Parameter selection and adaptation in Unified Particle Swarm Optimization

Abstract

The performance of the recently proposed Unified Particle Swarm Optimization method is investigated under different schemes for the determination and adaptation of the unification factor, which is the main parameter of the method, controlling its exploration and exploitation properties. Widely used benchmark problems are employed and numerous experiments are conducted along with statistical tests to yield useful conclusions regarding the effect of the parameter on the algorithm's performance as well as the most efficient adaptation schemes. © 2007 Elsevier Ltd. All rights reserved.