Biblio

Swarm intelligence is rather a simple implementation but has a good performance in function optimization. There are a variety of instances of swarm model and has its inherent dynamic property. In this study we consider a hybrid swarm model where agents complement each other using its native property. Employing popular swarm intelligence model Particle swarm and Firefly we consider hybridization methods in this study. This paper presents a hybridization that agents in Particle swarm selected by a simple rule or a random choice are changing its property to Firefly. Numerical studies are carried out by using complex function optimization benchmarks, the proposed method gives better performance compared with standard PSO.