基于重构相空间的延迟时间和嵌入维数这两个参数的选取不相关的观点,提出用互信息函数法确定延迟时间后,用CAO方法来确定嵌入维数的新思路.通过对几种典型的混沌动力学系统的数值验证,结果表明该方案能够确定出相空间重构的有效延迟时间和最佳嵌入维数.该方法能够从时间序列中有效地重构原系统的相空间,是混沌信号识别的一种有效途径.
The choice of delay time and embedding dimension have no co rrelation in the process of phase-space reconstruction. Based on this principle, a new idea for determining the embedding dimension is proposed. It uses CAO theory after determining the delay time by the mutual information function. Numerical simulations show that the method is applicable for determining the appropriate delay time and the best embedding dimension. The method can recover the original phase space from the time series effectively, and can be used as an effective method to identify the chaotic signal.