Performance of Mixed-integer linear programming (MILP) with intlinprog

Hello, I am writing a tool for the optimisation of satellite constellation and I use Mixed-integer linear programming (MILP) with the function intlinprog. I have a lot of Variables depending on the number of satellites and the repeat cycle of the orbit, already small problems need more than 1000 Variables. The function becomes super slow. It wouldn't finish in 8 hours, but also it is not using my full CPU. I would appreciate any idea to improve the performance of the tool a lot!

Thank you very much in advance!

Kira

Answer 1

[Alan Weiss]

MILP is a notoriously difficult problem type. For suggestions, see Tuning Integer Linear Programming.

Alan Weiss

MATLAB mathematical toolbox documentation

[Kira Lara Löper]

Thank you, this is very helpful!