Particles prefer walking along the axes: experimental insights into the behavior of a particle swarm

Authors:
Manuel Schmitt;Rolf Wanka
Affiliations:
University of Erlangen-Nuremberg, Erlangen, Germany;University of Erlangen-Nuremberg, Erlangen, Germany
Venue:
Proceedings of the 15th annual conference companion on Genetic and evolutionary computation
Year:
2013

Citing 3
Cited 0

The particle swarm - explosion, stability, and convergence in amultidimensional complex space

IEEE Transactions on Evolutionary Computation
Biases in Particle Swarm Optimization

International Journal of Swarm Intelligence Research
Particle swarm optimization almost surely finds local optima

Proceedings of the 15th annual conference on Genetic and evolutionary computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

We study the frequently observed phenomenon of stagnation in the context of particle swarm optimization (PSO). We show that in certain situations the particle swarm is likely to move almost parallel to one of the axes, which may cause stagnation. We provide an experimentally supported explanation in terms of a potential of the swarm and are therefore able to adapt the PSO algorithm slightly such that this weakness can be avoided.