低信噪比下的状态空间重构

针对淹没在噪声中的混沌信号的重构问题，基于动力系统理论和对现有重构方法的研究，提出了一种利用具有最大方差的主元子集重构状态空间的方法．同时论述了重构窗口的问题．利用洛伦兹混沌信号叠加白噪声，通过分析奇异谱和相图，对这一重构方法进行了验证 关键词：

STATE SPACE RECONSTRUCTION AT POOR SIGNAL TO NOISE RATIO

For state space reconstruction of the chaotic signals overwhelmed in the noise,based on dynamical systems theory and current methods of reconstruction, we propose a new reconstruction method by using the subset of principal components with the largest variance. We also discuss the reconstruction window. The new approach is verified by analyzing the singular spectra and the phase portraits of the Lorenz′s signal superposed with white noise.