larg number constraints in matlab
1 次查看(过去 30 天)
显示 更早的评论
I have an optimization problem which have a larger number of variables,N=20;
how i can define this constrain? please answer to me
Qe and Qi is a matrix (40*10)
Z=[I2g,I2h, W]';
W=sdpvar(1,2*N+2);
function [cineq,ceq]=nonlcon(Z)
N=20;
for(n=1:N)
cineq(1*n)= Qe(n,:)*PP*Z-W(1)*ones(N,1)<= -10000; %%for all N indicates a vector inequality
cineq(2*n)= -Qe(n,:)*PP*Z+W(2)*ones(N,1)<= 10000; %%for all N indicates a vector inequality
cineq(3*n)= Qi(n,:)*ZZ+500-W(3+n)<=0;
cineq(4*n)= -W(3+n)<=0;
cineq(5*n)= W(N+3+n)-Qi(n,:)*ZZ-500<=0;
cineq(6*n)= W(N+3+n)<=0; %%for all N indicates a vector inequality
cineq(7*n)=sum(W(3:N+2))-sum(W(N+3:2*N+2))<=N*Iavemax;
cineq(8*n)=500+Qi(n,:)*ZZ<=Imax*ones(N,1);
cineq(9*n)=-Imax*ones(N,1)<=500+Qi(n,:)*ZZ;
end
ceq=[];
end
0 个评论
采纳的回答
Walter Roberson
2021-9-14
Maybe?
sdpvar() is from the third-party package YALMIP https://yalmip.github.io so whether it works or not depends upon YALMIP not on MATLAB.
The [cineq,ceq] style like that is for nonlinear constraints defined by MATLAB. I am not aware that YALMIP has any facility to call nonlinear constraints using that style.
The individual elements you assign look plausible to me to be YALMIP constraint declarations on their right-hand side.
The locations you assign to on the left are broken. Consider than when n = 1, the fourth cineq(4*n) would be assigning to cineq(4*1) = cineq(4) . But when n = 2, the second cineq(2*n) would be assigning to cineq(2*2) = cineq(4) which would overwrite the previous entry. And you never write to any location with a prime index greater than 9. cineq(13) is never assigned to, for example.
5 个评论
Walter Roberson
2021-9-14
You have to move
const=sum(cineq);
SS=optimize(const,obj);
z=value(Z);
to before the function statement.
Also, that sum(cineq) will have to be changed to sum(sum(cineq)) or sum(cineq(:))
更多回答(1 个)
Johan Löfberg
2021-9-14
编辑:Walter Roberson
2021-9-14
You're better off posting YALMIP specific questions on YALMIP specific forums
https://github.com/yalmip/YALMIP/discussions (new, prefered)
另请参阅
产品
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!