Evolutionary Multi-modal Optimization with the Use of Multi-objective Techniques

Abstract

When evolutionary algorithms for solving multi-modal optimization problems are applied, the crucial issue to be solved is maintaining population diversity to avoid drifting and focusing individuals around single global optima. A lot of techniques have been used here so far. Simultaneously for last twenty years a lot of effort has been made in the area of evolutionary algorithms for multi-objective optimization. As the result at least several highly efficient algorithms have been proposed such as NSGAII or SPEA2. Obviously, also in this case maintaining of population diversity is crucial but this time, taking the specificity of optimization in the Pareto sense, there are built-in mechanisms to solve this issue effectively. If so, the idea arises of applying of state-of-theart evolutionary multi-objective optimization algorithms for solving not original multi-modal (but single-objective) optimization task but rather its transformed into multi-objective problem form by introducing additional dispersion-oriented criteria. The goal of this paper is to present some further study in this area