Partially connected topologies for particle swarm

Authors:
Carlos M. Fernandes;Juan Julián Merelo;Juan L.J. Laredo;Carlos Cotta;Agostinho C. Rosa
Affiliations:
University of Granada,, Granada, Spain & Technical University of Lisbon, Lisbon, Portugal;University of Granada, Granada, Spain;University of Luxembourg, Luxembourg, Luxembourg;University of Malaga, Malaga, Spain;Technical University of Lisbon, Lisbon, Portugal
Venue:
Proceedings of the 15th annual conference companion on Genetic and evolutionary computation
Year:
2013

Citing 3
Cited 0

Population structure and particle swarm performance

CEC '02 Proceedings of the Evolutionary Computation on 2002. CEC '02. Proceedings of the 2002 Congress - Volume 02
Population structure and particle swarm performance

CEC '02 Proceedings of the Evolutionary Computation on 2002. CEC '02. Proceedings of the 2002 Congress - Volume 02
Comprehensive learning particle swarm optimizer for global optimization of multimodal functions

IEEE Transactions on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

The effects of dynamic and partially connected 2-dimensional topologies on the particle swarm are studied. The particles are positioned on 2-dimensional grids of nodes, where they move according to a simple rule. The von Neumann neighborhood is used to decide which particles influence each individual. Structures with growing size are tested on a classical benchmark. The partially connected grids with von Neumann neighborhood structure perform more consistently than other strategies.