Swarm Intelligence Algorithms

Source-code of global version of PSO algorithm in Matlab 269

distances between all particles in the swarm (for example using Euclidean

distance) and then for each particle P

i

we select the best particle Lbest

i

among

the m nearest neighborhoods of particle P

i

. The social neighborhood is defined

on the basis of a topology which depends only on the particle index i [4]. In

practical application, the neighborhood of each particle is defined (at the start

of the PSO algorithm) as a list of particles which does not change with the

iterations. The topology which is more commonly used is a ring with a k = 2

(each particle has two neighbors – first on the left