Optimization methods

 

 

Date

05/11/2007

Author

F. Moussouni, S. Brisset, P. Brochet

Affiliation

L2EP EC Lille France

Email

fouzia.moussouni@ec-lille.fr

Method

Genetic Algorithm

References

[1] F. Moussouni, S. Brisset, P. Brochet, « Comparison of two multi-agent algorithms: ACO and PSO», ISEF 2007- 13th International Symposium on Electromagnetic Fields In Mechatronics, Electrical and Electronic Engineering, Prague, Czech Republic, September 13-15, 2007.

Description of the method

Particle swarm optimization (PSO) is inspired by social behavior of bird flocking developed by Eberhart and Kennedy in 1995.

PSO is based on the concept of cooperation of agents (particles), which can be seen like rustic animals, having little memory and faculties of reasoning. The exchange of information between these rudimentary agents enables them nevertheless to acquire an overall astute behavior to be able to solve hard problems.
The PSO algorithm is based on two rules:

  1. each particle has a memory that enables him to memorize its best position found in the past and it tends to be attracted by this point.
  2. each particle is informed about the best position find by its vicinity and it tends to go towards this point.

In other words, starting from some information, particles must be able to choose its next movement, i.e., to calculate its new velocity which is an updating operator for its position.

  

Back