本文将复杂非线性时间序列分析中的系统建模（ 或参数估计） 与噪声抑制这两个通常是分别或顺序完成的任务进行综合处理，提出一种交叉参考的统一框架，通过各自迭代算法中的交叉参考和积极作用改善性能．基于这一框架，建立了一个用于混沌时间序列分析的交叉参考迭代算法，仿真结果表明了该方法对性能的显著改善．根据不同的优化准则，还可获得各种变型算法 In this dissertation, the system modeling (or parameter estimation) and noise suppression in complex non-linear time series analysis, which are usually completed separately or in sequence, are comprehensively processed. A cross-reference unified framework is proposed. Through the respective iterative algorithms Cross-reference and positive effects improve performance. Based on this framework, a cross-reference iterative algorithm for chaotic time series analysis is established. The simulation results show that this method can significantly improve the performance. Depending on the optimization criteria, various variant algorithms are also available