Modeling spatial density in low earth orbits using wavelets and random search

The events that occurred after 2007 such as Chinese anti-satellite test, explosion of Briz-M upper stage, break up of cosmos-2421 and collision of cosmos-2251 with Iridium-33 satellites have completely changed the spatial density patterns in low earth orbits. This has increased the risk of collision between active satellites and debris created by them. Aftermath, the risk assessment of possible collisions called as conjunction analysis of working satellites from day to day has become more crucial. Spatial density models are useful in understanding the long-term likelihood of a collision in a particular region of space and helpful in pre-launch orbit planning. In this paper we present an algorithmic procedure for automatically estimating exact model parameters corresponding to the peak location and number of peaks using wavelets that will speed up the parameter estimation process for the models with peaks.     

基于瞬时相位的BPSK信噪比估计算法

通过对加性高斯白噪声条件下正弦型信号瞬时相位分布情况进行研究,提出了一种新的BPSK(Binary Phase Shift Keying)信噪比估计方法.含噪信号瞬时相位的方差与信噪比之间存在近似线性的一一对应关系,以此为基础通过将含噪信号的估计方差与理论值相比较获得信噪比估计值.该方法直接对基带过采样信号进行处理,不需要事先进行符号定时与同步.使用计算机仿真测试了算法的估计性能,并与高阶累积量方法和子空间分解方法进行了比较,结果表明该算法具有更高的估计精度.      

动量轮诊断测点配置与资源占用度成本评价

以动量轮为例开展诊断测点配置与成本评价方法研究.分析系统功能模块输出与故障模式之间的关系,建立DG图故障传播定性模型.对DG图回路进行处理,生成满足可检测性和可分离性的根节点,建立偶图.采用改进的贪婪算法选取关键测点,给出诊断测点配置方案.考虑测点对系统体积、重量、功耗、处理能力等资源的要求,提出用资源占用度的概念表示成本计算模型,合理设计了资源占用度的表达形式,可以突出某项重要资源要求对评价结构的影响,也可以避免某些重要指标被弱化,并且在需要考虑其他资源要求时可扩展.通过成本模型对测点配置方案进行评价.     

基于SINS/CNS组合导航系统的多模型自适应估计算法

针对单一模型滤波器在未知或不确定的系统参数下适应性较差的问题,提出了一种新的基于多模型自适应估计(multiple model adaptive estimation,MMAE)的滤波方法。该方法利用改进的卡尔曼滤波代替传统的卡尔曼滤波,比如扩展卡尔曼滤波(extended Kalman filter,EKF)和无迹卡尔曼滤波(unscented Kalman filter,UKF)。EKF和UKF被用来作为多模型自适应估计的子滤波器,从而实现对非线性系统的状态估计。同时,还将该方法应用于基于弹道导弹模型的组合导航中实现了系统仿真。仿真结果表明,与传统的EKF和UKF算法比较,改进的滤波方法可以解决传统模型滤波器适应性差的问题,并提高系统的导航精度。     

空间目标距离像畸变分析及运动参数估计

针对卫星目标的宽带雷达回波信号,从理论上详细分析了目标运动产生的多项式相位模型,指出三次以上的相位对一维距离像的畸变可以忽略。以频谱主瓣的展宽程度为判据,推导了二次相位需补偿时的目标速度边界条件。结果表明,在典型参数条件下卫星目标的径向速度远大于该临界速度（约560m/s）,需在成像前予以补偿。最后,结合雷达目标散射中心的稀疏分布特性,提出了基于L1范数的空间目标运动参数估计算法。结果表明,该算法对速度的估计误差接近克拉美-罗下界,且远小于临界速度,可用于卫星目标成像。     

SLX 1746-311 in outburst

We report on RXTE-PCA observations of SLX 1746-311 during a 100 day outburst in April 2003. We present the 3–20 keV PCA energy spectra of the source and study evolution of spectral parameters. The burst spectrum is soft with T_in 1.3 keV at the beginning of the outburst. There is a clear evidence of a hard power law component throughout the outburst. There is also an indication of a broad iron line. The source exhibits three state transitions with considerable change in the hardness ratio at low count rates. We discuss the possible scenario under which this unusual behaviour can be explained.     

一种基于计算机视觉的飞行器姿态估计算法

改进了一种基于图像统计信息的飞行器姿态估计算法.改进后的算法可以迅速地估计出飞行器的横滚角与俯仰角,减少了运算量,提高了姿态估计的实时性和鲁棒性,精度满足飞行器的姿态控制要求.将不同的算法进行了仿真比较,提出了一种应用方案,解决了姿态估计过程中横滚角与俯仰角耦合的问题.改进后的算法适用于装备以视觉系统作为导航系统的飞行器,尤其是只能装备小型视觉系统的微小型飞行器(MAV, Micro Air Vehicles).将应用该算法的视觉导航系统与微惯性测量单元(MIMU, Miniature Inertial Measurement Unit)组合使用,可以增大飞行器姿态角估计的范围,进一步提高估计精度.      

基于遥测数据动态特征的卫星异常检测方法

基于遥测参数分析异常是保证卫星正常在轨运行的基础,通常采用阈值法判断遥测参数是否超差来判断卫星工作状态,由于其无法检测在阈值范围内变化的卫星遥测数据异常,因而会导致故障漏报.本文利用遥测参数动态变化特性,提出一种基于遥测数据变化规律检测异常的方法.利用周期图谱法求解遥测参数周期,根据遥测数据各周期之间参数值的相似性,按照遥测参数周期对数据进行采样,得到平稳差分序列,对其建立自回归移动平均混合模型,通过精确的预测结果与实测遥测数据比较来发现异常.利用该方法对实际在轨运行的某卫星2012年5月太阳能帆板转动异常故障进行验证,结果表明其能够有效避免故障漏报.     

An optimization iterative algorithm based on nonnegative constraint with application to Allan variance analysis technique

It is well known that inertial integrated navigation systems can provide accurate navigation information. In these systems, inertial sensor random error often becomes the limiting factor to get a better performance. So it is imperative to have accurate characterization of the random error. Allan variance analysis technique has a good performance in analyzing inertial sensor random error, and it is always used to characterize various types of the random error terms. This paper proposes a new method named optimization iterative algorithm based on nonnegative constraint applied to Allan variance analysis technique to estimate parameters of the random error terms. The parameter estimates by this method are nonnegative and optimal, and the estimation process does not have matrix nearly singular issues. Testing with simulation data and the experimental data of a fiber optical gyro, the parameters estimated by the presented method are compared against other excellent methods with good agreement; moreover, the objective function has the minimum value.     

离散不确定系统的优化鲁棒滤波方法及在飞行试验中的应用

提出了一种离散系统的优化鲁棒滤波方法。为了得到滤波的逼近计算式，通过优化加权矩阵得到了上界不等式逼近和等效系统矩阵，得到了鲁棒滤波的时间更新算法；通过优化加权矩阵得到了下界不等式逼近和等效观测矩阵，得到了鲁棒滤波的测量更新算珐，并且给出了鲁棒滤波算法收敛的条件。飞行试验数据处理的结果表明，提出的方法是有效的。