混沌系统延迟时间确定的符号化方法研究

Research on Determination of Delay Time in Chaotic Systems by Symbolic Analysis Method

采用等概率符号分析方法估算混沌系统的互信息,按互信息极小规则确定混沌系统中的最佳延迟时间,分析了在等概率符号化过程中,短序列长度L对符号序列的影响.指出:如果L太小,导致符号序列丢失了一些确定性信息;如果L太大,数据的相关性变差,从而影响最佳延迟时间的确定.通过对Lorenz系统和脑电信号的仿真实验得到的结果表明:短序列长度L影响着最佳延迟时间的确定,且不同系统对于L变化,其延迟时间变化敏感程度不同.

The mutual information of chaotic system was estimated by the equal-probablity symbolic anal- ysis method, and delay time was determined by the rule of minimum mutual information. The effect of the short-sequences length L on symbol sequence was analyzed when the time series was symbolized by equal-prob- ability selection. It was pointed out that some deterministic information would be lost if L was too small, and the relevance of the data would become worse if L was too large, so that the best delay time couldn＇t be determined exactly. The simulation experiments for both Lorenz system and EEG showed that the determination of the optimal delay time was affected by the short-sequences length L, and the sensitivity to delay time was different for the variance of L for different system.