RECOGNIZING HIDDEN FREQUENCIES IN A CHAOTIC TIME-SERIES

Using the method of phase space reconstruction (PSR), it is possible t o detect dynamically connected hidden frequency components in a chaoti c time series. Here, we suggest to use the method of singular value de composition (SVD) in tandem with the PSR. With this new scheme, the pr ocedure to search for the hidden frequencies can be simplified substan tially and the search efficiency can be improved greatly. We use the L orenz model, the Rossler model, and the hyperchaotic Rossler model to illustrate how this PSR/SVD scheme works. (C) 1998 Elsevier Science B. V.