查询词典 method

This article tries to employ the angles of historical amparalive method ,stalistical method,drawing method,and the analytical method of sense of a word and so on ,and adopt the theories from means. dialectology,cultural anthropology and sociaology,etc.

本文试图从历史比较法、统计法、图表法、词义分析法等多个角度，采用语言学、方言学、文化人类学、社会学等理论，多视角地对罗布方言词汇进行了比较深入的探讨。

Thirdly, in order to improve the learning efficiency and stability of feedforward neural network, a fast learning algorithm for neural networks based on CGM-OC approach is presented. Compared with other learning methods such as BP method, Broyden Flecher Goldfarl Shanno method and Powell method in which the derivative is replaced by the difference quotient etc, simulation results show that the proposed method is an efficient and fast one.

第三，为了提高前馈神经网络权值的学习效率及稳定性，提出一种基于正交校正共轭梯度优化方法的神经网络学习算法，通过与其它学习算法（如：BP算法、变尺度法、用差商近似代替导数的Powell法等）的比较，经仿真试验表明，本算法是一种高效、快速的学习算法。

Based on the theories of the first two chapters, chapter 5 introduces the theoretical foundation and method of the software design of the license plate recognition system in detail. In this system, a license plate is fixed in a position and cut apart by verge examination based on the differential operator and threshold algorithm. Then outline method and projection method are used as methods of characteristic extracting, lastly neural network is used as a method to recognition the single character.

第五章是在前两章的理论基础上，详细介绍了车牌识别系统软件设计的实现方法和实验结果，应用图像分割的理论对车牌进行边缘检测和提取，利用轮廓和投影的方法提取字符的特征，最后用神经网络的方法对所得字符进行识别。

A fuzzy predictive control method based on improved T-S fuzzy modeling method thus is acquired. A series of predictive model is obtained by improved T-S fuzzy modeling method. The model includes input and output data before current time and future input value. This method does not need to solve Diophantine equation and nonlinear programming.

并在此基础上给出了一种基于改进递阶聚类模糊T-S模型建模方法的模糊预测控制算法，根据预测控制中的预测步长，建立多个包括当前时刻以前的输入输出数据、未来时刻的输入量的模糊预测模型，可以在不求解Diophantine方程的情况下通过最优化性能指标得到预测控制律。

During the calculation the results got by quintic spline kernel was the best so it was chosen as the proper kernel for the SPH simulation. By comparison of the simulation results got by SPH and finite volume method, it was proved that the buffer cells method proposed in the SPH simulation to implement the Dirichlet boundary conditions was accurate, and this is of great importance for the simulation of flows pass a pipe. At last, the SPH method was improved to simulate the liquid-solid two phase flows and the results show that this method was appropriate for the simulation of flows containing viscous fluid and macroscopic solid components. Now the two methods are both able to predict the continuum flows and rarefied flows severally, but the hybrid program still has some aspects to be improved.

首先用标准算例验证了SPH方法的理论模型和计算程序的准确性；通过比较确定选用五次样条核作为计算低雷诺数下的粘性连续流动时的核函数；采用SPH方法和有限体积法分别对同一管道流动进行了对比计算，证明了本文提出的边界缓冲区方法在实现SPH方法中的流体边界条件时的有效性，解决了SPH方法模拟管道流动时的一个关键问题；巧妙地使用SPH方法对化工分离流动中常见的固液两相流及固液耦合作用进行了模拟尝试，获得了较好的结果；在跨尺度计算方面，目前两个子程序已经能够分别模拟连续的和稀薄的二维管流问题，二者的耦合计算程序还处于调试和优化阶段，各个参数对于计算精度和计算速度的影响规律仍在探索中。

Firstly, the conception of network system reliability and method of calculation reliability for network system are introduced in the thesis. Secondly, calculation method analysis and detailed realization of searching node method and method of calculation of disjoint form minimal path sets are the emphases of this paper.

本文首先介绍了网络系统的概念和网络系统可靠性分析的相关概念及方法，然后着重于本文要采用的节点遍历法和最小路集不交化方法做了算法分析和具体实现，最后将计算程序简单界面化并加以实际应用。

The second one is a method based on Independent Component Analysis with reference signal. Unlike the former, this method can separate the estimating value of the watermark signal separately by using the information provided by the reference signal, so the computing complexity of the first method can be greatly decreased, at the same time, the shortcomings of the ICA based method, such as disorderliness and randomicity can be overcome.

第二种方法以带参考信号的独立分量分析算法为基础，通过参考信号提供的信息直接提取所需要的水印信号估计值，克服了ICA算法在分离各路源信号时所具有的无序性和任意性的缺点，降低了第一种方法的运算复杂度。

By means of the projection method,adding a small disturbing function to the objective function, the penalty method, andthe gradient method, this dissertation proposes several neural networks for solvingsaddle point problems and minimum radius of ball-coverings problems, and investi-gates the stability of the proposed networks by the LaSalle invariance principle andthe Lyapunov method, so as to design neural networks which avoid getting stuck inthe local minimum.

本文针对约束鞍点问题和球覆盖最小半径的计算问题，分别利用投影法、对目标函数加上很小"扰动函数"的逼近法、罚函数法和梯度法等建立神经网络模型进行求解，并基于LaSalle不变集原理和Lyapunov直接法等工具，对模型的动态特性进行研究，从而设计出避免陷入局部极小的优化计算神经网络模型。

The available depreciation methods consist of the straight-line method, unit of production method, double declining balance method, sum of the years digits method, etc.

可选用的折旧方法包括年限平均法、工作量法、双倍余额递减法和年数总和法等。

Fixed asset depreciation method choice and to enterprise tax payment influence Starts according to January 1, 2007 to execute "Business accounting Criterion", Our country Enterprise may select depreciation method including age limit method of average, work mensuration, double remaining sum decreasing progressively law and year number method of summation and so on.

问题补充：根据2007年1月1日开始施行的《企业会计准则》，我国企业可选用的折旧方法包括年限平均法、工作量法、双倍余额递减法和年数总和法等。

Finally it offers the analysis of the fault-tolerance system as well as its test verification.

这样，运行于空间环境中的系统的设计目标就是在保证系统实时性

However, it is still difficult to find out the real reason of the increase of the keratinized gingiva during the orthodontic extrusion whether it is due to the proliferation of the gingival tissue or to its elastic nature, that is because the proliferation of the fibroblasts as well as the connective tissue changes in volume and the intra-fibers spaces were all immeasurable in most of the studies.

然而，仍然难以找到真正的原因增加的角化牙龈在正畸挤压它是否是由于扩散的牙龈组织或其弹性性质，这是因为增殖的成纤维细胞以及作为结缔组织的变化量和内部纤维空格都是不可估量的，大多数的研究。