混沌背景中信号参数估计的新方法

怎样提取混沌背景中信号的参数具有重要意义.在重构的相空间中,叠加有其他信号的混沌信号时间序列重构的点集会偏离混沌吸引子所在光滑流形,依据这一性质并综合利用混沌背景中信号本身的特性,提出一种参数估计的新方法：最小相对奇异值(MRSV)法.该方法先建立逆滤波器,由其输出重构相空间,然后改变其参数,使输出信号在嵌入空间中作局部奇异值分解的相对奇异值最小,来实现参数估计.AR模型参数和正弦信号频率估计的仿真结果验证了该方法的有效性. 关键词： 混沌 参数估计 最小相对奇异值(MRSV) 逆滤波器     

道路交通流状态的多参数融合预测方法

针对道路交通流普遍存在的混沌特性以及单交通参数不足 以全面反映交通流状态的实际情况,考虑交通动力学系统中多个 交通参数之间的关联关系,提出一种新的多参数混沌时间序列预 测算法.该算法在相空间重构理论的基础上,借助Bayes估计将多个参数在 同一高维相空间中进行相点最优融合,从而增加重构相空间的系统信息量, 使得相空间的相点轨迹更加逼近原交通系统的动力学行为.同时借鉴单 参数混沌时间序列预测方法,从不同角度对动力学系统的运动状态进行描 述,以实现多参数时间序列的混沌预测.实验结果表明,通过融合多交通参数时 间序列,获得了更加完整的交通流状态变化特征.与单交通参数时间序列的预测 结果相比,其预测误差显著降低,均衡系数相应增大,提高了交通流状态预测的准确率.     

输油管道压力时间序列混沌特性研究

输油管道压力波动的内在动态特征,采用非线性的分析方法探讨了实测输油管道压力波动中存在混沌动态特性的可能性. 以6个典型的输油管道压力实测数据集为研究对象,重构相空间、求得了分形维数和Lyapunov指数谱、验证了数据的平稳性和非线性. 通过对所得结果的分析,从理论上证明了输油管道压力信号具有严格混沌动态特性. 为基于输油管道压力时间序列的研究提供了混沌理论基础. 关键词： 压力时间序列 Lyapunov指数 分形维数 混沌     

基于条件熵扩维的多变量混沌时间序列相空间重构

提出一种多变量混沌时间序列相空间重构的条件熵扩维方法.首先使用互信息法求解每个变量的时间延迟,其次按条件熵最大原则逐步扩展相空间的嵌入维数,使得重构坐标从低维到高维的转换保持较强的独立性,最终的重构相空间具有较低的冗余度,为多变量时间序列的预测构造了有效的模型输入向量.通过对几个经典多变量混沌时间序列进行数值实验,结果表明该方法比单变量预测和已有多变量预测方法具有更好的预测效果,说明了该重构方法的有效性. 关键词： 多变量混沌时间序列 相空间重构 条件熵 神经网络预测     

多元混沌时间序列的多核极端学习机建模预测

多元混沌时间序列广泛存在于自然、经济、社会、工业等领域. 对多元混沌时间序列进行建模预测有助于人类更好地管理, 控制与决策. 针对多元混沌时间序列的建模预测问题, 本文提出一种基于多核极端学习机的预测方法. 首先对多元混沌时间序列进行相空间重构, 将多元混沌时间序列序列的时间相关性转化为空间相关性. 提出一种结合多核学习算法与核极端学习机模型的多核极端学习机建立相空间中输入输出数据的非线性映射. 多核极端学习机模型结合了多核学习算法的数据融合能力以及核极端学习机的训练简便优势. 基于Lorenz混沌时间序列预测和San Francisco河流月径流量预测的仿真实验表明, 与其他常见混沌时间序列预测方法相比, 本文提出的基于多核极端学习机的多元混沌时间序列预测方法具有更小的预测误差.     

基于滑动窗口子空间的混沌信号降噪算法

基于子空间投影理论，利用混沌信号特性，提出了一种基于滑动窗口子空间投影的混沌信号降噪算法，通过相空间重构和多次迭代运算，较好地解决了低信噪比下的混沌信号降噪问题。对算法的复杂性作了研究，并对重构维数、延迟时间、迭代次数等参数的选择进行了讨论。利用Lorenz模型进行仿真得到了比较好的效果，与传统算法相比，新算法具有降噪效果较好、局部特性损失较少、精度较高等优点。     

基于小波变换的混沌信号相空间重构研究

应用小波变换和非线性动力学方法研究了混沌信号在相空间中的行为，指出混沌时间序 列的小波变换实质上是在重构的相空间中，混沌吸引子向小波滤波器向量所张的空间中的投 影，与Packard等人提出的相空间重构方法本质上是一致的.实验结果表明，混沌信号经过 小波变换后，吸引子轨迹与原有轨迹具有相似的结构，同时，系统的关联维数、Kolmogorov 熵等非线性不变量仍然得到保留.这些结果表明，利用小波变换研究混沌信号是有效的. 关键词： 小波变换 相空间重构 混沌信号 脑电信号     

