Adaptive particle swarm optimization: Detection and response to dynamic systems

Abstract

This paper introduces an adaptive PSO, which automatically tracks various changes in a dynamic system. Different environment detection and response techniques are tested on the parabolic and Rosenbrock benchmark functions, and re-randomization is introduced to respond to the dynamic changes. Performance on the benchmark functions with various severities is Anal.yzed. © 2002 IEEE.