Saved in:
| Main Author: | |
|---|---|
| Format: | Artículo científico |
| Language: | en |
| Published: |
Instituto Politécnico Nacional
2012
|
| Subjects: | |
| Online Access: | https://www.redalyc.org/articulo.oa?id=402640460002 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- Constricted Particle Swarm Optimization based Algorithm for Global Optimization Gonzalo Nápoles Isel Grau Rafael Bello Computación Local optima Random Samples Global Optimization Premature Convergence Variable Neighborhoods Particle Swarm Optimization (PSO) is a bioinspiredmeta-heuristic for solving complex global optimization problems.In standard PSO, the particle swarm frequently gets attracted bysuboptimal solutions, causing premature convergence of thealgorithm and swarm stagnation. Once the particles have beenattracted to a local optimum, they continue the search processwithin a minuscule region of the solution space, and escapingfrom this local optimum may be difficult. This paper presents amodified variant of constricted PSO that uses random samples invariable neighborhoods for dispersing the swarm whenever apremature convergence (or stagnation) state is detected, offeringan escaping alternative from local optima. The performance ofthe proposed algorithm is discussed and experimental resultsshow its ability to approximate to the global minimum in each ofthe nine well-known studied benchmark functions. 2012 artículo científico 1870-9044 https://www.redalyc.org/articulo.oa?id=402640460002 en http://www.redalyc.org/revista.oa?id=4026 Polibits application/pdf Instituto Politécnico Nacional Polibits (México) Vol.46