Existing algorithms for filtering noise from contaminated chaotic signals require relatively high signal-to-noise ratios (SNRs). A preprocessing method to be used in conjunction with existing algorithms is proposed for initial noise reduction in severe SNRs. This preprocessing method is applicable to noisy data from any nonlinear dynamical system because no knowledge of the system dynamics is required. [References: 7]
展开▼
