773*x(1)+3.1007*x(2)+3.2178*x(3)+ 3.8021*x(4)+3.6727*x(5)+3.0442*x(6)+1.6362*x(7)+3.0479*x(8)+3.3917*x(9)+3.7468*x(10)+3.8989*x(11)+3.1073*x(12)+3.3207*x(13)+ 1.6382*x(14);
% if there is no constraints then comments all c0 lines below    
c0 = zeros(1,0, 'like', x);
CTI =0.3;
A(1) =2.6773*x(1);   B(1) =4.2191*x(6);    %first row
A(2) =3.1007*x(2);   B(2) =5.0777*x(1);    %second row
A(3) =3.1007*x(2);   B(3) =2.1891*x(7);    %third row
A(4) =3.2178*x(3);   B(4) =3.9908*x(2);    %fourth row
A(5) =3.8021*x(4);   B(5) =4.1278*x(3);    %fifth row
A(6) =3.6727*x(5);   B(6) =5.0785*x(4);    %sixth row
A(7) =3.0442*x(6);   B(7) =5.8436*x(5);    %seventh row
A(8) =1.6362*x(7);   B(8) =5.8436*x(5);    %eighth row
A(9) =3.0442*x(6);   B(9) =2.1962*x(14);   %nighth row
A(10)=1.6362*x(7);   B(10)=7.3310*x(13);   %tenth row
A(11)=3.0479*x(8);   B(11)=2.1891*x(7);    %eleventh row
A(12)=3.0479*x(8);   B(12)=5.4250*x(9);    %tweleventh row
A(13)=3.3917*x(9);   B(13)=4.9527*x(10);   %thirteenth row
A(14)=3.7468*x(10);  B(14)=5.2862*x(11);   %fourteenth row
A(15)=3.8989*x(11);  B(15)=3.8989*x(12);   %fiveteenth row
A(16)=3.1073*x(12);  B(16)=7.3310*x(13);   %sixteenth row
A(17)=3.1073*x(12);  B(17)=2.1962*x(14);   %seventennth row
A(18)=3.3207*x(13);  B(18)=4.4351*x(8);    %eighteenth row
A(19)=1.6382*x(14);  B(19)=5.0777*x(1);    %nighnteenth row
A(20)=1.6382*x(14);  B(20)=5.4250*x(9);    % twenty row


c0(1)=  B(1) -A(1)-CTI;
c0(2)=  B(2) -A(2)-CTI;
c0(3)=  B(3) -A(3)-CTI;
c0(4)=  B(4) -A(4)-CTI;
c0(5)=  B(5) -A(5)-CTI;
c0(6)=  B(6) -A(6)-CTI;
c0(7)=  B(7) -A(7)-CTI;
c0(8)=  B(8) -A(8)-CTI;
c0(9)=  B(9) -A(9)-CTI;
c0(10)= B(10)-A(10)-CTI;
c0(11)= B(11)-A(11)-CTI;
c0(12)= B(12)-A(12)-CTI;
c0(13)= B(13)-A(13)-CTI;
c0(14)= B(14)-A(14)-CTI;
c0(15)= B(15)-A(15)-CTI;
c0(16)= B(16)-A(16)-CTI;
c0(17)= B(17)-A(17)-CTI;
c0(18)= B(18)-A(18)-CTI;
c0(19)= B(19)-A(19)-CTI;
c0(20)= B(20)-A(20)-CTI;

if isa(x, 'sym')
    D = arrayfun(@(C) piecewise(C < 0, -C*1000, C), c0);
else
    D = c0;
    mask = D < 0;
    D(mask) = -1000*D(mask);
end
Z = sum(D);
f=of+Z;                      % fitness function
end

[tahseen alshmary]

[Walter Roberson]

You posted your entire code, and that code did not have a call

func(GBEST);

If you want to know the value of the function at the best location, just examine

GBEST.O

and the best location found is

GBEST.X

Answer 2

[tahseen alshmary]

Andrea Strappato

Thank you so much

Answer 3

[tahseen alshmary]

