Identificación del sistema de Rössler: Enfoque algebraico y algoritmos genéticos

This article presents a method to determine the parameters of Rössler's attractor in a very approximated way, by means of observations of an available variable. It is shown that the system is observable and identifiable algebraically with respect to the chosen output. This fact allows to construct a differential parametrization of the output and its derivatives. Using this parametrization an identification scheme based on least mean squares is established and the solution is found with a genetic algorithm.