LMI Feasibility, a quick question!

[Hassan D]

Hi professor Löfberg, 

I have a quick question. I am working on time-delay systems, and some methods can lead to linear matrix inequalities.

I used your toolbox with different solvers, but as I am not familiar with optimization, I can't interpret the report from the solver, so, using the value command, I reconstruct the LMIs. Then, I check their eigenvalues to see whether they are neg or pos. 

Is this a right thing to do?!! I have attached the file I have written!

by the way, is there a file or something that explains what the report says?

Thank you for your time and consideration.

Hassan

[Johan Löfberg]

Why don't you create PI_1 etc symbolically to begin with, and then after solving the problem you simply use value(PI_1)

or you apply the command check on your constraints after solving

http://users.isy.liu.se/johanl/yalmip/pmwiki.php?n=Commands.Check

[Hassan D]

Thank you for your comment.

I didn't know about the check command, but is checking the eigenvalues wrong?

[Johan Löfberg]

same thing. read the manual on check.