A compositional method to model dependent failure behavior based on PoF models

In this paper, a new method is developed to model dependent failure behavior among failure mechanisms. Unlike the existing methods, the developed method models the root cause of the dependency explicitly, so that a deterministic model, rather than a probabilistic one, can be established. Three steps comprise the developed method. First, physics-of-failure(PoF) models are utilized to model each failure mechanism. Then, interactions among failure mechanisms are modeled as a combination of three basic relations, competition, superposition and coupling. This is the reason why the method is referred to as ‘‘compositional method". Finally, the PoF models and the interaction model are combined to develop a deterministic model of the dependent failure behavior. As a demonstration, the method is applied on an actual spool and the developed failure behavior model is validated by a wear test. The result demonstrates that the compositional method is an effective way to model dependent failure behavior.     

Stark broadening of B IV lines for astrophysical and laboratory plasma research

Stark broadening parameters for 36 multiplets of B IV have been calculated using the semi-classical perturbation formalism. Obtained results have been used to investigate the regularities within spectral series and temperature dependence.     

Stark widths of Ar III spectral lines in the atmospheres of subdwarf B stars

Using semiclassical perturbation approach in impact approximation, we have calculated Stark widths for 32 spectral lines of doubly charged argon (Ar III). Oscillator strengths are calculated using Hartree–Fock method with relativistic correction (HFR) and an atomic model including 17 configurations. Energy levels are taken from NIST database. For perturbing levels for which the corresponding energy does not exist in NIST database, the calculated energies are used. Our widths are compared with the experimental results. The results presented here are of interest for modelling and investigation of stellar atmospheres since argon in different ionization stages is observed in many astrophysical objects. Finally, the importance of Stark broadening mechanism is studied in the atmospheric conditions of sdB stars. Electron impact Stark widths are compared to thermal Doppler widths as a function of temperature and optical depth of atmospheric layers.     

基于随机森林的航空器到达时刻预测

航空器到达时刻（ ETA）预测是进场排序与调度的基础，因此进场飞行时间的快速与准确预测显得尤为重要。通过分析航班信息、天气信息以及空中交通信息，基于随机森林算法构建了航空器到达时刻预测模型。选取上海浦东机场进场航班进行仿真验证，仿真结果表明，预测模型可以实现航空器到达时刻的快速与准确预测。     

最小二乘法在计量经济模型中的应用

本文从古典计量经济模型的建立谈起，又从异方差模型，自相关模型和联立方程模型中，论述了由最小二乘法派生出的加权最小二乘法、广义最小二乘法和间接最小二乘法等。     

CFAR DETECTORS FOR HIGH RESOLUTION RADAR IN NON-GAUSSIAN CLUTTER

讨论了球不变随机过程（ＳＩＲＰ）非高斯杂波背景下高分辨雷达距离扩展目标检测问题。文中提出了ＣＡ－ＣＦＡＲ，ＯＳ－ＣＦＡＲ两种距离扩展目标检测器，并证明了它们的形状因子和尺度因子未知的情形下具有恒虚警性，即双参数恒虚警性，分析了这两种检测器的检测性能。这两种检测器亦可用作ＳＩＲＰ杂波背景下低分辨雷达检测器。     

基于新体制的单基线定位技术研究

结合工程应用,研究单基线利用模糊相位差对地面固定目标的高精度快速定位方案与关键技术,分析自适应随机惯性权值粒子群算法的性能。仿真结果表明,通过优化和改进,基于新体制的该算法能获得较好的定位性能。     

高阶矩谱分析与空间等离子体非线性相互作用识别

利用高阶矩统计分析方法对ＡＵＲＥＯＬ－３卫星观测数据进行分析处理，所涉及的信号分别为ＥＬＦ频率范围内的电场扰动Ｅｈ（ｔ）和等离子体密度波动δＮｅ／ｎｅ（ｔ），处理结果表明：利用高阶矩谱分析方法可以很好地识别空间等离了波－粒相互作用过程中的非线性成分     

非金属电热膜材料及工艺分析

本文论述了制作电热膜的工艺过程.对电热膜的材料、烧结工艺及使用寿命进行了试验,获得了有价值的结论.并论述了提高电热膜产品寿命应注意的问题.     

基于LQG/LTR方法的鲁棒飞行控制系统设计

针对飞行控制系统设计中存在的建模误差以及实际飞行中外界干扰的影响,采用LQG/LTR鲁棒控制方法完成了某型攻击机横航向控制设计,解决了在飞机模型不确定性、随机干扰下控制系统可能出现的不稳定和控制精度不够的问题。系统仿真结果表明,控制系统实现了横航向指令的精确跟踪,不但具有良好的鲁棒性,而且调节性能良好,满足飞机横航向控制的要求。