在算法上

The basic approach of protecting people from being hurt or killed in an accident is to improve crashworthiness of vehicles. This paper starts with discussing theories and methods for vehicle passive safety design, which included experiential methods, analytic methods, multi-body dynamics methods, crash test methods and the finite element method. Emphasis will be paid to the basic FEM theories and algorithms of impact problems. Topics discussed include the governing equation, element discretization, hourglass control, time integration, material model, shell element algorithms and contact-impact algorithms. For more reliable simulation results, this paper then conducts study on how to accurately get material model parameters by combining the FEM method with experimental method, and develops relevant material test machine and parameter calculating software. Based on that, basic typical impact simulation and test validation study are conducted. On the basis of the above work, the paper studies the thin-shell structures'crashworthiness, which was affected by the following factors: spot-weld features, shell thickness, cross-section and pre-deformation. As to vehicle parts'crashworthiness, the paper conducts simulation study of designing energy-absorbing steering system and adaptive airbag system. A practical energy-absorbing steering wheel is designed as an example. In further depth research, the paper suggests several inverse quantificational methods for vehicle crashworthiness design based on the ideal crash characteristics. The above theories and methods are applied with good results through several practical vehicles' crashworthiness design and improvement.

文中对汽车碰撞安全性的设计理论和方法进行了归类和总结，其中包括经验法、解析法、多刚体动力学法、试验法以及有限元方法等；重点介绍了碰撞有限元法的基本理论和有关算法，涉及到有限元求解控制方程、单元离散、沙漏模态控制、时间积分、材料模型和应力修正、薄壳单元算法以及接触碰撞界面算法等，探讨了有限元计算中的材料模型参数获取技术，开发了相应的材料试验装置及材料模型参数反求软件，在此基础上进行了基本的碰撞仿真算例和试验对比研究；文中通过应用有限元方法研究了薄壁构件的碰撞吸能特性，指出了设计薄壁吸能构件时需要考虑的几个主要影响因素，即焊点、壁厚、横截面和预变形等；在关于车辆部件的碰撞安全性能设计中进行了吸能转向机构和自适应安全气囊的仿真研究，提出了吸能转向机构和自适应安全气囊的仿真研究方法；文中最后探讨了整车碰撞性能的设计与改进方法，提出了以理想碰撞特性为目标，采用分段加速度或者根据碰撞时间进行设计的反推设计法、部件吸能仿真优化设计法等量化设计方法，并通过综合应用文中所讨论的相关技术进行了整车碰撞性能的设计与改进实例研究，所选定的车型是大众化的普通轿车和碰撞安全性基础较差的微型面包车，研究结果表明，本文所提出的设计与改进方法是适用而有效的，具有重要的工程实用意义和价值。

It is the first important step in video analysis and will directly affects the effectiveness of indexing Shot boundary detection is one of our major research interests and we will tackle the following existing problems the ambiguity between gradual change and camera motion, the discontinuity during gradual change, false detection caused by illumination variation and flashlight, automatic threshold selection Firstly, we discuss shot boundary detection in non-compressed domain In chapter 2, we compare some of the commonly used detection methods which are based on frame difference and point out that single feature will not generate good results As a conclusion, we use fuzzy logic to combine multiple features Presently, most frame difference based shot boundary detection algorithms rely on threshold and hence the selection of such thresholds will greatly affect the performance of boundary detection We propose a membership function to define frame difference and calculate the membership with self adaptation according to the statistic distribution of frame differences to satisfy different type of video clips Experiments show that the proposed fuzzy shot boundary detection algorithm can be used with different video types and has a high detection precision and recall In chapter 3, we discuss model-based shot boundary detection algorithms regarding chromatic and spatial editing effects such as fade-in, fade-out, dissolve and wipe Various parameters are proposed to better describe the characteristics of each editing type.

