Abstract：Based on the phase space reconstruction of chaotic time series, an adaptive prediction method of chaotic vibration is presented and the adaptive algorithm of model parameters is derived. The effect of the neighborhood size, embedding dimension on the prediction is studied and results show that when the parameters are reasonable, the method can be use to predict chaotic time series efficiently. Finally, this method is applied to predict experimental signal of the X- type nonlinear vibration isolation system successfully.
2009, Vol. 28 
