Particle Swarm Optimization is a parallel algorithm that spawns particles across a search space searching for an optimized solution. Though inherently parallel, they have distinct synchronizations points which stumbles attempts to create completely distributed versions of it. In this paper, we attempt to create a completely distributed peer-to-peer (P2P) particle swarm optimization in a cluster of heterogeneous nodes. Since, the original algorithm requires explicit synchronization points we modified the algorithm in multiple ways to support a P2P system of nodes. We also modify certain aspect of the basic PSO algorithm and show how certain numerical problems can take advantage of the same thereby yielding fast convergence. This paper is based on one of our earlier work where we demonstrated the use of peer-to-peer systems for single objective optimizations functions. In this paper, we present the modifications that have been made to the previous system and test several benchmark functions.
CITATION STYLE
Dewan, H., Nayak, R. B., & Devi, V. S. (2014). A peer-to-peer single objective particle swarm optimizer. In Advances in Intelligent Systems and Computing (Vol. 258, pp. 689–708). Springer Verlag. https://doi.org/10.1007/978-81-322-1771-8_60
Mendeley helps you to discover research relevant for your work.