Can I get the code for Economic Load Dispatch using Gradient Descent (matlab)?

I'm not getting how to proceed after this!!

clc
a=[500;400;200];
b=[5.31;5.5;5.8];
g=[0.004;0.005;0.009];
lambda=input('enter vvalue of lambda');
iter=0;
Pd=800;
n=length(a);
delP=5;tcost=0;
tolerance=0.0001;
while abs(delP)>tolerance
    iter=iter+1;
    x=0; 
    for i=1:n
        p(i)=(lambda-b(i))/(2*g(i));
        x=x+(1/(2*g(i)));
        cost=(a(i)+(b(i)*p(i))+(g(i)*p(i)*p(i)));
  

[Suzzane M]

Let's assume F1= 561+7.92P1+0001562P1*P1

F2= 310+7.85P2+0.00194P2*P2

F3=78+7.97P3+0.00482P3*P3

Total load demand=850MW

P1=400MW

P2=300MW

P3=150MW

Ploss=0.00003P1*P1+0.00009P2*P2+0.00012P3*P3

[mahesh kumar]

HI! which paper/text are you following..for gradient descent algr..

(other than allen j. wood??)

[Suzzane M]

Hi, I am basically following allen.J wood along with this-http://home.iitk.ac.in/~saikatc/EE632_files/chap2.pdf

I've made some progress in code but I still don't know how to include loss?

clear all

clc

linedata=[ 0.001562 7.92 561 400;

0.00194 7.85 310 300;

0.00482 7.97 78 150; ]

a = linedata(:,1);

b = linedata(:,2);

c = linedata(:,3);

p = linedata(:,4);

for iter = 1:9

for i = 1:3

df(i) = b(i)+(2*a(i)*p(i));

end

lamda = sum(df)/3;

lamda

iter

for i = 1:3

pn(i) = p(i)-(df(i)-lamda);

end

pn

pt=sum(pn)

for i = 1:3

fc(i) = c(i)+b(i)*p(i)+(a(i)*(p(i)^2));

end

fulcost = sum(fc)

for i = 1:3;

p(i) = pn(i)

end

end

[mahesh kumar]

nice sister, did you get your answer?

[mahesh kumar]

you seem to have bn using too many for loops; try this..

clc
clear
linedata=[  0.001562  7.92  561  400;
            0.00194   7.85  310  300;
            0.00482   7.97  78   150;  ];
          dP=1;
a  = linedata(:,1);
b  = linedata(:,2);
c  = linedata(:,3);
p  = linedata(:,4);
while abs(dP)>0.001
    
df=b+2*a.*p;
lamda = sum(df)/3;
disp(lamda)
pn=p-df+lamda;
disp(pn)
% pt=sum(pn)
% for i = 1:3
%     fc(i) = c(i)+b(i)*p(i)+(a(i)*(p(i)^2));
% end
dP=p-pn
p=pn; disp(p); disp(sum(p))
end
fc=c+b.*p+a.*p.*p;
fuelcost = sum(fc);
disp(fuelcost)

[Suzzane M]

Hi. Yes it runs perfectly. But how to include losses in the above code ?

[mahesh kumar]

I'll send something today,by evening,ok?

[Suzzane M]

Yes. Thankyou so much!

[Suzzane M]

Ploss=0.00003P1*P1+0.00009P2*P2+0.00012P3*P3 This is the loss equation to be incorporated in the above code.

[mahesh kumar]

try this::

B  = [.00003 0 0;0 .00009 0 ; 0 0 .00012]; dB=diag(B); load=850;
x=max(b); dP=1;i=0;
z=randperm(10000,length(b))'; P=load*z/sum(z);
while abs(dP)>0.000001
    i=i+1; disp(i);
    dP=load+P'*B*P-sum(P); % loss=P'*B*P;
    x=x+dP*2/(sum(1./a));    
    P=(x-b-2*(B*P-dB.*P))./(2*a+2*x*dB);
%     plot(x,P,'x'); hold on; pause(1);
%     C=c+b.*P+a.*P.*P;
%     plot(x,sum(C)/100,'o');
end

[Suzzane M]

Yes, it solves my problem. Thankyou so much for the help. Was badly stuck in for days. Fianlly got some clarity. Thankyou again

Answer 1

[mahesh kumar]

some answer, if helps everyone