Swarm Intelligence

chapter

eight

Variations and Comparisons

In this chapter, we explore implementa-

tions of the particle swarm paradigm, look

at some variations of the algorithm, ask

whether the particle swarm is an evolution-

ary algorithm, and compare the performance

of various versions of particle swarms for

optimization. This chapter is more technical

than some of the others.

In the “Variations” section, we first take

apart the algorithm to see what makes it tick,

focusing on ways to facilitate convergence

and prevent “explosion” of the swarm. Ap-

proaches include limiting the maximum al-

lowed particle velocity, including V

max,

a con-

striction coefficient, and an inertia