The implementation frameworks of meta-heuristics hybridization with dynamic parameterization

Abstract

The hybridization of meta-heuristics algorithms has achieved a remarkable improvement from
the adaptation of dynamic parameterization. This paper proposes a variety of implementation
frameworks for the hybridization of Particle Swarm Optimization (PSO) and Genetic
Algorithm (GA) and the dynamic parameterization. In this paper, taxonomy of the PSO-GA
with dynamic parameterization is presented to provide a common terminology and
classification mechanisms. Based on the taxonomy, thirty implementation frameworks are
possible to be adapted. Furthermore, different algorithms that used the implementation
frameworks with sequential scheme and dynamic parameterizations approaches are tested in
solving a facility layout problem. The results present the effectiveness of each tested algorithm in comparison to the single PSO and constant parameterization.

Keywords: hybridization; PSO; GA; implementation frameworks; dynamic parameterization.