大型不规则椭球回转面的曲面重构与仿真加工

提出了大型不规则椭球回转面的一种新的曲面重构与仿真加工方法。基于GeomagicStudio软件,处理大型不规则椭球回转面的采集数据点云,重构了实测曲面的等距面模型;采用Pro/E对实测曲面的等距面模型进行等距操作,生成实测曲面模型;利用GeomagicQualify软件对实测曲面的等距面模型相对于原始数据点云,以及实测曲面模型相对于理论曲面模型进行误差分析;最后利用Pro/NC模块对实测曲面模型进行数控编程和仿真加工。研究结果表明,该方法提高了大型不规则椭球回转面的曲面重构精度,生成的标准数控加工程序具有可移植性,解决了原加工方法对专用设备的依赖。     

反舰导弹命中概率模型研究

根据现代海战的特点,结合敌我武器装备的实际,对反舰导弹的命中概率计算模型进行了构造和分析,同时针对具体实例,进行了仿真计算及结果分析。     

高可扩展格子Boltzmann方法

格子Boltzmann方法(Lattice Boltzmann Method,LBM)是计算流体力学中的一种常用方法.由于LBM中的格点仅与相邻的格点间存在数据传递,因此具有良好的并行性.LBM并行算法中的数值通信部分通常采用的是格点上的微观量——分布函数.每次传递的分布函数具有多个不同的速度方向,为了进一步减少LBM并行算法的通信开销,从格子 Boltzmann方法的物理特性以及相应的串行程序为切入点,深层次地挖掘可并行的因子,设计了专门用于通信面的类,使用数量较少的宏观量进行通信,降低了通信所占的比重,缩短了通信时间,提高了加速比和效率.实验表明,在4 096个计算核上依然有良好的加速比和效率.     

Predicting the vulnerability of spacecraft components: Modelling debris impact effects through vulnerable-zones

The space environment around the Earth is populated by more than 130 million objects of 1 mm in size and larger, and future predictions shows that this amount is destined to increase, even if mitigation measures are implemented at a far better rate than today. These objects can hit and damage a spacecraft or its components. It is thus necessary to assess the risk level for a satellite during its mission lifetime. Few software packages perform this analysis, and most of them employ time-consuming ray-tracing methodology, where particles are randomly sampled from relevant distributions. In addition, they tend not to consider the risk associated with the secondary debris clouds. The paper presents the development of a vulnerability assessment model, which relies on a fully statistical procedure: the debris fluxes are directly used combining them with the concept of vulnerable zone, avoiding the random sampling the debris fluxes. A novel methodology is presented to predict damage on internal components. It models the interaction between the components and the secondary debris cloud through basic geometrical operations, considering mutual shielding and shadowing between internal components. The methodologies are tested against state-of-the-art software for relevant test cases, comparing results on external structures and internal components.     

基于统计分析模型的飞行器落点预报方法

目前飞行器落点预报主要采用人机交互方式进行处理,而不同人员计算出的结果存在较大差异.针对该问题,提出了一种基于四阶龙格库塔计算方法的落点预报分布式计算统计分析模型.该模型根据飞行器的运行轨迹特性和在无动力、无控制、常质量段的预报落点具有集中分布的特点,采用零假设和备择假设统计方法对预报落点过滤选择,并对有效预报落点求其数学期望值.采用实际测量数据对所提出的模型进行了可行性和有效性验证,试验结果表明,该模型不仅具有较强的实时性、稳定性,而且当飞行器运行距离在数千千米时,该方法计算出的平面位置误差仅为几十米.     

热发射测速技术

本文叙述了一种通过探测高温两相流中热粒子的热辐射来测量粒子速度的热发射测速技术,该技术曾多次用于测量电弧加热器射流和火箭弹尾喷火焰中的粒子速度,取得了比较满意的结果。     

Data Grids and High Energy Physics: A Melbourne Perspective

The University of Melbourne, Experimental Particle Physics group recognises that the future of computing is an important issue for the scientific community. It is in the nature of research for the questions posed to become more complex, requiring larger computing resources for each generation of experiment. As institutes and universities around the world increasingly pool their resources and work together to solve these questions, the need arises for more sophisticated computing techniques. One such technique, grid computing, is under investigation by many institutes across many disciplines and is the focus of much development in the computing community. ‘The Grid’, as it is commonly named, is heralded as the future of computing for research, education, and industry alike. This paper will introduce the basic concepts of grid technologies including the Globus toolkit and data grids as of July 2002. It will highlight the challenges faced in developing appropriate resource brokers and schedulers, and will look at the future of grids within high energy physics. This revised version was published online in August 2006 with corrections to the Cover Date.     

基于高分四号数据的多时相多通道云检测算法

针对高分四号（GF-4）卫星影像波段较少导致传统云检测算法难以区分云与冰雪像元的问题,提出一种多时相多通道云检测算法。该算法首先对GF-4卫星影像进行辐射定标和配准,然后利用云与典型地表的光谱差异得到潜在云像元,之后利用序列GF-4卫星影像之间的差异识别出移动的云像元,最后利用中红外波段反演地表亮度温度来去除冰雪像元。该算法在海南、辽宁和安徽3个研究区域进行验证,并将检测结果与传统单时相云检测算法、支持向量机（SVM）云检测算法和实时差分（RTD）云检测算法的检测结果进行对比。结果表明,该算法优于其他3种云检测算法,准确识别率均达到90%以上,误检率均低于5%,有利于GF-4卫星影像的进一步利用。     

基于CAD二次开发的飞行器隐身性能计算方法

在飞行器同雷达对抗时的不同威胁情况下,分析了在雷达仰视照射下飞行器的隐身性能。通过绘制雷达仰角图,得到了目标飞行器在某典型雷达照射下的暴露与隐身区域,从而量化了目标飞行器的隐身性能。通过CATIA二次开发技术实现了CAD模型的便捷修改,为优化外形隐身特性提供支持。最后通过一个具体算例验证种方法的效果,并说明方法具有实用意义。     

一种空间非合作目标的稀疏点云配准算法

点云数据配准是三维重构的关键技术之一,为了提高空间非合作目标的稀疏扫描点云数据配准的速度和精度,提出一种改进的基于四点算法的全局配准算法进行初始配准,再使用迭代最近点算法精确配准.针对直接扫描所得到点云数据量大的问题,本文提出一种基于KD Tree点云均匀采样简化算法,并且对传统基于四点算法中的阈值参数进行了统一,确定了各误差阈值参数和点云密度之间的关系.仿真结果表明,该方法能够快速、有效地实现卫星稀疏点云的配准,改进的四点算法配准耗时仅为几何哈希算法的42.49%.