动力系统实测数据相空间重构的改进方法

A New Algorithm of the Phase Space Reconstruction for the Data Obtained in Dynamical Systems

应用本征值分解技术对动力系统实测数据嵌入空间矩阵的本征值进行了计算 ,提出了具体计算嵌入空间矩阵本征值及其本征向量的改进计算方法 ,以及嵌入空间矩阵基的改进选取方法 .并将所得到的动力系统实测混沌数据向已正交化、单位化了的本征向量上投影 ,从而完成了动力系统实测混沌数据的相空间重构工作 .

The new method of the singular value decomposition is used to calculate the eigenvalue of embedding space matrix, and the corresponding algorithm to calculate eigenvectors and to obtain the basis of embedding vector space is put up in this paper. The projection on the orthogonal basis generated by eigenvectors of timeseries data, and concrete paradigms are given the follow. Meanwhile the fundamental work for the state space reconstruction was done already.