How can I tune PID controller using GENETIC ALGORITHM.......????

num=6;
den=[48 44 12 1];
gp=tf(num,den); % PLANT TRANSFER FUNCTION
%%OVERALL TRANSFER FUNCTION
gpgcr=gp*gcr;
sysy4=feedback(gpgcr,1);
%%BY GENETIC ALGORITHM
N=5; % NO OF POPULATION
M=4;
iter=12; % no.of iterations
maxiter=1;
%%initialization of pbest and gbest
for i=1:N
for j=1:iter
for k=1:M
pbest(i,j,k)=rand;
gbest(i,j,k)=rand;
end
end
end
%%initialization of kp,kd,ki and calculation of ISE
for i=1:N
for j=1:iter
for k=1:M
kp(i,j,k)=rand;
kd(i,j,k)=rand;
ki(i,j,k)=rand;
%%controller transfer function
numgc=[kd(i,j,k),kp(i,j,k),ki(i,j,k)];
dengc=[1,0];
gc=tf(numgc,dengc); % controller T.F
g=gp*gc;
GCL=feedback(g,1); % gives c(s)/r(s)
[num,den] = tfdata(GCL,'v'); % 'V' return the numerator and denominator directly as
......row vectors rather than as cell arrays
GCL1=1-GCL; % GIVES e(s)/r(s)
[nume,dene]=tfdata(GCL1,'v');
dene1=[dene,0]; % for e(s)
[r,p,k1]=residue(nume,dene1);
n=length(p);
for y=1:n
h(y)=ilaplace(r(y)/(s-p(y)));
end
h;
q=0;
for z=1:n
q=h(1,z)+q; % gives e(t)
end
q;
q=vpa(q,5); % Variable-precision arithmetic
qsq=q^2;
lqsq=laplace(qsq); % to get laplace transform
isee=limit(lqsq,s,0);
ise=vpa(isee,5);
x(i,j,k)=ise %CAN WE TAKE IT AS FITTNESS FUNCTION..??
end
end
end

3 个评论

were u able to find the fitness function by using the above codes..if not what modifications did u use to get it???
gcr is controller transfer function.

请先登录,再进行评论。

回答(3 个)

there are many mistakes in this code, so you have to fix it and remove some commands that you do not need.
Error using InputOutputModel/feedback (line 138) The first and second arguments of the "feedback" command must have compatible I/O sizes.
Error in gapid (line 6) sysy4=feedback(gpgcr,1);

评论:

2019-1-22

Community Treasure Hunt

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

Start Hunting!

Translated by