Constraint implementation for the Ipopt sovler with a non-contiguous target range

Question

In a project, Ipopt is used to solve a problem. I am wondering if I can add a new constraint to the problem, but I am not very deep into the topic. Basically, a variable should be greater than a certain value, or it can be zero. The latter could be a problem since the target range is not contiguous.

I was thinking of using binary variables, but as far as I can see, the Ipopt solver does not support them. Is there any way to implement my condition?

Take a look into this post, it describes almost identical problem. https://stackoverflow.com/questions/16470413/how-to-assign-binary-variable-in-ampl-in-respect-to-another-variable — CaptainTrunky, Dec 06 '22 at 10:15
I dont think so, they assume that binary variables are possible. — Lukas Malcher, Dec 06 '22 at 15:35
Ipopt is for continuous problems. You are looking for an MINLP solver. — Erwin Kalvelagen, Dec 28 '22 at 12:01

score 0 · Answer 1 · answered Jan 31 '23 at 02:38

While it would be possible to formulate this condition with a nonconvex nonlinear constraint, I would also suggest that you look for a solver for MINLP and use a binary or semicontinuous variable.

Sometimes you have to change the tool when your problem changes.

Constraint implementation for the Ipopt sovler with a non-contiguous target range

1 Answers1