Index problem with Yalmip

[Michael@PolyU]

Dear Prof. John,

Thank you very much for your relentless support. As I said earlier, I am a newbie to yalmip, the problem I'm encountering in my model is index problem exceeding matrix dimensions. Please, how can I rectify this, the attached file is the code I developed. you can make v=1, and Ta=25, for the testing.

Thanks, 

Michael

[Johan Löfberg]

Don't send code which requires input of data...(i.e. edit the file so that it runs)

and fix the glaring problems

>> Warning: Strict inequalities will only (if at all) be supported on binvar/intvar variables in future versions of YALMIP. 

>> Most likely you meant to write a + b, but you wrote a +b

This can easily lead to bugs, as [a +b] is a vector with two elements

If you really want to use unitary plus, you will have to edit sdpvar/uplus

and delete this message

and regarding the question, have you even tried to debug?

K>> PWt(k,:)

Error using sdpvar/subsref (line 161)

Error using sdpvar/subsref>subsref2d (line 271)

Index exceeds matrix dimensions. 

K>> PWt

Linear scalar (real, integer, 1 variable)

Coefficients range: 20 to 20

K>> k

k =

     2

[Michael@PolyU]

Thanks for your reply.. i have checked the model again but i don't really understand whats wrong, nHours=24, and i defined the affected decision variables as sdpvar(nHours, 1) e.g PWt=sdpvar(nHours,1) while k = i:nHours.

Please, how can i rectify those errors? the attached file is the modified code but is still showing some warnings that i don't understand where things are wrong (still on the learning process of Yalmip)

Thannks

[Johan Löfberg]

PWt=nwt*Pwind

Linear scalar (real, integer, 1 variable)

Coefficients range: 20 to 20

and you should install the most recent version of YALMIP, where you will get even stronger warnings to fix some issues

and regarding the a+b vs a +b warning, you are doing

 = [1+2...

      +3]

thinking it will return 1+2+3 but MATLAB will return [3 3]

[Michael@PolyU]

Thank you Prof.

i have done the necessary corrections as suggested, but I encountered another problem again, 'solver not applicable' using CPLEX, and the problem is a MILP.

the attached file is the modified code.

Thanks

[Johan Löfberg]

No, it is not a MILP

>> Batstored_constr1

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

|    ID|                            Constraint|   Coefficient range|

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

|    #1|   Equality constraint (bilinear) 24x1|      0.02 to 1.0417|

|    #2|          Element-wise inequality 24x1|              1 to 1|

[Michael@PolyU]

Prof. Johan,

Thanks for your reply... i have corrected the code, but i still have the same error 'solver not applicable'

The attached file is the modified code

[Johan Löfberg]

You're still ding the same thing...

>> capacity_constr

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

|    ID|                                Constraint|   Coefficient range|

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

|    #1|    Element-wise inequality (integer) 24x1|        0.01728 to 1|

|    #2|               Element-wise inequality 2x1|           1 to 1500|

|    #3|   Element-wise inequality (bilinear) 24x1|              1 to 1|

[Michael@PolyU]

Thank you very much for the reply, please, how can i eliminate the bilinear error?

Thanks,

Michael

[Johan Löfberg]

I don't see any way of getting rid of the bilinearity, beyond a piecewise affine approximation

[Michael@PolyU]

Thanks Johan, i have found a way to modified the problem, but i run into an infeasible problems by CPLEX.

The attached file is the code file. Please, help me have a look

[Johan Löfberg]

Your code does not run

Error using  ==  (line 15)

Error using  -  (line 195)

Matrix dimensions must agree.

Error in xcvxcvxcv (line 339)

power_balance = [power_balance,

PVt(k,:)+PWt(k,:)+Prce(k,:)+Pice(k,:)+Psofc(k,:)+Pmofc(k,:)+Ppemfc(k,:)-Pbch(k,:)+Pbdch(k,:)+Pgrid(k,:)-Psell(k,:)==Pecin(k,:)+Pashp(k,:)+Pwshp(k,:)+Pelez(k,:)+LD']; 

[Michael@PolyU]

i have done the correction in the attached file, but is still showing infeasible problem.

Thanks

[Johan Löfberg]

Already optimize(power_balance) is infeasible...

[Johan Löfberg]

and violently trivially so as it is infeasible already with

power_balance=[];

k=1

[Michael@PolyU]

you mean the problem is in the power balance constraint?

please how can i correct it?

[Johan Löfberg]

at least parts of the problems are there, since already that alone is infeasible

you correct it by finding out why you are asking for something impossible

[Johan Löfberg]

my guess you want LD(k), because now the lhs is a scalar and the rhs is a vector