How to add constraints to a optimization problem matrix

Hello everybody,

I have another problem to solve. It is a linear problem but it is pretty big and thats why Matlab needs more than 30 minutes to solve

[x,fval] = linprog(M, A, b, Aeq, beq);

My matrix A is approximately 40.000 x 20.000, so I have about 20.000 variables in my vector x.

I would like to add some constraints that the solver doesnt take to much time. I would like to add for example a constraint where I give him a possible x with a low function value M*x. My idea is that the branch and bound algorithm cuts more branches off because I added a guilty constraint with a very good value. How can I do this ? Do I have to set this as x_0 ??

Or is it possible to put this constraint in A. But I have no idea how to do this.

Actually I would like to add a row in my matrix A with this constraint, but I dont know how to combine this with my function value

Thank you very much for your help !

Best regards

Answer 1

[Pruthvi Muppavarapu]

Hi,

The following document might help you understand linear constraints and a way to use them:

https://in.mathworks.com/help/optim/ug/linear-constraints.html#brhkghv-14

Feel free to refer to the following example, which shows the usage of constraints in an optimization problem:

https://in.mathworks.com/help/optim/ug/example-linear-programming.html

Hope this answer helps.

Regards,

Pruthvi

[Kernel7364]

Hey Pruthvi,

thank you very much for the answer.

The examples and links you posted here were very helpful.

I solved my problem thanks to the links you gave me.

Have a great day and best regards,

Michael