Title :
Simple adaptive cognition for PSO
Author :
Monson, Christopher K.
Author_Institution :
Google, Inc., Pittsburgh, PA, USA
Abstract :
A simple, effective, computationally cheap, and easily tuned method is presented for improving PSO performance by automatically adapting acceleration coefficients. While this approach can be shown to be effective on its own as a source of swarm diversity on difficult functions, it is also capable of enhancing other adaptive strategies commonly employed with PSO. Significantly, and unlike many other PSO enhancements designed to improve swarm diversity, this approach does not typically harm the performance of the underlying method, allowing it to work well on easy and difficult functions alike.
Keywords :
cognition; particle swarm optimisation; PSO enhancement; PSO performance; adaptive strategy; automatically adapting acceleration coefficient; simple adaptive cognition; swarm diversity; tuned method; Acceleration; Cognition; Convergence; Equations; Schedules; Topology; Tuning;
Conference_Titel :
Evolutionary Computation (CEC), 2011 IEEE Congress on
Conference_Location :
New Orleans, LA
Print_ISBN :
978-1-4244-7834-7
DOI :
10.1109/CEC.2011.5949814