Swarm Intelligence

implemented, with ϕ=4.1. One particle starts out at +20 and the other

at −20, with initial velocities of 0.0. As can be seen, in this trial the parti-

cles’ trajectories weave around one another and quickly converge on the

optimum x = 0.0. Naturally, every trial follows a different path; in this

example, the solid-lined particle is initially attracted upward toward the

dotted-lined one’s better value, and then both race down toward the

optimum.

In a typical particle swarm, neighborhood size can vary from three

(the particle and its two neighbors) to the population size (in which case

there is just one neighborhood).

Neighborhood Topology

Human social