A new approach to nonlinear adaptive prediction of chaotic time series is presented by means of an adaptive sine-Volterra predictive filter based on the deterministic and nonlinear characteristics of chaotic time series. The named adaptive sine-Volterra predictive filter introduces sine function to approximate the Volterra filter based on linearization decomposition. Numerical simulation results show that the nonlinear adaptive filter proposed here has good performance for making adaptive prediction of chaotic time series.
展开▼