Swarm control designs applied to a mechanical oscillator
Abstract
This paper presented the particle swarm optimization approach for nonlinear system identification and for reducing the oscillatory movement of the nonlinear systems to periodic orbits. We analyzes the non-linear dynamics in an oscillator mechanical and demonstrated that this model has a chaotic behavior. Chaos control problems consist of attempts to stabilize a chaotic system to an equilibrium point, a periodic orbit, or more general, about a given reference trajectory. This approaches is applied in analyzes the nonlinear dynamics in an oscillator mechanical. The simulation results show the identification by particle swarm optimization is very effective. © 2008 IEEE.
How to cite this document
Chavarette, Fabio Roberto; Guilherme, Ivan Rizzo. Swarm control designs applied to a mechanical oscillator. 2008 International Conference on Computational Intelligence for Modelling Control and Automation, CIMCA 2008, p. 756-760. Available at: <http://hdl.handle.net/11449/231215>.
Language
English
