在算法上

Based on the existed theory of mareoids and fuzzy matroids, this thesis studies the closed regular fuzzy matroid and its fundamental sequence, the fuzzy base and its algorithm of closed fuzzy matroids, the fuzzy circuit and its algorithm of closed fuzzy matroids and so on. The main contributions of this thesis are as follows: 1 The necessary and sufficient condition of closed regular fuzzy matroid and a property of its fuzzy dual matroid are found by studying some properties of closed regular fuzzy matroid. 2 By studying some properties of fuzzt bases of closed fuzzy matroid, the necessary and sufficient condition of judging fuzzy bases of closed fuzzy matroids and some corollaries are found. In the end, an algorithm of obtaining a fuzzy base is given. 3 By studying some properties of fuzzt circuits of closed fuzzy matroid, some necessary and sufficient conditions of using its fundamental sequence to express fuzzy circuits are found. An algorithm of obtaining a fuzzy circuit is given. 4 By studying the fundamental sequence of closed regular fuzzy matroid, some necessary and sufficient conditions of fundamental sequence of closed regular fuzzy matroid are found.

本文在现有拟阵和模糊拟阵理论的基础上，研究了闭正规模糊拟阵及其基本序列，闭模糊拟阵的模糊基及算法、模糊圈及算法等内容，现分述如下： 1研究了闭正规模糊拟阵的一些性质，得到了闭正规模糊拟阵的充要条件及其模糊对偶拟阵的一个性质； 2研究了闭模糊拟阵模糊基的性质，找到了闭模糊拟阵模糊基的充要条件和几个推论，最后还给出了求模糊基的算法； 3研究了闭模糊拟阵模糊圈的性质，找到了用基本序列来表达模糊圈的几个充要条件，并给出了求模糊圈的算法； 4研究了闭正规模糊拟阵的基本序列，找到了闭正规模糊拟阵的基本序列的几个充要条件。

To find the minimum of energy function E is a part of work in the test generation based on neural network.

给出了神经网络测试生成的遗传进化方法的并行实现算法；提出了异步拟牛顿法、异步割线法和异步区间算法等三种异步并行算法，并对每一算法的收敛性及收敛效率做了分析，为神经网络测试生成方法在多机并行系统上的仿真实现提供了新的途径。

By the analyses on the existing detector generating algorithms, we achieved the improvement of the linear time detector generating algorithm, which was widely used at present. The efficiency of the linear algorithm was improved by eliminating the reluctant detectors, and at the same time the improved algorithm was ensured to cover the nonself space as much as possible.

在分析已有的检测器生成算法的基础上，改进了目前使用最多的线性检测器生成算法，通过消除冗余的检测器提高了线性检测器生成算法的效率，同时还保证了改进后的算法能够尽可能多地覆盖"非我"空间。

By iteratively combining these two procedures we achieve a way of training the configuration of RBFNN. In addition, the algorithm is very robust with respect to the noise level. Simulation experiments show that the proposed algorithms are sound. In chapter 3, which is the extension of chapter 2, the improvements of the architectures of RBFNNs and the solutions of their relating problems are discussed, Firstly, an new distance metric is advanced and a forward orthogonal least square selection procedure is applied to learning the parameters of classification function and selecting the important input nodes.

首先分析了目前存在的同时确定RBFNN结构和参数的方法的缺陷，在此基础上提出了用基于拉马克的进化学说的进化编程算法来改进算法，以克服某些缺陷；然后针对受到严重噪声污染的系统，如何提高RBFNN的泛化能力的问题，利用基于AIC的适应度函数的改进遗传算法学习结构和参数；最后介绍基于MDL原理的方法，将优化网络的结构和参数分为两个阶段：训练和进化，先自适应地改变RBFNN基函数的中心和宽度，同时训练输出线性权值，再用基于MDL原理的适应度函数的标准GA来优化隐层节点，通过交替使用这两过程达到训练RBFNN的结构和参数的目的，该算法具有较强的鲁棒性。

The problem of huge numbers of datas are disposed in the process of disposed three-dimensional laser scanning datas, so on the base of carefully analszing the serial algorithm of plane segmentation and multiple datasets registration in the process of disposing three-dimensional laser scan datas, according to the method of parallel programming basing on MPI, the paper used the right means of datas distributing to design and realize the parrel algorithm of plane segmentation and multiple datasets registration.