镜头边界的检测是把视频自动地分割为一个个镜头，作为基本的索引单元，因此它是视频分析重要的第一步，直接影响到视频检索的成败。镜头边界的检测是本文研究的重点之一。目前镜头边界检测算法主要存在以下问题：渐变与镜头运动难以区别；渐变过程中的不连续与停顿、光照条件的变化及闪光灯等特殊情况会引起误检测；自动选择阈值比较困难等。本文首先针对非压缩域视频进行了镜头边界检测的研究。在第二章中我们采用了比较流行的基于帧间差的方法。在比较各种帧间差计算方法的基础上，指出使用单一的特征难以取得很好的检测效果，提出用模糊逻辑综合使用各种特征。目前大多数基于帧间差的镜头边界检测算法都采用阈值法进行镜头转换的判别。阈值选择的误差对检测性能有较大的影响，本文提出用隶属度函数定义帧间差较大、中等较大和较小等概念，并根据帧间差的统计分布自适应地确定隶属度，以适应不同类型的视频片断。实验结果表明这种基于糊逻辑的镜头边界检测算法可以适应不同的视频，并具有较高的检测精度和检出率。在第三章中采用基于模型的方法进行镜头渐变的检测，研究了淡入／淡出、慢转换和扫换的模型。

In chapter three, the elliptic curve discrete logarithm problems on the elliptic curve finite group are discussed.

第3章在介绍有限域上的离散椭圆曲线的基础上，深入讨论了椭圆曲线有限群上的椭圆曲线离散对数问题，研究了目前已知的椭圆曲线离散对数问题的几类求解算法，分析了这些算法的特点和应用范围，并总结归纳出了一系列的安全椭圆曲线选取准则。

It produces corresponding nonredundent rule by direct generalization of lattice's nodes; This paper also improves the incremental concept formation and Hasse diagram update and apply it to incremental rule generation. We give examples to illustrate the idea of the algorithm and corresponding experimental results. 6 Rough set theory and concept lattice are similar in some aspects, the relationship between them is interesting.

在概念格结构上提出了一种在一定条件下更有效的，在已建造好的概念格上提取规则的算法，这种方法主要依据格结点的直接泛化来产生相应无冗余规则；改进了一种的渐进式更新概念格与相应Hasse图的算法，并将之应用于渐进式规则生成。

Aimed at the characteristic of multiple types of faults possibly happened in nuclear power plant, large scale of training sample, and requirement of quick and accurate diagnosing, after the analysis of large sample number generated by large fault type exceeds limitation of 64K paragraph in DOS and large number of local minmum in error surface, measurements of same format of data file, Win95/NT operating system platform, resetting weight learning rate, dynamic training set in quick learning algorithm and improvement on quick learning algorithm using homotopy method which can avoid local minmum points in error surface have been adopted to ensure quickly and effective process of the course of neural network's training and testing.

针对核动力装置可能发生的故障种类多，训练样本规模大，故障诊断需要快速准确的特点等，分析了样本多和故障种类多产生的训练样本量超出DOS段大小和网络误差曲面上局部极小值多的情况，采取了使用相同格式数据文件、Win95/NT操作系统平台，对快速学习算法采用重置神经网络权值学习率和动态训练集、并采用能够有效克服网络误差曲面上局部极小点的同伦方法对学习算法进行改进等措施保证神经网络训练测试过程的快速、有效地进行；针对装置发生的故障须快速、准确诊断的需要，分析了故障的产生对装置参数变化的影响及操纵员对故障诊断的基础，在核动力装置发生故障时参数曲线的变化量与正常运行时参数曲线的变化量存在明显差异的基础上，提出采用参数曲线的变化量作为神经网络的输入，并围绕参数变化量的方法采用二次曲线拟合滤波求变化量和阈值技术来保证神经网络得到精确的装置参数变化量，从而得到准确的诊断结果。

To conquer limitations of existing techniques for developing loop invariants, professor Xue Jinyun, who is imbursed under two National Nature Science Foundation of China, systematically research the theory and method of loop invariants and advanced the new definition and developing strategies for loop invariants. Based on above, he presented efficiency method of developing algorithmic program, Partition-and-Recur Approach, and its developing environment which bring into important play in formal method.

