能谱涨落统计特征分析在Logistic混沌序列上的应用

混沌现象普遍存在于现实生活中,混沌时间序列的分析是非线性系统研究的主要内容.基于Anderson 模型,提出了一种新的混沌序列的分析方法.通过对Logistic序列的分析,使用能谱涨落分析方法,发现能级斥力的大小能与最大lyapunov指数相对应.结果表明,该方法能较好的预测混沌序列,而且具有较强的抗噪声的能力,不受初值敏感性的影响.

Spectral analysis approach to chaotic series

Chaos is ubiquitous in real life. Identification of chaotic time series is the main content of nonlinear field. We map a time series to a 1-dimensional random lattice. By using random matrix theory, we find that the statistical characteristics of level spacing are consistent very well with the lyapunov exponent behaviors. The result show that the method has strong anti-noise ability and non-initial-state-sensitivity.