objective function and max ,min
2 views (last 30 days)
Show older comments
clc
clear
close all
Npar = 3;
VarLow=[-5.12 -5.12 -5.12];
VarHigh = [5.12 5.12 5.12];
N=100;
MaxIter=100;
XBest = rand(1,Npar).* (VarHigh - VarLow) + VarLow;
FBest=fitnessFunc(XBest);
GB=FBest;
t = cputime;
X = zeros(N, Npar);
F = zeros(N, 1);
for ii = 1:N
X(ii,:) = rand(1,Npar).* (VarHigh - VarLow) + VarLow;
F(ii) = fitnessFunc(X(ii,:));
end
for it=1:MaxIter
num=zeros(1,Npar);
for ii=1:N
for jj=1:Npar
num(jj)=num(jj)+(X(ii,jj)/F(ii));
end
end
den=sum(1./F);
%centre of mass
Xc=num/den;
for ii=1:N
for jj=1:Npar
New=X(ii,:);
New(jj)=Xc(jj)+((VarHigh(jj)*rand)/it^2);
end
New=limiter(New,VarHigh,VarLow);
newFit=fitnessFunc(New);
if newFit<F(ii)
X(ii,:)=New;
F(ii)=newFit;
if F(ii)<FBest
XBest=X(ii,:);
FBest=F(ii);
end
end
end
GB=[GB FBest];
end
t1=cputime;
fprintf('The time taken is %3.2f seconds \n',t1-t);
fprintf('The best value is :');
XBest
FBest
figure(1)
plot(0:MaxIter,GB, 'linewidth',1.2);
title('Convergence');
xlabel('Iterations');
ylabel('Objective Function (Cost)');
grid('on')
function newP=limiter(P,VarHigh,VarLow)
newP=P;
for i=1:length(P)
if newP(i)>VarHigh(i)
newP(i)=VarHigh(i);
else
if newP(i)<VarLow(i)
newP(i)=VarLow(i);
end
end
function fitness = fitnessFunc(x)
fitness = x(1)^2 - 10*cos(2*pi*x(1)) + 10;
fitness= fitness+ x(2)^2 - 10*cos(2*pi*x(2)) + 10;
fitness= fitness+ x(3)^2 - 10*cos(2*pi*x(3)) + 10;
end
please, Can anyone help me how can l store the max value of the fitness function(worst value) also how can l plot the max,min and mean of the fitness function in one figure?
0 Comments
Answers (1)
Samatha Aleti
on 28 Apr 2020
Hi,
If you want to find maximum, minimum and mean values of an array, you can use the MATLAB functions “max”, “min”, “mean” respectively.
To store the maximum value of “fitness”, you can add the following lines of code :
Fmax = 0 % Initialize
FBest=fitnessFunc(XBest); % Call to function
Fmax = max(FBest, Fmax); % Maximum value
0 Comments
See Also
Categories
Find more on Two y-axis 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!