Walter Roberson

what do you mean

the full objective function below

of =2.6773*x(1)+3.1007*x(2)+3.2178*x(3)+ 3.8021*x(4)+3.6727*x(5)+3.0442*x(6)+1.6362*x(7)+3.0479*x(8)+3.3917*x(9)+3.7468*x(10)+3.8989*x(11)+3.1073*x(12)+3.3207*x(13)+ 1.6382*x(14);
% if there is no constraints then comments all c0 lines below    
c0=[];
CTI =0.3;
A(1) =2.6773*x(1);   B(1) =4.2191*x(6);    %first row
A(2) =3.1007*x(2);   B(2) =5.0777*x(1);    %second row
A(3) =3.1007*x(2);   B(3) =2.1891*x(7);    %third row
A(4) =3.2178*x(3);   B(4) =3.9908*x(2);    %fourth row
A(5) =3.8021*x(4);   B(5) =4.1278*x(3);    %fifth row
A(6) =3.6727*x(5);   B(6) =5.0785*x(4);    %sixth row
A(7) =3.0442*x(6);   B(7) =5.8436*x(5);    %seventh row
A(8) =1.6362*x(7);   B(8) =5.8436*x(5);    %eighth row
A(9) =3.0442*x(6);   B(9) =2.1962*x(14);   %nighth row
A(10)=1.6362*x(7);   B(10)=7.3310*x(13);   %tenth row
A(11)=3.0479*x(8);   B(11)=2.1891*x(7);    %eleventh row
A(12)=3.0479*x(8);   B(12)=5.4250*x(9);    %tweleventh row
A(13)=3.3917*x(9);   B(13)=4.9527*x(10);   %thirteenth row
A(14)=3.7468*x(10);  B(14)=5.2862*x(11);   %fourteenth row
A(15)=3.8989*x(11);  B(15)=3.8989*x(12);   %fiveteenth row
A(16)=3.1073*x(12);  B(16)=7.3310*x(13);   %sixteenth row
A(17)=3.1073*x(12);  B(17)=2.1962*x(14);   %seventennth row
A(18)=3.3207*x(13);  B(18)=4.4351*x(8);    %eighteenth row
A(19)=1.6382*x(14);  B(19)=5.0777*x(1);    %nighnteenth row
A(20)=1.6382*x(14);  B(20)=5.4250*x(9);    % twenty row

[Walter Roberson]

Please post your entire code.

[tahseen alshmary]

of =2.6773*x(1)+3.1007*x(2)+3.2178*x(3)+ 3.8021*x(4)+3.6727*x(5)+3.0442*x(6)+1.6362*x(7)+3.0479*x(8)+3.3917*x(9)+3.7468*x(10)+3.8989*x(11)+3.1073*x(12)+3.3207*x(13)+ 1.6382*x(14);

% if there is no constraints then comments all c0 lines below

c0=[];

CTI =0.3;

A(1) =2.6773*x(1); B(1) =4.2191*x(6); %first row

A(2) =3.1007*x(2); B(2) =5.0777*x(1); %second row

A(3) =3.1007*x(2); B(3) =2.1891*x(7); %third row

A(4) =3.2178*x(3); B(4) =3.9908*x(2); %fourth row

A(5) =3.8021*x(4); B(5) =4.1278*x(3); %fifth row

A(6) =3.6727*x(5); B(6) =5.0785*x(4); %sixth row

A(7) =3.0442*x(6); B(7) =5.8436*x(5); %seventh row

A(8) =1.6362*x(7); B(8) =5.8436*x(5); %eighth row

A(9) =3.0442*x(6); B(9) =2.1962*x(14); %nighth row

A(10)=1.6362*x(7); B(10)=7.3310*x(13); %tenth row

A(11)=3.0479*x(8); B(11)=2.1891*x(7); %eleventh row

A(12)=3.0479*x(8); B(12)=5.4250*x(9); %tweleventh row

A(13)=3.3917*x(9); B(13)=4.9527*x(10); %thirteenth row

A(14)=3.7468*x(10); B(14)=5.2862*x(11); %fourteenth row

A(15)=3.8989*x(11); B(15)=3.8989*x(12); %fiveteenth row

A(16)=3.1073*x(12); B(16)=7.3310*x(13); %sixteenth row

A(17)=3.1073*x(12); B(17)=2.1962*x(14); %seventennth row

A(18)=3.3207*x(13); B(18)=4.4351*x(8); %eighteenth row

[Andrea Strappato]

What about the x variable? It's seems not defined.

[tahseen alshmary]

