不同铝源对碳热还原法合成氮化铝粉末的影响

采用碳热还原法制备了氮化铝粉末，探讨了不同的起始原料、反应温度、不同添加剂对合成氮化铝粉末的影响。采用XRD、SEM、化学分析等手段对制备的氮化铝粉末进行分析。结果表明，采用活性碳和以Al(OH)3为铝源有助于抗热还原氮化反应，能提高产物的氮含量；采用CaF2具有良好的催化效果，而引入AlN晶种能有效地提高铝源的转化率。     

Expectations for Crater Size and Photometric Evolution from the Deep Impact Collision

The NASA Discovery Deep Impact mission involves a unique experiment designed to excavate pristine materials from below the surface of comet. In July 2005, the Deep Impact (DI) spacecraft, will release a 360 kg probe that will collide with comet 9P/Tempel 1. This collision will excavate pristine materials from depth and produce a crater whose size and appearance will provide fundamental insights into the nature and physical properties of the upper 20 to 40 m. Laboratory impact experiments performed at the NASA Ames Vertical Gun Range at NASA Ames Research Center were designed to assess the range of possible outcomes for a wide range of target types and impact angles. Although all experiments were performed under terrestrial gravity, key scaling relations and processes allow first-order extrapolations to Tempel 1. If gravity-scaling relations apply (weakly bonded particulate near-surface), the DI impact could create a crater 70 m to 140 m in diameter, depending on the scaling relation applied. Smaller than expected craters can be attributed either to the effect of strength limiting crater growth or to collapse of an unstable (deep) transient crater as a result of very high porosity and compressibility. Larger then expected craters could indicate unusually low density (< 0.3 g cm⁻³) or backpressures from expanding vapor. Consequently, final crater size or depth may not uniquely establish the physical nature of the upper 20 m of the comet. But the observed ejecta curtain angles and crater morphology will help resolve this ambiguity. Moreover, the intensity and decay of the impact “flash” as observed from Earth, space probes, or the accompanying DI flyby instruments should provide critical data that will further resolve ambiguities.     

广义奇异摄动降阶方法在控制器降阶中的应用

 首先推导了广义奇异摄动降阶和奇异摄动降阶的 Schur方法,解决了均衡变换可能出现病态的问题,并且算法具有较好的鲁棒性。然后,将其应用于控制器降阶,并同其它方法比较,认为原系统带宽附近的控制器降阶误差对降阶效果影响较大,而广义奇异摄动降阶法通过参数α的选择,能取得较好的降阶效果。可供工程设计参考。     

超声速射流离散频率噪声的屏蔽抑制及对流动的影响

根据对超声速射流离散频率噪声产生的反馈机理的认识,提出了一种能够有效地破坏反馈环的形成,从而抑制射流离散频率噪声的喷嘴屏蔽方法.这种方法是通过阻隔反馈声波使其不能到达喷嘴唇口、同时屏蔽物不与射流接触来实现降噪的目的.实验结果表明,对于合适的屏蔽罩参数,正常降噪效果可达5分贝以上.应用激光多普勒测速技术测量了超声速自由射流在采取屏蔽降噪措施前后轴线速度的分布,发现屏蔽后射流势核长度增加近50%,表明这种降噪方法有助于射流推力的提高,但不利于射流的掺混.     

实现DDS的波形存储表幅度值压缩方法

简介直接数字频率合成器的原理 ,提出在自行设计 DDS时采用波形存储表幅度值压缩方法以降低输出频谱的杂散。文中阐述正弦相位差法、Nicholas优化结构等幅度值压缩方法的有关原理和具体实现方法 ,并提出一种在现有方法上进一步提高压缩比的改进办法 ,最后 ,给出在 Matlab中 Simulink下的仿真验证结果     

低地轨道空间碎片环境建模与分析

根据“箱中粒子”（PIB）的模型和气体分子运动学理论，建立了估算低地轨道（LEO）上空间碎片总数的微分方程，参考国外文献提供的有关空间碎片统计数据和初始条件，求解方程，并分析了空间碎片环境的短期和长期变化趋势。     

基于高阶谱约简的变结构模糊神经网络

针对先验知识不完备和不确定的情况下海量数据造成的冗余和互斥，模糊神经网络结构变得复杂化并不能很快逼近和分类输出对象的情况，提出了一种基于高阶谱完成规则约简的变结构模糊神经网络的模型。相同结论属性的模糊规则的条件属性值可以被认为是由若干个谐波成分组成的平稳信号，并且此信号可以采用高阶谱分析来估计其谐波成分，规则的最小约简集与谐波对应。在完成了谐波估计后，神经网络结构和连接权值发生改变，神经网络的性能也得到优化。最后给出了此模型在航迹融合中应用的一个例子，得到了较好的结果。     

对接机构隔框锁的接触碰撞动力学研究及仿真

为了给空间对接飞行器提供地面试验的参考依据。验证隔框锁结构设计的正确性和对接过程的可行性。本文针对隔框锁实例，在其运动平面内以均匀B样条曲线抉合锁钩外轮廓。并建立起接触碰撞动力学微分方程和仿真模型，分析了刚体碰撞的过程，刚性接触的单边约束条件和考虑变形接触的法向计算方法，文章还提出了基于几何引擎的接触点的实时判别算法，并给出了对接完成锁钩的接触力，动能及动量的仿真曲线图。计算过程稳定，所得结果合理，具有重要的参考价值。     

A multi-disciplinary plan for easier access, management, and analysis of science data

NASA's COST LESS Team is pursuing strategies to reduce the cost and complexity of planning and executing space missions. The team's technical goal is to reverse the trend of constructing unique solutions for similar problems. To this end, the team is exploring ways to represent mission functionality in terms of building blocks and is discovering approaches that could accommodate the same building blocks for seemingly disparate activities, such as organizing processed telemetry data, controlling onboard experiments, searching science archives, reducing and presenting information to science users, and supporting educational outreach. Reusable object technology (UOT), a research undertaking by the authors, is showing promise in recognizing similarities in functions which were previously viewed as unique because they appeared in different programs or mission phases. Since UOT is aimed at being implementation independent (i.e. the function performed could be accomplished manually, by an automated process, by a specialized instrument, etc.), no premature judgment for automation or autonomy need be made. In this paper, the authors attempt to strike a balance between theory and reality as they describe UOT, including its beginnings, its underpinning, its utility, and its potential for achieving substantive reductions in cost and complexity for the Agency's space programs. The authors discuss their collaboration with the Center for EUV Astrophysics, University of California, Berkeley to reduce the cost and complexity of science investigations. Their multi-disciplinary plan incorporates both UOT and a complementary technology introduced in this paper, called interactive archives.