基于数据融合的多变量相空间重构方法

针对单变量时间序列和多变量时间序列相空间重构所存在的问题,提出一种新的多变量融合的相空间重构方法. 通过Bayes估计理论,将多变量在同一相空间中进行相点的最优融合,得到了更为理想的融合相空间. 应用所提出的方法对Lorenz系统及耦合Rssler系统进行了多变量融合的相空间重构. 通过多变量重构图与单变量重构图的比较,发现基于数据融合的多变量相空间重构图包含了所有单变量相空间重构图的重要信息,使重构的相空间更加完备,较全面地反映出吸引子的全貌信息. 最后应用该方法对转子油膜涡动故障得到的多变量时间序列 关键词： 多变量时间序列 相空间重构 数据融合 Bayes估计     

基于径向基神经网络预测的混沌时间序列嵌入维数估计方法

根据Takens定理,研究了混沌时间序列相空间重构嵌入维数的选取问题.提出了基于径向基函数神经网络预测模型性能的嵌入维数估计方法,即根据嵌入维数与混沌时间序列预测模型性能的变化关系来确定嵌入维数.通过对几种典型混沌动力学系统的数值验证,结果表明该方法能够确定出合适的相空间重构嵌入维数. 关键词： 混沌 相空间重构 嵌入维数 预测     

两种实用的相空间重构方法

研究了目前流行的几种相空间重构技术,分析了它们存在的不足,在此基础上提出了两种适合于不同情况下的重构方法,并初步分析了它们重构的质量.在仿真实验中,由于两种方法都运用了小波降噪技术而具有一定的降噪能力,同时还可以看出第一种方法比第二种方法好.因此,在通常情况下,可以用第一种方法对信号进行重构.而在工程应用中,如需要重构的嵌入维较小时,可用第二种方法来对信号进行重构 关键词： 混沌 相空间重构 最大李雅普诺夫指数     

基于信息准则的局域预测法邻近点的选取方法

基于信息准则,提出了选取局域预测法中邻近点个数的定量方法. 并用此方法分析Lorenz模型生成的混沌时间序列和Santa Fe时间序列竞赛的激光数据(Data A). 实验结果表明用该方法选取邻近点的局域预测法的一步和多步预测性能较好,在满足预测精度较高的条件下,计算量较小. 关键词： 邻近点 邻域 局域预测 信息准则     

地球变化磁场Z分量的混沌动力学特性分析

对地球变化磁场进行了混沌动力学特性分析,主要进行了以下两个方面的工作:1)运用多种方法对多个地磁台站同一时段变化磁场分量数据(分为磁扰较小和磁扰较大时段)进行了混沌特性分析,主要是为了以更充分的判据来验证地球变化磁场是否具有混沌特性;2)同一地磁台站不同时段变化磁场分量数据的混沌特性分析,主要是验证地球变化磁场是否具有变参数混沌特性.分析结果表明:1)变化磁场时间序列确实具有混沌特性;2)从混沌时间序列相空间重构的角度可初步认为模拟地球变化磁场动力学系统所必须的独立变量应该为6左右;3)变化磁场时间序列不仅具有混沌特性而且其参数也是在缓慢变化着的,具有变参数混沌系统的特性.     

混沌时间序列的局域高阶Volterra滤波器多步预测模型

依据相空间邻近轨道演化相似性特点建立训练模式,提出了基于自适应高阶非线性Volterra滤波器（HONFIR）的混沌时间序列多步预测模型（MSP-HONFIR）;通过定义距离相似度、趋势相似度来衡量轨道演化相似度,提出了混沌吸引子邻近轨道判别的新方法;从模型训练充分性角度出发探讨了MSP-HONFIR滤波器模型训练集规模控制的依据.数值研究表明MSP-HONFIR滤波器模型的多步预测性能优于原有HONFIR滤波器模型. 关键词： 混沌 非线性自适应预测 Volterra滤波器模型 训练模式     

Using chaos to generate variations on movement sequences

We describe a method for introducing variations into predefined motion sequences using a chaotic symbol-sequence reordering technique. A progression of symbols representing the body positions in a dance piece, martial arts form, or other motion sequence is mapped onto a chaotic trajectory, establishing a symbolic dynamics that links the movement sequence and the attractor structure. A variation on the original piece is created by generating a trajectory with slightly different initial conditions, inverting the mapping, and using special corpus-based graph-theoretic interpolation schemes to smooth any abrupt transitions. Sensitive dependence guarantees that the variation is different from the original; the attractor structure and the symbolic dynamics guarantee that the two resemble one another in both aesthetic and mathematical senses. (c) 1998 American Institute of Physics.     

