Building a MINLP Heuristic Model in Python

Question

I am currently building a MINLP model which has around 200k decision variables and upto 100 constraints. I have access to only open source solvers which are BONMIN and COUENNE. When I try to solve the problem, I see that the solver keeps on running for more than 2 hours. I have been reading the BONMIN documentation and there I see various heuristic algorithms as options. I wanted to know is there any options list i can pass to this BONMIN solver which will trigger a heuristic algorithm that will give me a suboptimal solution in ~15 minutes?

I am working with the Pyomo package. Thanks in Advance!

score 0 · Answer 1 · answered Nov 26 '18 at 23:34

0

See this section of the Pyomo documentation on sending options to a solver: https://pyomo.readthedocs.io/en/latest/working_models.html#sending-options-to-the-solver

answered Nov 26 '18 at 23:34

Bethany Nicholson

2,723
1
11
18

I am also looking for an option list which will trigger a heuristic algorithm – linearprogrammer Nov 27 '18 at 05:49
Available options are solver specific and you should refer to the solver documentation to see what is available. – Bethany Nicholson Nov 28 '18 at 14:31

Building a MINLP Heuristic Model in Python

1 Answers1