Home Install Documentation Problems Tree Applications Forum

Save your money
via IT outsourcing!

Ukrainian HI-TECH Initiative


From OpenOpt

Jump to: navigation, search
Multi-Objective Problems (MOP)

\vec f(\vec x) \to \vec F \\
\text{subjected to} \\
\vec g(\vec x) & \le \vec 0 \\
\vec h(\vec x) & = \vec 0 \\
\vec x_l \le & \vec x \le \vec x_u \\
\vec x \in & R^n \\
\vec f: R^n \to R^m\\
\vec F \in (R \cup \{-\infty,\infty\})^m\\

Solver License Made by Info
(since v 0.38)
BSD Dmitrey Can handle global constrained MOPs with specifiable accuracies and both discrete and continuous variables. You can be 100% sure your result covers whole Pareto front according to the required tolerances on objective functions. If you want to speedup interalg on a MOP, first of all pay attention to its parameter maxActiveNodes and interalg MOP parameter sigma, and nProc (default: 1) for multi-CPU systems. Trajectories for different nProc should be same, but sometimes differ due to roundoff errors.

future plans for interalg include some speedup and better multiprocessing, possibility to start from a Pareto front obtained by another (inexact) MOP solver

See also:

  • wikipedia webpage for multi-objective optimization
  • NLP (nonlinear problems)
  • MINLP (mixed-integer nonlinear problems)
  • GLP (global nonlinear problems)
Retrieved from "http://openopt.org/MOP"
Personal tools
    Latest OOSuite 0.38

    from 2012-03-15

    Next release: