离散

Discrete Fourier Transform and Discrete Cosine Transform are used to get the descriptors of lip contours in the unsymmetrical lip contours model automatically. Hidden Markov Model is trained by using both of the descriptors as the eigenvector of lip contours, and then recognition ability is tested.4. Feature fusion is used to improve the classifiable power.

利用离散傅里叶变换和离散余弦变换分别得到描述唇形轮廓特征的傅里叶描述子和离散余弦变换描述子，然后将两类描述子作为唇形轮廓的特征向量，采用隐马尔可夫模型进行学习和识别，分析了两类描述子刻画唇形轮廓特征的能力。4。

Discrete Fourier Transform and Discrete Cosine Transform were used to get the descriptors of lip contours in the asymmetrical lip contour model.Then the Hidden Markov Model was trained using two kinds of descriptors as the eigenvectors of lip contours.

通过离散傅里叶变换和离散余弦变换分别得到描述唇形轮廓特征的傅里叶描述子和离散余弦变换描述子，然后将两类描述子作为唇形轮廓的特征向量，采用隐马尔可夫模型进行学习和识别。

And then, we gain some definitions that in relation to regular surfaces which include discrete Gauss curvature and discrete mean curvature.

由此获得了一系列的相对于光滑曲面的离散定义，如离散平均曲率，离散Gauss曲率等。

For reducing boundary distortion area, the elevations of grid DEM are viewed as a discrete sequence under a generalized coordinate, so that the boundary problem of two-dimensional discrete wavelet transform is turned into the boundary problem of one-dimensional discrete wavelet transform, which not only greatly reduces the boundary distortion area, improves the precision of reconstructed DEM data near the boundary, but also decreases the number of zeros adding on the boundary for wavelet transform, and boosts the compression ratio. On the basis of analyzing the distribution characteristics of coefficients in low and high frequency bands, a mixed entropy coding plan is applied, and a high efficient quantizator called"self-adaptive hard threshold"is established, which is helpful for improving the precision of the reconstructed DEM data and for enhancing compression ratio.

为了减小小波变换山区格网DEM数据压缩中边界失真区域，将山区格网DEM高程值视为广义坐标下的离散随机序列，把二维离散小波变换中的边界问题转换为一维离散小波变换中的边界问题，不仅大大减少了小波变换边界失真区域，提高了重构DEM数据边界点的精度，而且减少了边界补零个数，提高了压缩比；通过对山区格网DEM数据小波变换低频区和高频区变换系数分布特点的分析，采取了霍夫曼—游程混合熵编码方案，设计了高效、低失真的"自适应硬阈值"量化器，即对低频区变换系数不进行量化，直接进行霍夫曼编码；对高频区变换系数则先用"自适应硬阈值"方法量化处理，再进行游程编码。

The dissertation focuses on performance gradient estimation and SRO algorithm of DES aswell as other related topics,such as modeling of DES and discrete event simulation,performancemeasure Hessian matrix estimation,Stochastic Approximationand so on.General Semi-Markov Processis adapted to describe behaviour of DES and discrete eventsimulation.A general result is obtained and applied to analyze strong consistency of performancegradient Perturbation Analysisestimator.Conditions for effectiveness of InfinitesimalPerturbation Analysisare also discussed.New PA algorithms,such as server PA,PA usingDirac generalized function,job sequence PA etc.,are proposed to deal with discreteparameters.Furthermore,strategies of combining PA algorithms with certain iterative rules areanalyzed widely in order to optimize DES parameters and some SRO algorithms are constructed toslove DES optimization problems.

论文将DES性能分析领域中性能梯度估计和单运行仿真优化作为主要的研究对象，探讨了与性能梯度估计有关的DES和离散事件仿真过程建模、高阶导数估计以及随机逼近等理论和方法；建立了具有一般性的DES和离散事件仿真的广义半Markov过程模型；获得了分析性能梯度估计量强相合性的理论结果和无穷小摄动分析算法应用的条件；提出了诸如服务台数目的摄动分析、奇异摄动分析、串联生产线和加工车间中工件加工序的摄动分析等一系列摄动的思想和算法，对连续参数和离散参数都进行了深入的研究；对摄动分析算法同优化规则结合实现仿真优化的策略也进行了广泛的分析，构造了新型的变界截尾的随机逼近算法，并用于DES的单运行仿真优化研究；最后，给出了DES性能分析方法的实证性研究结果。

In this paper a discretization algorithm based on importance of cut point. Firstly, the paper gives some concept of rough set theory:decision table,indiscernible relation,information entropy and condition entropy.

粗糙集理论可以用在数据挖掘的各个阶段，该文研究了粗糙集理论在数据挖掘的数据离散化阶段的应用，并提出了一个新的决策表离散化算法。l 基本概念及离散化问题的描述定义1 。

The digital algorithms of integral and discrete wavelet transform are presented for purpose of electric power system protection, based on the analysis of wavelet transform for sampled digital signals under the Shannon sampling theorem.

从电力系统继电保护应用的角度，结合香农采样定理，深入探讨了连续信号和离散信号的小波变换的关系，给出了离散信号的积分小波变换和离散小波变换的数字计算方法。从理论上研究了信号小波变换的误差与尺度函数和小波函数的关系。

To the discrete-time systems, we introduce a new discrete-time norm, the algorithm given is not depend upon the initial data. For a class of uncertain discrete-time systems, we also adopt the method of iterative learning control, we convert the problem of learning control into a global stability of trivial solution for lower triangular matrix dynamic systems, and obtain the result of tracking to desired state of the system.

对于一般非线性离散系统，引进了离散的范数，所给的算法不依赖于学习控制算法中的初始数据，对于一类不确定离散系统的状态跟踪问题，我们也采用迭代学习控制的方法，将相应的学习控制问题转化为一个下三角动力系统的平凡解全局稳定性问题，获得了对系统理想状态的跟踪结果，并对所给的算法的收敛性进行了分析。

For pratical reasons, the method of increasing the number of iteration m so as to enhance the security can only be used in the case that parameters can match exactly, like a digital encryption system.

本文对以上的离散实际混沌加密系统采用多精度自扰动的离散化策略，从而避免了传统的离散化方法所带来的缺陷，使这类数字混沌加密系统走向实用。

The convection of onflow mode with ROE format disperse, diffuse with the center difference divides format disperse, the control equation after dimensional disperse advances a format to seek solution with explicit time of much pace Runge-Kutta.

湍流模式的对流项用ROE格式离散，扩散项用中心差分格式离散，空间离散后的控制方程用多步Runge-Kutta显式时间推进格式求解。

The split between the two groups can hardly be papered over.

这两个团体间的分歧难以掩饰。

This approach not only encourages a greater number of responses, but minimizes the likelihood of stale groupthink.

这种做法不仅鼓励了更多的反应，而且减少跟风的可能性。