Quantum Particle Swarm Optimization: Performance analysis for various particle neighborhood topologies

EasyChair Preprint 2515

Abstract

Developed in 1995 by Kennedy and Eberhart, Particle Swarm Optimization (PSO) is a population-based metaheuristic inspired by the movements of bird swarms. It has the advantage of being able to solve non-derivable problems and to be efficient in global search. However, it has some disadvantages, such as a weakness in local search, observable when the velocity of the particles is high (due to a poor adaptation of the parameters controlling the velocity), a risk of premature convergence (particularly when the global particle remains static) as well as parameters that are difficult to set. We had previously proposed a new variant of PSO, called QUAntum Particle Swarm Optimization (QUAPSO) in order to solve some of these problems. In this paper, we propose to evaluate the performance of QUAPSO for various particle neighborhood topologies.  

Keyphrases: Algorithme auto-adaptatif, Intelligence en essaim, Optimisation par essaims particulaires

@booklet{EasyChair:2515,
  author    = {Arnaud Flori and Hamouche Oulhadj and Patrick Siarry},
  title     = {Quantum Particle Swarm Optimization: Performance analysis for various particle neighborhood topologies},
  howpublished = {EasyChair Preprint 2515},
  year      = {EasyChair, 2020}}