Exponential sine chaotification model for enhancing chaos and its hardware implementation

Chaotic systems have been intensively studied for their roles in many applications, such as cryptography, secure communications, nonlinear controls, etc. However, the limited complexity of existing chaotic systems weakens chaos-based practical applications. Designing chaotic maps with high complexity is attractive. This paper proposes the exponential sine chaotification model (ESCM), a method of using the exponential sine function as a nonlinear transform model, to enhance the complexity of chaotic maps. To verify the performance of the ESCM, we firstly demonstrated it through theoretical analysis. Then, to exhibit the high efficiency and usability of ESCM, we applied ESCM to one-dimensional (1D) and multi-dimensional (MD) chaotic systems. The effects were examined by the Lyapunov exponent and it was found that enhanced chaotic maps have much more complicated dynamic behaviors compared to their originals. To validate the simplicity of ESCM in hardware implementation, we simulated three enhanced chaotic maps using a digital signal processor (DSP). To explore the ESCM in practical application, we applied ESCM to image encryption. The results verified that the ESCM can make previous chaos maps competitive for usage in image encryption.     

Ehrenfest Time at the Transition from Integrable Motion to Chaotic Motion

Ehrenfest time depends differently on the Planck constant in integrable motion and chaotic motion. We study how its dependence on the Planck constant changes when there is a continuous transition from regular motion to chaotic motion. We find that the dependence is a weighted compromise between its two distinct dependences in regular and chaotic motions. The study is carried out with the system of periodically driven anharmonic oscillator. As the system is quite typical, the result may apply generally.     

Global behavior of gear system using mixed cell mapping

In some mechanical nonlinear systems, the transient motion will be undergoing a very long process and the attractor-basin boundaries are so complicated that some difficulties occur in analyzing the system global behavior. To solve this problem a mixed cell mapping method based on the point mapping and the principle of simple cell mapping is developed. The algorithm of the mixed cell mapping is studied. A dynamic model of a gear pair is established with the backlash, damping, transmission error and the time-varying stiffness taken into consideration. The global behaviors of this system are analyzed. The coexistence of the system attractors and the respective attractor-basin of each attractor with different parameters are obtained, thus laying a theoretical basis for improvement of the dynamic behaviors of gear system.     

Global behavior of gear system using mixed cell mapping

In some mechanical nonlinear systems, the transient motion will be undergoing a very long process and the attractor-basin boundaries are so complicated that some difficulties occur in analyzing the system global behavior. To solve this problem a mixed cell mapping method based on the point mapping and the principle of simple celll mapping is developed. The algorithm of the mixed cell mapping is studied. A dynamic model of a gear pair is established with the backlash, damping, transmission error and the time-varying stiffness taken into consideration. The global behaviors of this system are analyzed. The coexistence of the system attractors and the respective attractor-basin of each attractor with different parameters are obtained, thus laying a theoretical basis for improvement of the dynamic behaviors of gear system.

     

Wave chaos and mode-medium resonances at long-range sound propagation in the ocean

We study how the chaotic ray motion manifests itself at a finite wavelength at long-range sound propagation in the ocean. The problem is investigated using a model of an underwater acoustic waveguide with a periodic range dependence. It is assumed that the sound propagation is governed by the parabolic equation, similar to the Schrodinger equation. When investigating the sound energy distribution in the time-depth plane, it has been found that the coexistence of chaotic and regular rays can cause a "focusing" of acoustic energy within a small temporal interval. It has been shown that this effect is a manifestation of the so-called stickiness, that is, the presence of such parts of the chaotic trajectory where the latter exhibit an almost regular behavior. Another issue considered in this paper is the range variation of the modal structure of the wave field. In a numerical simulation, it has been shown that the energy distribution over normal modes exhibits surprising periodicity. This occurs even for a mode formed by contributions from predominantly chaotic rays. The phenomenon is interpreted from the viewpoint of mode-medium resonance. For some modes, the following effect has been observed. Although an initially excited mode due to scattering at the inhomogeneity breaks up into a group of modes its amplitude at some range points almost restores the starting value. At these ranges, almost all acoustic energy gathers again in the initial mode and the coarse-grained Wigner function concentrates within a comparatively small area of the phase plane.     

基于Coulomb摩擦效应的一类非线性相对转动系统的混沌研究

研究一类非线性相对转动系统在负载Coulomb摩擦效应下的混沌运动行为. 根据Lagrange方程建立一类含非线性负载Coulomb摩擦阻尼的两个质量相对转动系统的动力学方程. 利用Cardano公式讨论自治系统的特征值, 在此基础上, 应用待定系数法给出系统同宿轨道的存在性, 并借助Silnikov定理研究了系统的混沌行为. 最后数值模拟了给定参数下系统的混沌运动, 并给出在Coulomb摩擦阻尼变化下系统由周期、倍周期通向混沌的途径, 验证了理论分析的正确性.