Standard deviation can't be output.

Question

Qiandong Dong 2023-2-22

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1916700-standard-deviation-can-t-be-output

编辑： KSSV 2023-2-22

clear;clc;close all;
%% Parameters
p=5;
s=5;
t=10;
for lambda=0.2:0.2:0.2
    for n=1:5
        pc=80+10*ones(p,s,t);%reshape(xlsread('data.xlsx','sheet1','A2:A33'),p,s,t);
        pc1=120+10*ones(p,s,t);
        oc=200+150*ones(p,s,t);%reshape(xlsread('data.xlsx','sheet1','B2:B33'),p,s,t);
        hc=5+5*ones(p,t);%reshape(xlsread('data.xlsx','sheet1','G2:G17'),p,t);
        tc=10+8*ones(s,t);%xlsread('data.xlsx','sheet1','D2:D5');
        D0=500+100*ones(p,t);%reshape(xlsread('data.xlsx','sheet1','H2:H17'),p,t);
        sc=800+400*ones(p,s,t);%reshape(xlsread('data.xlsx','sheet1','C2:C33'),p,s,t);
        eo=2+ones(p,s,t);%reshape(xlsread('data.xlsx','sheet1','P2:P33'),p,s,t);
        eh=1+2*ones(p,t);%reshape(xlsread('data.xlsx','sheet1','N2:N17'),p,t);
        ev=2+3*ones(p,s,t);%reshape(xlsread('data.xlsx','sheet1','I2:I33'),p,s,t);
        w=350+200*ones(1,t);%xlsread('data.xlsx','sheet1','O2:O5');
        dis=100+50*ones(1,s);%[120 150];
        cc=4000;
        r=[0.01,0.1,0.1,0.05,0.15];
        cp=50;%60+10*rand;
        b=2+0.3*rand; gam=2;
        eta1=500000; epsilon=0.9; eta=0.1;theta=0.1;%lambda=0.9;
        theta1=150;p0=[0.3;0.6;0.1];e0=[1;2;3];e=e0/norm(e0,1);
        %p0=xlsread('data.xlsx','box','ah2:ah81');e0=xlsread('data.xlsx','box','z2:z81');e=e0/norm(e0,1);
        m=3;
        x=sdpvar(p,s,t);
        y=binvar(p,s,t);
        x1=sdpvar(p,s,t);
        y1=binvar(p,s,t);
        z=sdpvar(p,t);
        sh=sdpvar(p,t);
        u=sdpvar(1);beta=sdpvar(1);beta0=sdpvar(1);tao=sdpvar(m,1);tao0=sdpvar(m,1);
        r1=sdpvar(1);r2=sdpvar(m,1);r3=sdpvar(m,1);r4=sdpvar(m,1);
        lambda1=sdpvar(1);lambda2=sdpvar(m,1);lambda3=sdpvar(m,1);lambda4=sdpvar(m,1);
        r10=sdpvar(1);r20=sdpvar(m,1);r30=sdpvar(m,1);r40=sdpvar(m,1);
        lambda10=sdpvar(1);lambda20=sdpvar(m,1);lambda30=sdpvar(m,1);lambda40=sdpvar(m,1);
        %% Objective function
        obj=lambda*beta+(lambda/(1-epsilon))*(tao'*p0+(lambda2'+lambda3')*e*gam)+(1-lambda)*(sum(sum(sum(eta1.*D0)))*ones(1,m)*p0...
            +(r2'+r3')*e*gam)+10*sum(squeeze(sum(sum((x+x1),1).*tc,3)).*dis) + sum(sum(sum(oc.*(y+y1))))+sum(sum(hc.*z))+u*cp...
            +sum(theta1.*r.*(sum(sum(y+y1,1),3)))+sum(sum(sum((x1).*pc1)))+sum(sum(eta1.*(-sum(x,2)-sum(x1,2))))+sum(z(:,10))...
            +sum(sum(sum((x1).*pc1)))+sum(sum(sum((x).*pc)));%
        
        %% Constraint
        C=[x(:)>=0];
        C=[C,x1(:)>=0];
        C=[C,z(:)>=0];
        C=[C,sh>=0];
        %% disruption
        for i=1:2
            y(:,i,:)=0;
        end
        C=[C,r2(:)>=0];C=[C,r3>=0];C=[C,r4(:)>=0];
        C=[C,lambda2(:)>=0];C=[C,lambda3>=0];C=[C,lambda4(:)>=0];
        C=[C,r20(:)>=0];C=[C,r30>=0];C=[C,r40(:)>=0];
        C=[C,lambda20(:)>=0];C=[C,lambda30>=0];C=[C,lambda40(:)>=0];
        C=[C,tao(:)>=0];C=[C,tao0(:)>=0];
        C=[C,x<=sc.*y];
        C=[C,x1<=(sc-x).*y1];
        C =[C,sum(z,1)<=w];
        for i=1:p
            for j=1:t
                C =[C,sum(r.*y1(i,:,j),1)<=0.2];
            end
        end
        C =[C,-squeeze(sum(x(:,:,1),2))-squeeze(sum(x1(:,:,1),2))+D0(:,1)+z(:,1)>=0];
        for i=2:t
            C =[C,-z(:,i-1) - squeeze(sum(x(:,:,i),2)) - squeeze(sum(x1(:,:,i),2))+D0(:,i)+z(:,i)>=0];
        end
        
        C=[C,sum(sum(sum(eta1.*D0)))==r1*e-r2+r3];
        C=[C,tao==lambda1*e-lambda2+lambda3];
        C=[C,tao>=sum(sum(sum(eta1.*D0)))-beta*e];
        
        for i=1:p
            for j=1:t
                C=[C,lambda*beta0+(lambda/(1-epsilon))*(tao0'*p0+(lambda20'+lambda30')*e*gam)+(1-lambda)*(theta*D0(i,j)*ones(1,m)*p0...
                    +(r20'+r30')*e*gam)>=eta*sum(x(i,:,j)+x1(i,:,j),2)];
            end
        end
        for i=1:p
            for j=1:t
                C=[C,theta*D0(i,j)==r10*e-r20+r30];
            end
        end
        C=[C,tao0==lambda10*e-lambda20+lambda30];
        for i=1:p
            for j=1:t
                C=[C,tao0>=theta*D0(i,j)-beta0*e];
            end
        end
        
        C=[C,cc+u >= sum(sum(sum((y1+y).*eo)))+sum(sum(z.*eh))+sum(sum(sum((x1+x).*ev))) + b*sum(sum(sum((x1+x),1),3).*dis)];
        
        optimize(C,obj)
        
        f(n)=obj;
    end
    l=round(lambda/0.2);
    g(l)=mean(f);
    h(l)=std(f);
end
value(h)

3 个评论
显示 1更早的评论隐藏 1更早的评论

Qiandong Dong 2023-2-22

There is no error message, and the Matlab solver shows that it is always busy.

Dyuman Joshi 2023-2-22

I do not have the YALMIP toolbox right now to run your code.

However, you can certainly improve your code -

All the variables defined before x (first use of sdpvar) are constants, bring them out of the loops. (I do not know whether sdpvar variables are constants are not)

And it's not clear to me what is the purpose of the 2nd for loop (n as the loop variable), when there is no change in the code w.r.t n.

请先登录，再进行评论。

请先登录，再回答此问题。