I am using MATLAB's 'fmincon' to solve some nonlinear constrained optimisation problem. But it is very slow. What are the possible ways to speed up the simulation?
What is the best alternative to 'fmincon' to speed up the optimisation process so that I can use it with MATLAB?
Hi, generally fmincon is an efficient solver. In your case, perhaps your problem is particular leading to a hudge computational time. If your goal is to speed up the computation, you could use FORTRAN solvers instead of MATLAB. Very efficient open sources exist:
DONLP2 see http://www.mathematik.tu-darmstadt.de/fbereiche/numerik/staff/spellucci/DONLP2/
IPOPT: https://projects.coin-or.org/Ipopt
CONMAX from NETLIB : http://www.netlib.org/opt/index.html
You can use Artelys KNITRO as an alternative. The syntax is almost the same as fmincon on MATLAB. Note that it is not open source but you can get a trial license to try your problem.
Besides these answers, you might use LINGO since it has nonlinear constrained optimization solver.
You can use a multistrat points. See https://fr.mathworks.com/help/gads/how-globalsearch-and-multistart-work.html
The alternative may be AMPL or Opti Toolbox of MATLAB available freely online.
https://www.inverseproblem.co.nz/OPTI/
https://www.inverseproblem.co.nz/OPTI/index.php/GetStart/Overloads
https://www.inverseproblem.co.nz/OPTI/index.php/Solvers/Solvers
http://www.i2c2.aut.ac.nz/Resources/Software/OptiToolbox_Original.html
https://github.com/jonathancurrie/OPTI
- Badges
- Science topic