function f = ofun(x)   % objective function (minimization) 
of =2.6773*x(1)+3.1007*x(2)+3.2178*x(3)+ 3.8021*x(4)+3.6727*x(5)+3.0442*x(6)+1.6362*x(7)+3.0479*x(8)+3.3917*x(9)+3.7468*x(10)+3.8989*x(11)+3.1073*x(12)+3.3207*x(13)+ 1.6382*x(14);
% if there is no constraints then comments all c0 lines below    
c0=[];
CTI =0.3;
A(1) =2.6773*x(1);   B(1) =4.2191*x(6);    %first row
A(2) =3.1007*x(2);   B(2) =5.0777*x(1);    %second row
A(3) =3.1007*x(2);   B(3) =2.1891*x(7);    %third row
A(4) =3.2178*x(3);   B(4) =3.9908*x(2);    %fourth row
A(5) =3.8021*x(4);   B(5) =4.1278*x(3);    %fifth row
A(6) =3.6727*x(5);   B(6) =5.0785*x(4);    %sixth row
A(7) =3.0442*x(6);   B(7) =5.8436*x(5);    %seventh row
A(8) =1.6362*x(7);   B(8) =5.8436*x(5);    %eighth row
A(9) =3.0442*x(6);   B(9) =2.1962*x(14);   %nighth row
A(10)=1.6362*x(7);   B(10)=7.3310*x(13);   %tenth row
A(11)=3.0479*x(8);   B(11)=2.1891*x(7);    %eleventh row
A(12)=3.0479*x(8);   B(12)=5.4250*x(9);    %tweleventh row
A(13)=3.3917*x(9);   B(13)=4.9527*x(10);   %thirteenth row
A(14)=3.7468*x(10);  B(14)=5.2862*x(11);   %fourteenth row
A(15)=3.8989*x(11);  B(15)=3.8989*x(12);   %fiveteenth row
A(16)=3.1073*x(12);  B(16)=7.3310*x(13);   %sixteenth row
A(17)=3.1073*x(12);  B(17)=2.1962*x(14);   %seventennth row
A(18)=3.3207*x(13);  B(18)=4.4351*x(8);    %eighteenth row
A(19)=1.6382*x(14);  B(19)=5.0777*x(1);    %nighnteenth row
A(20)=1.6382*x(14);  B(20)=5.4250*x(9);    % twenty row


c0(1)=  B(1) -A(1)-CTI;
c0(2)=  B(2) -A(2)-CTI;
c0(3)=  B(3) -A(3)-CTI;
c0(4)=  B(4) -A(4)-CTI;
c0(5)=  B(5) -A(5)-CTI;
c0(6)=  B(6) -A(6)-CTI;
c0(7)=  B(7) -A(7)-CTI;
c0(8)=  B(8) -A(8)-CTI;
c0(9)=  B(9) -A(9)-CTI;
c0(10)= B(10)-A(10)-CTI;
c0(11)= B(11)-A(11)-CTI;
c0(12)= B(12)-A(12)-CTI;
c0(13)= B(13)-A(13)-CTI;
c0(14)= B(14)-A(14)-CTI;
c0(15)= B(15)-A(15)-CTI;
c0(16)= B(16)-A(16)-CTI;
c0(17)= B(17)-A(17)-CTI;
c0(18)= B(18)-A(18)-CTI;
c0(19)= B(19)-A(19)-CTI;
c0(20)= B(20)-A(20)-CTI;

% for d = 1: length(c0)
%      if c0(d)<0
%          c(d)=1;
%      else
%          c(d)=0;
%      end
% end
% penalty = 10000;           % penalty on each constraint violation
% E = penalty*sum(c);
for i=1:length(c0)
    if c0(i)<0
        c0(i)=c0(i)*1000;
    end
    D(i) = abs(c0(i));
end
Z = sum(D);
f=of+Z;                      % fitness function

[Walter Roberson]

Okay, but please show one of the calls to ofun . We need to see what you are passing in to ofun()

[tahseen alshmary]

noP=10; %number of population

maxIter=1000; %number of iteration

func=@ofun;

visFlag = 1; % set this to 0 if you do not want visualization

[Walter Roberson]

I do not see any call to ofun there. I see you constructing a function handle and assigning it to func, but I do not see you calling ofun or func .

The code you posted hints to me that you might have created your own genetic algorithm code ?

[tahseen alshmary]

clear all
close all
clc
tic
warning off
addpath('PSO')
% Problem preparation 
problem.nVar = 14;
problem.ub = 1.1 * ones(1, 14);
problem.lb = 0.05 * ones(1, 14);
problem.fobj = @ofun;
%--------------------------------------------------------------------------
noP=30;                 %number of population
maxIter=1000;   %number of iteration
func=@ofun;
visFlag = 1; % set this to 0 if you do not want visualization
RunNo  = 30; 
BestSolutions_PSO = zeros(1 , RunNo);
%%
[GBEST,cgcurve ] = PSO( noP , maxIter, problem , visFlag ) ;

[Walter Roberson]

Mathworks does not supply a function named PSO . The Mathworks particle swarm function is named particleswarm()

So we would need to know where you found the PSO.m function .

Is it https://www.mathworks.com/matlabcentral/fileexchange/67429-a-simple-implementation-of-particle-swarm-optimization-pso-algorithm ?

[tahseen alshmary]

Yes it is