C08 - Examination of the performance of PSO algorithm with time–varying population

Title: Examination of the performance of PSO algorithm with time–varying population
Authors: Dimou CK, Charalampakis AE
Type: Conference paper
Conference: 9th HSTAM International Congress on Mechanics
Venue: Limassol, Cyprus
Date: 2010
Language: English

[abstract]

Although an increase of population size usually improves the average performance of the Particle Swarm Optimization (PSO) algorithm, in some cases, however, it can become detrimental on robustness, or it can result to high computational cost. In this study, the relative performance of various PSO variants implementing time-varying population schemes is examined against the PSO algorithm with inertial parameter. The variation of the population is based on the Saw – Tooth oscillation scheme, implemented within various Evolutionary Algorithms (EAs) for demanding optimization problems. The performance is examined for a wide selection of unimodal and multimodal functions.

[cite as]

Dimou CK, Charalampakis AE. Examination of the performance of PSO algorithm with time–varying population. Proc 9th HSTAM International Congress on Mechanics, Limassol, Cyprus; 2010.


[ paper]