Scalability study of particle swarm optimizers in dynamic environments

Authors:
Barend J. Leonard;Andries P. Engelbrecht
Affiliations:
Department of Computer Science, University of Pretoria, South Africa;Department of Computer Science, University of Pretoria, South Africa
Venue:
ANTS'12 Proceedings of the 8th international conference on Swarm Intelligence
Year:
2012

Citing 4
Cited 0

Dynamic Search With Charged Swarms

GECCO '02 Proceedings of the Genetic and Evolutionary Computation Conference
Particle swarms and population diversity

Soft Computing - A Fusion of Foundations, Methodologies and Applications
Heterogeneous particle swarm optimization

ANTS'10 Proceedings of the 7th international conference on Swarm intelligence
Multiswarms, exclusion, and anti-convergence in dynamic environments

IEEE Transactions on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

This study investigates the scalability of three particle swarm optimizers (PSO) on dynamic environments. The charged PSO (CPSO), quantum PSO (QPSO) and dynamic heterogeneous PSO (dHPSO) algorithms are evaluated on a number of DF1 and moving peaks benchmark (MPB) environments that differ with respect to the severity and frequency of change. It is shown that dHPSO scales better to high severity and high frequency DF1 environments. For MPB environments, similar scalability results are observed, with dHPSO obtaining the best average results over all test cases. The good performance of dHPSO is ascribed to its ability to explore and exploit the search space more efficiently than CPSO and QPSO.