本文针对三维激光扫描数据处理过程中所处理的数据量大这一问题，在认真分析了三维激光扫描数据处理过程中平面分割和多视点配准串行算法的基础上，根据基于MPI并行程序设计方法，采用合适的数据分配方案，设计了平面分割和多视点配准的并行算法，并给出了效果图，以及并行算法加速比和效率的实际数据，最后对并行算法效果进行了分析。

Dynamic programming algorithm was introduced to solve the problem. Based on discussion of calculation complexity and effectiveness, the exact DP Algorithm was successfully used in operative schedule of the collecting bucket in multi-car-single-layer mode, which has few service nodes. For problems with large number of service nodes, restricted DP algorithm was constructed to avoid rapid expand in calculation and memory demands during programming process. The restricted DP algorithm, successfully used in single-car-multilayer and compound service modes, gets balance between near optimal programming results and reasonable programming time, so as to meet the real time schedule demands.

文中首先在TSP-STW问题的动态规划模型基础上，论述了计算机精确动态规划解法的求解策略和方法以及算法的有效性和复杂性，并将其成功应用于服务节点数相对较少的多小车单层配药服务模式中集中料斗的集药动作规划；在进行服务节点数较多的单小车多层服务模式中取药小车的运动规划时，为限制计算量和存储空间需求的迅速膨胀，引入以合理限制动态规划算法在每一阶段保留的最大状态元素数为核心的有限动态规划法，以便在求解结果的优化程度和规划时间上取得均衡，达到实时调度的要求。

Based on the analysis and research of the domestic and overseas EIT algorithms, a practical and fast EIT algorithm with better resolution---Fast Newton's One-Step Error Reconstructor dynamic algorithm is studied and realized in this dissertation. Based on a great deal of experiment researches, the EIT algorithm and performance of hardware-measuring device needed by imaging system are analyzed.

本论文在分析、研究国内外有关电阻抗成像重建算法及其研究状况的基础上，研究实现了一种实用、快速、具有一定分辨率的电阻抗成像算法——快速牛顿一步误差重构动态算法，对成像算法进行了大量的实验研究和数据分析，对成像系统所采用的硬件测量装置的性能进行了研究。

In this paper,we present an improved digital image watermarking method of singular value based on matrix multiplication of strassen and comparability measurement between original image and watermarking image;the new method is compared to the methods based on SVD and Block-SVD.

在基于奇异值分解算法的基础上，提出了一种基于Strassen矩阵乘法的奇异值分解水印算法；提供了原图像和水印图像的相似性度量方法；给出了本算法与SVD及Block-SVD算法的时间对比分析。

Time domain algorithm,range-doppler algorithm and wave number domain algorithm are being discussed and compared on this foundation.

在此基础上，分别讨论时域算法、距离——多普勒算法、波数域算法并对其进行了比较，并对波数域算法的并行性做了初步研究。

This paper will analyze the reason that the performance of imbalanced data is weak,and improve this classical Boosting algorithm by restraining from overfitting and controlling the F-Measure of minority class,and propose an improved algorithm named RIFBoost,and then compare this new algorithm with some traditional classified algorithms on WEKA system.

从分析非平衡数据分类性能差的原因入手，通过抑制过度拟合与对少数类的F度量的控制对经典推进算法进行改进，提出了一种改进算法RIFBoost，然后将算法在WEKA系统上与几个传统的分类算法进行了比较。

But we don't care about Battlegrounds.

但我们并不在乎沙场中的显露。

Ah! don't mention it, the butcher's shop is a horror.

啊！不用提了。提到肉，真是糟透了。