A method for distinguishing dynamical species in chaotic time series

Abstract

Two statistical quantities are proposed in this paper,that is Q and R,Q is calle d the exponent of dynamical autocorrelation factor, and R is called the exponent of dynamical cross_correlation factor. Using Q and R,we can identify the dynami cal property of time series or their implicative dynamical structural levels and complexity by estimating the “distance” between diffierent time series or dif fierent segments of same time series.This method has been verified to be valid b y some typical instances.