Automatic change detection in dynamical system with chaos based on model, fractal dimension and recurrence plot

Abstract

Automatic change detection is the important subject in dynamical systems. There are known techniques for linear and some techniques for nonlinear systems, but merely few of them concern deterministic chaos. This paper presents automatic change detection technique for dynamical systems with chaos based on three different approaches neural network model, fractional dimension and recurrence plot. Control charts are used as a tool for automatic change detection. We consider the dynamical system described by the univariate time series. We assume that change parameters are unknown and the change could be either slight or drastic. Methods are checked by using small data set and stream data. © Springer-Verlag Berlin Heidelberg 2007.