Research on Phase Space Reconstruction of Nonlinear Time Series

A new method called C-C-2 is presented based on the analysis for the C-C method which is usually used to reconstruct the phase space of chaotic time series. First, the C-C-2 method improves the correlation integral algorithm of time series on the computing way and the parameters. Moreover, the method proposes a new way to determine the optimal delay time window by finding the first local minimum peak value of Scor(t) belonging to the period N of the chaotic time series based on the theory of the chaotic system period N. Finally, the method estimates the optimal delay time only by finding the first local minimum of the average ?S2(t). The experimental results shows that the C-C-2 method are more stable and more appropriate and also improves the robustness of C-C method from 30% to 80%.