查询词典 iterative computation

Besides, as to inpainting problem in wavelet domain, we introduce a new iterative model. In this model low frequency and high frequency will carry on processing separately, enormous speed up the image repair convergence rate, and even if as high as 90% wavelet coefficient have been lost, this model will also be able to have the very good repair effect. In the old model, to compute the curvature we have to transform the coefficients to the pixel domain, and then transform back to the coefficient domain for several times and this method can solve this problem and speed up the iterative convergence.

其次针对小波域中图像修复的问题，本文提出了一种新的小波域图像修复模型，该模型根据小波变换后系数相关性的特点，将低频和高频分别进行处理，极大的加快了图像修复的收敛速度，并且即使丢失高达90%的小波系数，该模型也会有很好的修复效果，有效的解决了传统的TV- Wavelet模型在计算图像的平均曲率WCURV时，需要反复进行小波分解与重构的计算，迭代速度慢的问题。

In the bit-interleaved coded modulation systems with iterative decoding, a good labeling is one of the key factors that dictate the system performance. In this thesis, Extrinsic Information Transfer chart, a popular tool to analyze the convergence behavior of an iterative decoding system, is employed for the labeling design.

外来资讯转换图是一个在迭代解码系统中很好的分析解码收敛行为的工具，在位元交错调变码的解码系统中，位元对应的设计对於提升系统效能是非常重要的，藉由不同的平均位元差异数来改变讯号侦测器转换曲线的斜率，因而从中选出一个最适合外部码的位元对应方式。

In Chapter 1, we summarize several iterative methods and their convergence condition, the techniques in proving the iterative methodss convergence theorem.

现在，迭代法的研究日益成为解决各种非线性问题的核心，迭代法优劣的选择直接影响到各种非线性问题的结果的良好，所以迭代法的研究有着十分重要的科学价值和实际意义。

With the cone theory and monotone iterative technique,the existence and uniqueness of the solution for the ordering symmetric contraction mixed monotone operator equations are studied,and the error estimations for the convergent iterative sequence are also given.

摘 要 利用锥理论和单调迭代技巧，研究序对称压缩的混合单调算子方程解的存在唯一性，并给出迭代序列收敛于解的误差估计。

A new definition and a new iterative scheme are introduced and the result that the iterative sequence is strongly convergent to common zero point s of maximal monotone operators Ai(i=1,2,…,m) is obtained by using the techniques of Lyapunov functional and generalized projection operator,etc.

将引进一个新定义、给出一种新迭代算法，并利用Lyapunov泛函与广义投影算子等技巧，证明迭代序列强收敛于Ai的公共零点，i=1，2，…，m。

A method of building the fuzzy model with constant-value consequence for MIMO complex processes is presented by means of the iterative learning algorithm of fuzzy model consequent parameters, and the convergence of the iterative learning algorithm is proved according to the convergent theorem of bounded positive-term series, moreover, the effectiveness of the presented fuzzy modeling method is also demonstrated by a simulation example.

深入讨论了常值后件模糊规则模型与映射的关系，并证明这种模糊模型所表示的模糊系统具有在紧集上逼近任意实值连续函数到任意精度的能力，为将这种模糊模型用于复杂过程的建模和控制作了一些理论准备。

Let C be a nonempty closed convex subset of a Hilbert space H, P:H→C be a nearest point projection, and T:C→H be a nonexpansive mapping satisfying the weakly inwardness condition, and :C→C be a fixed contractive mapping. Let the implicit iterative sequences {x},{y} and the explicit iterative sequences {x},{y}.

设C为实Hilbert空间H的非空闭凸子集，P：H→C为最近点投影映射，T：C→H为非扩张映象，且T满足弱内向条件，：C→C为压缩映象。t∈(0，l)，定义了2种隐式粘性迭代序列{x}，{y}和2种显式粘性迭代序列{x}，{y}，证明了T有不动点当且仅当序列{x}，{y}，{x}，{y}有界。

It can be found that the IPCA, as a non-iterative method, can obtain one principal component by only one calculating process. Therefore the IPCA is not only fast in calculation, but can avoid the cumulative error suffered by iterative methods.

使用模拟和实验HPLC-DAD数据对IPCA进行了验证，结果表明IPCA是一种非迭代方法，只需要完成一次计算过程即可得到一个主成分，因此IPCA不仅能够显著提高运算速度，而且可以避免迭代法中的误差积累问题。

According to the principle of the diameter of a circle corresponding to the right angle and the principle of short side for little angle in right triangle, projecting the initially-chosen point to the hyper-planes of the linear equation group in which every linear equation can be regarded as a hyper-plane and the projection points can be obtained. The initially-chosen point, and one arbitrary projection point, and the solution point are all on the surface of the relative hyper-geometry ball, therein-to, there is a projection point which is nearest to the solution point and it can be regarded as the next iterative initially-chosen point, so the solution problem of the linear equation group can be changed as an iterative problem of approaching the solution on the surface of a hyper-geometry ball.

根据直径对应的圆周角足直角以及直角三角形中短边对小角的原理进一步知道，当将初始点向线性方程组中各个方程所代表的超平面上投影得到投影点时，初始点和其任何一个投影点及方程组的解点都将位于一个相应的超球面上，其中必定存在一个投影点离问题解点的距离最短，即把该点作为下一次迭代的初始点，从而可将线性方程组求解的问题变成球面上逼近解点的迭代问题。

Although some results on convex and concave iterative roots are known, there are no results about convexity for more general iterative equations. In this chapter, convexity of both increasing solutions and decreasing solutions is investigated by the divided difference theory and fixed point theory.

本章在连续函数构成的紧凸集上构造一个连续自映射算子，利用均差理论和不动点理论证明了线性型迭代方程的凹凸解的存在唯一性及连续依赖性。

Nowadays, most of research are to build a transmutative Petri Nets through adding controlling place sets, controlling arc sets and controlling policy to the basic Petri Nets, while the Controlled Petri Nets could be used to argue many controlling theory problems conveniently and to induce many logically and physically supervisory and solve the Event Feedback Controlling Problems and State Feedback Controlling Problem in DEDS supervisory theory.

目前大多数的研究表现为在变形后的受控Petri网基础上，利用各种方法求得各种逻辑型、结构型控制器，解决DEDS监控理论中的事件反馈控制问题与状态反馈控制问题。

On one hand, there are discussions with the works council and union about extension of short time working up to the end of September.

一方面，有讨论，工程理事会和联盟关于延长工作时间短至9月底。