#Using MINLP solvers in JuMP

1 messages · Page 1 of 1 (latest)

balmy tendon
#

I've been solving simple non-convex MINLP problems with Juniper/Ipopt/Cbc succesfully so far, but for larger MINLP's (still acceptable in size) Juniper/Ipopt/Cbc ends up running indefinitely.

I tried using AmplNLWriter with Couenne, but after running I only get the following:

Couenne 0.5.8 -- an Open-Source solver for Mixed Integer Nonlinear Optimization
Mailing list: [email protected]
Instructions: http://www.coin-or.org/Couenne
couenne:

Is Couenne at least trying to solve the problem or is it just returning nothing?

prisma forge
balmy tendon
#

Hey! thanks for the response. Using the model_summary I get the following output

  • Solver : AmplNLWriter

  • Status
    Result count : 0
    Termination status : OTHER_ERROR
    Message from the solver:
    "Error calling the solver. Failed with: ProcessFailedException(Base.Process[Process('C:\Users\andre\.julia\artifacts\a193c7545d24bf6d69aa7f92ccfbe86bc05be3f6\bin\couenne.exe' 'C:\Users\andre\AppData\Local\Temp\jl_NXN3Dz\model.nl' -AMPL, ProcessExited(3221225477))])"

  • Candidate solution (result #1)
    Primal status : NO_SOLUTION
    Dual status : NO_SOLUTION

  • Work counters
    Solve time (sec) : 2.25700e+00

Apparently I can't call the solver properly?

prisma forge
#

Maybe try switching to HiGHS to see if it works?