MeSwarm: Memetic particle swarm optimization

Bo Fu Liu*, Hung Ming Chen, Jian Hung Chen, Shiow Fen Hwang, Shinn-Ying Ho

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

5 Scopus citations


In this paper, a novel variant of particle swarm optimization (PSO), named memetic particle swarm optimization algorithm (MeSwarm), is proposed for tackling the overshooting problem in the motion behavior of PSO. The overshooting problem is a phenomenon in PSO due to the velocity update mechanism of PSO. While the overshooting problem occurs, particles may be led to wrong or opposite directions against the direction to the global optimum. As a result, MeSwarm integrates the standard PSO with the Solis and Wets local search strategy to avoid the overshooting problem and that is based on the recent probability of success to efficiently generate a new candidate solution around the current particle. Thus, six test functions and a real-world optimization problem, the flexible protein-ligand docking problem are used to validate the performance of MeSwarm. The experimental results indicate that MeSwarm outperforms the standard PSO and several evolutionary algorithms in terms of solution quality.

Original languageEnglish
Title of host publicationGECCO 2005 - Genetic and Evolutionary Computation Conference
EditorsH.G. Beyer, U.M. O'Reilly, D. Arnold, W. Banzhaf, C. Blum, E.W. Bonabeau, E. Cantu-Paz, D. Dasgupta, K. Deb, al et al
Number of pages2
StatePublished - 1 Dec 2005
EventGECCO 2005 - Genetic and Evolutionary Computation Conference - Washington, D.C., United States
Duration: 25 Jun 200529 Jun 2005

Publication series

NameGECCO 2005 - Genetic and Evolutionary Computation Conference


ConferenceGECCO 2005 - Genetic and Evolutionary Computation Conference
CountryUnited States
CityWashington, D.C.


  • Evolutionary computation
  • Memetic algorithms
  • Numerical Optimization
  • Particle Swarm Optimization
  • Solis and Wets Local Search strategy

Fingerprint Dive into the research topics of 'MeSwarm: Memetic particle swarm optimization'. Together they form a unique fingerprint.

Cite this