多尺度动力学机制下的时变信号分析方法研究

Multiple time-scales dynamics embedded in rhythmic process, multiple.time-scales dynamical models and its design methods were researched in this project.The main contributes are listed. First, several methods have been proposed for the analyzing multi time-scales process especial for cardiac rhythm. On the basis of the presented methods, cardiac chaos was stated in a reliable way. Second, on the basis of time-scale transformation, a so-called Lissajous chaotic map was proposed, which.break through the limit for one dimensional chaotic maps that the map must be described by nonlinear function. Third, several evolutionary computation based.algorithms have been proposed for designing multiple time-scales dynamical model.We expect the obtained achievements will be useful in the area of biomedical.engineering and net information security.

将多尺度动力学概念引入非线性信号分析领域，创建基于动力学尺度信息补偿的吸引子重构策略、动力学机制下不同尺度上信息提取准则和方法为重要组成部分的多尺度动力学时变信号分析方法。在挖掘一类时变信号中尚未被认识的客观规律、突破当前生理节律信号非线性分析遇到的困难以及开拓新信号分析方法方面均具有极其重要的理论和实际意义。