Multiobjective Particle Swarm optimizer with dynamic epsilon-dominance sorting

In this paper, we propose a general dynamic epsilon non-domination sorting procedure to replace the exhaustive search approach used in the literature to tune the numerical values of the epsilon parameter of each objective in a multiobjective optimization problem (MOP). We integrate this approach into an MOPSO (Dyn-ε -MOPSO). Comparative evaluations using several multi-objectives test problems demonstrate the merit of our proposed dynamic epsilon dominance sorting.