On the premature convergence of particle swarm optimization
Rie B. Larsen, Jérôme Jouffroy, Benny Lassen · 2016
This paper discusses convergence issues of the basic particle swarm optimization algorithm for different parameters. For the one-dimensional case, it is shown that, for a specific range of parameters, the particles will converge prematurely, i.e. away from the actual minimum of the objective function. We illustrate the proposed results on the tuning of a well-known anti-windup technique in a numerical experiment.