MINLP non-convex optimality

Awaiting user input

Tanzin Ahmed

April 08, 2026 06:01

• 
• 
• 

For a MINLP non-convex problem, the solver do not give optimal result if the timelimit is not set. If I set a timelimit, then the result is not optimal. For different parameter tunning, I get different result. In this case, how can I get a optimal result?

Related to

• Nonconvex
• MINLP

0

• 

  Riley Clement

  • Gurobi Staff

  April 08, 2026 08:13

  Hi Tanzin,

  If I understand correctly you are letting the model run until it terminates via the default MIPGap criteria?

  If you are getting different results with different parameters, it is because Gurobi is following different solution paths.  In theory with a MIPGap of 0, the optimal objective value reported should be the same, and not affected by the solution path.  In practice this may not be true if numerics are causing issues.

  I'm happy to take a quick look at your model if you can post a link to where I can download it (Google Drive, Dropbox, etc).

  - Riley

  0

Please sign in to leave a comment.