摘要
重构相空间对于研究混沌时间序列有着重要的理论与现实意义,目前采用的分别估计嵌入时延和最小嵌入维数的技术路线,割裂了这两个参数所具有的天然联系.为此提出时延法重构相空间的双重构参数联合估计方法,根据两个重构参数的取值标准,利用迭代的方法,同步估计出时延法重构相空间双参数.应用所提出的方法,分别对高斯白噪声和Lorenz系统两个时间序列进行了数值验证,分析表明计算结果是可信的,可以应用于时间序列的相空间重构.
Phase space reconstruction is important in theory and reality for researching into chaotic time series, and at present most of the researchers estimate the delay time and minimum embedding dimension respectively, which dissever the relationship of the two parameters. The new method of optimal phase space dual parameters of delay-time coordinates is put forward. According to the choosing criterion of two reconstruction parameters, this method estimates delay-time coordinates phase space dual parameters in-phase by the iterative approach. A numerical simulation on the Gauss white noise time series and Lorenz system time series is conducted. The results show that the proposed method is effective, accurate and robust for time series phase space reconstruction.
出处
《大连理工大学学报》
EI
CAS
CSCD
北大核心
2009年第4期540-544,共5页
Journal of Dalian University of Technology
关键词
混沌时间序列
相空间重构
嵌入时延
最小嵌入维数
chaotic time series
phase space reconstruction
embedding time delay
minimum embedding dimension
