how can I model contains summation of different sdpvar inside a "FOR loop"

[Ahmed Rabee]

I want to model this equation:

for t = 1:24, for n=1:5,  Do    

       Sum[ FP (k,t) ] + P(n,t) = L(n,t)   where, k selected from K =[ 1:6 ] based on n  

and

FP  is Linear matrix variable 6x24 (full, real, 120 variables)

P is linear matrix variable 5x24 (full, real, 120 variables)

L is 5*24 double.

Please refer to the attached file lines 34:46, 

No error but these constraints not working !!!!

[Johan Löfberg]

WEll, thr ode is correct matlab code, so you're simply not writing the matlab code you think.

Since you only say it is incorrect but not how, it i impossible to understand what you want and what you got. There is no sum in the code though which you have in you post test

[Ahmed Rabee]

I use the following lines to make the summation: line 36-38

for l_n=1:length(lineNo)           % for all connected lines connected to node n

                FP_n(n,t)  = FP_n(n,t) + (-1)^dir(l_n) * FP(lineNo(l_n),t);

end

then I use the sumation in Lines 42  or 44:

                Constraints     = [Constraints, FP_n(n,t) + P(gen_index,t)  == load_n(n)] ;  

or             Constraints     = [Constraints, FP_n(n,t) == load_n(n)] ;      

where, l_n is k of the equation, lineNo is a subset of K, gen_index is n if there, Load_n is L

so we can write this loop like:

for  k =1: K_sub          % for all connected lines connected to node n

                FP_n(n,t)  = FP_n(n,t) + (-1)^dir(l_n) * FP(k,t);

end

thank you, professor, for your help

[Johan Löfberg]

so you've solved your problems?

this looks fishy though

FP_n  = sdpvar(N, TN, 'full');

           for l_n=1:length(lineNo)           % for all connected lines connected to node n

                FP_n(n,t)  = FP_n(n,t) + (-1)^dir(l_n) * FP(lineNo(l_n),t);

            end

as that means FP_n will be the sum of FP etc, plus the decision variable that initially was in FP_n, and since there are no constraints on those, FP_n is effectively completely free

Anology

x = sdpvar(1) % Variable 1

y = sdpvar(2,1); Variable 2 and 3

for i = 1:n

 x = x + y(i);

end

Will lead to y = variable1 + variable2 + variable3.