为了克服已有循环不变式开发技术存在的局限性，薛锦云教授在两个国家自然科学基金课题&若干新的算法程序设计和证明方法研究&和&实用的软件形式化方法及其开发工具的研究&的资助下，对循环不变式的理论和方法进行了系统的研究，提出了关于循环不变式的新定义和新的开发策略，并在此基础上形成了一种实用的算法程序形式化开发方法及其开发环境，在复杂算法程序及软件形式化开发中发挥了重要作用。

The paper puts forward immunity arithmetic used in distribution network on the basis of analyzing common uncertainty arithmetic and immunity arithmetic theory. It gets an optimizing project for distribution network, in which the equilibrium of load is regarded as target function. It has been exampled that this project is of higher constriction speed and stronger global search ability.

3在分析常见的非确定性算法和进化免疫算法原理的基础上，提出了在配电网络重构中采用免疫算法，并求解以负荷均衡为目标函数的配电网络重构的最优方案，实例表明该方法具有较快的收敛速度和较强的全局搜索能力；同时本文还为实时配电网络重构定义了启动条件，避免在配电网络正常运行时频繁启动网络重构。

According to the function of distributed emulator, optical information processing and image multi-object tracking system, the following three aspects were emphasized in this dissertation. Firstly, a vast virtual terrain, ground radar models and canyons were created for the distributed emulator with modeling software Creator. The radars were driven to scan during emulating and the real time terrain code was sent to Radar module. Secondly, an optics channel simulating the optics camera of a virtual aircraft was implemented with simulating software Vega. The 2-D information got from the channel was translated and sent to multi-sensor information fusion system as an optics information source. Thirdly, the dissertation studied some motion estimation methods and analyzed the factors which influence the speed and the precision of the block matching algorithms. Then an improved block matching method was proposed for image multi-object tracking system and applied in the system combined with frame difference method.

论文根据多传感信息融合实验系统中的分布式飞行仿真系统、光学信号处理器和图像多目标跟踪系统三个子系统的功能要求，具体完成了以下三方面的工作：首先利用建模软件Creator创建分布式飞行仿真系统中的大面积地形、场景中的地面雷达模型和峡谷地形，并在仿真过程中实时驱动雷达模型的扫描，同时为后续雷达仿真模块提供实时地形编码；其次利用仿真软件Vega，在分布式仿真系统中为光学信号处理器设计实现了光学通道，模拟飞行器的光学摄像机的功能，并且对由此通道获得的目标二维光学信息进行转化处理，作为光学信息源发送给后续的多传感信息融合器；最后针对图像多目标跟踪系统，在研究图像分割运动估计算法的基础上，通过分析影响块匹配算法速度和精度的因素，改进了块匹配运动估计算法，结合差分法可应用到目标跟踪系统中。

Based on the numerical results of tensile test of single-crystalline nano-scale copper bar, the average and local tensile stresses are compared, the advantages and disadvantages of these two methods are discussed in detail. Different external loads are considered during the simulations.

比较两种算法在计算垂直于拉伸方向截面上的平均拉伸应力和在该截面上不同位置处的局部拉伸应力，分析在不同外载荷下两种方法的计算结果，最后总结出两种算法各自的优点和缺点。

Use the mew accelerate function of OpenGL to accelerate the units display. Have adopted the three dimension graphics and OpenGL's technology, it makes use of Stencil test to carry on cutting operation, which reduces complexity of programming and has raised the speed and efficiency of displaying.

在后处理过程中，引进了DSI插值方法，改进了以往等值线插值，提高了算法速度和效率，满足了实时性要求，在扫描线算法的基础上，引进了OpenGL插值方法，改进了云图实现方法，提高了图形显示精度，在动画格式上，我们可以根据自己需要实现不同精度的不同压缩格式动画文件，提高了后处理显示应用场合，改变了以往动画格式单一的形式。

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。