Skip to content

Adaptive Heterogeneous Improved Dynamic Multi-Swarm PSO (A-HIDMS-PSO) Algorithm. Source code for the paper: IEEE SSCI https://ieeexplore.ieee.org/document/9660115

Notifications You must be signed in to change notification settings

FTVarna/Adaptive-HIDMS-PSO

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 

Repository files navigation

A-HIDMS-PSO: Adaptive-HIDMS-PSO

HIDMS-PSO Algorithm with an Adaptive Topological Structure.

Paper Link: https://ieeexplore.ieee.org/document/9660115

Cite as:
F. T. Varna and P. Husbands, "HIDMS-PSO Algorithm with an Adaptive Topological Structure," 2021 IEEE Symposium Series on Computational Intelligence (SSCI), Orlando, FL, USA, 2021, pp. 1-8, doi: 10.1109/SSCI50451.2021.9660115.

Abstract:

This paper presents a new variant of the state-of-the-art PSO (particle swarm optimisation) variant HIDMS-PSO (heterogeneous improved dynamic multiswarm PSO) algorithm. The proposed variant improves the fixed topology of the unit structure introduced in the HIDMS-PSO variant. The new master and slave dominated topologies significantly change the dynamics of the unit structure by fluctuating the behavioural heterogeneity in individual units via the use of adaptive topologies. In addition, several existing components in the standard HIDMS-PSO were trimmed down to simplify the algorithm. The efficacy of the adaptive HIDMS-PSO variant was tested by conducting three experiments on the CEC'05 and CEC'17 benchmark test suites at 30 and 50 dimensions using 11 baseline metaheuristics and 15 state-of-the-art PSO variants. The proposed algorithm outperformed all comparison algorithms in all of the conducted experiments. Additionally, adaptive HIDMS-PSO's convergence rate and population diversity maintenance capability were compared with the inertia weight PSO and the standard HIDMS-PSO. The empirical evidence suggests that the proposed algorithm is capable of converging faster to a better solution while mostly maintaining a better population diversity during the search.