Conic scalarization solving stuck (Dicopt)

Problems with modeling
Post Reply
player
User
User
Posts: 8
Joined: 3 years ago

Conic scalarization solving stuck (Dicopt)

Post by player »

Hi,

I am working on a min. problem with Conic Scalarization. In the subproblems of the model (NLP), it stucks on 0.10% (while: w1=1, w2=1, alpha=0.9, means that 9th iteration). It is not giving with a best integer solution. Is there a way of jumping of it. I get txt output file in each loop. Maybe, it will jump and generate an empty txt output file. Then, I can udnerstand that for these weights, there is no solution or how can I accept that solution with 0.10%?

Best,
gregbowers
User
User
Posts: 2
Joined: 6 months ago

Re: Conic scalarization solving stuck (Dicopt)

Post by gregbowers »

To handle subproblems getting stuck with a small optimality gap in mixed-integer programming (MIP) with conic scalarization:

Adjust solver tolerance.
Set a time limit for subproblems.
Implement feasibility heuristics.
Analyze output files for patterns.
Consider parallel processing.
Track dual bounds.
Explore branch-and-bound.
Experiment with constraint relaxation.
Post Reply