时序频谱的自适应估计

本文对AR模型的频谱作出自适应估计。文中首先对AR(p)的模型参数作递推自适应估计，然后用最大熵谱估计作自适应谱估计。在样本容量(短数据)较小的场合，给出了估计的精度分析。文中所提供的方法．对于时序变化剧烈的情况，具有较好的跟踪效果以及较高的分辨率和精度。     

尺寸放大对材料参数分布影响的估计方法及其指标

本文讨论了造成实际构件静强度材料参数比试件实验结果更具分散性的原因,主要是材料化学成分分散性增大、机械加工公差带来的分散性增大及热处理不均匀性带来的分散性增大。给出了估计尺寸放大对材料参数影响的一般方法和用试件实验结果估计实际构件材料参数的线性均方估计方法。提出了评价尺寸放大影响的指标:相关系数γ。以LC4的静强度极限σ_b为例进行了分析,说明方法是合理的。     

Analytic estimation and numerical modeling of actively cooled thermal protection systems with nickel alloys

Actively cooled thermal protection system has great influence on the engine of a hypersonic vehicle, and it is significant to obtain the thermal and stress distribution in the system. So an analytic estimation and numerical modeling are performed in this paper to investigate the behavior of an actively cooled thermal protection system. The analytic estimation is based on the electric analogy method and finite element analysis（FEA） is applied to the numerical simulation. Temperature and stress distributions are obtained for the actively cooled channel walls with three kinds of nickel alloys with or with no thermal barrier coating（TBC）. The temperature of the channel wall with coating has no obvious difference from the one with no coating, but the stress with coating on the channel wall is much smaller than that with no coating. Inconel X-750 has the best characteristics among the three Ni-based materials due to its higher thermal conductivity, lower elasticity module and greater allowable stress. Analytic estimation and numerical modeling results are compared with each other and a reasonable agreement is obtained.     

无人机测控通信系统中智能天线技术

针对无人机测控与信息传输系统中地面设备笨重、跟踪速度慢等问题,提出地面天线采用智能天线的方法,利用MUSIC(MUltiple SIgnal Classification,多重信号分类)算法实现对机载终端的DOA(Direction Of Ar-rival,波达方向)估计,并添加去相关操作对抗多径,利用基于LCMV(Linearly Constrained Minimum Variance,线性约束最小方差)准则的波束成形算法完成波束赋形,相较于传统的机械跟踪方式,具有设备质量小、成本低、机动性灵活性强和跟踪速度快等优势.仿真和实际测试结果表明,采用DOA估计和波束成形算法的智能天线能够在保证高增益的同时,实时准确地估计出有用信号和多径信号的来波信号方向,并正确完成波束指向,具有很高的工程应用价值.     

平尾积冰对飞机纵向气动参数的影响

建立飞机纵向动力学模型,基于最大似然参数估计原理,设计用于辨识飞机纵向气动参数的辨识系统,并对辨识系统的正确性和精确度进行了验证.以DHC-6飞机飞行试验数据为依据,对未积冰飞机和两种平尾积冰冰型的飞机进行纵向气动参数辨识,通过对比3种情况下飞机纵向气动参数的辨识结果,定量分析了平尾积冰对飞机纵向气动参数的影响.结果表明:平尾积冰将导致飞机纵向气动特性恶化,俯仰阻尼可减小15％,升降舵效率可降低20％,对飞行稳定性、操纵性以及飞行安全构成一定的威胁.     

Accurate single-observer passive coherent location estimation based on TDOA and DOA

This paper investigates the problem of target position estimation with a single-observer passive coherent location(PCL) system. An approach that combines angle with time difference of arrival(ATDOA) is used to estimate the location of a target. Compared with the TDOA-only method which needs two steps, the proposed method estimates the target position more directly. The constrained total least squares(CTLS) technique is applied in this approach. It achieves the Cramer–Rao lower bound(CRLB) when the parameter measurements are subject to small Gaussian-distributed errors. Performance analysis and the CRLB of this approach are also studied. Theory verifies that the ATDOA method gets a lower CRLB than the TDOA-only method with the same TDOA measuring error. It can also be seen that the position of the target affects estimating precision.At the same time, the locations of transmitters affect the precision and its gradient direction.Compared with the TDOA, the ATDOA method can obtain more precise target position estimation.Furthermore, the proposed method accomplishes target position estimation with a single transmitter,while the TDOA-only method needs at least four transmitters to get the target position. Furthermore,the transmitters' position errors also affect precision of estimation regularly.     

利用跟踪-微分器构造机动目标估计模型

在分析目标运动特点的基础上,提出了一种新的基于跟踪-微分器的机动目标估计模型。该估计模型与卡尔曼滤波算法相结合,能够在动态过程完全未知的情况下估计出目标运动参数,且估计模型简单,物理意义明确,能够适应于目标的各种机动行为模式。仿真结果表明,具有跟踪-微分器结构的估计模型不但具有很强的适应性,而且能够获得满意的估计精度。     

基于循环互相关的窄带回波多普勒频移估计

针对雷达窄带回波信号的循环互相关估计多普勒频移问题进行了研究。利用信号循环平稳特性,分析了存在多普勒频移条件下雷达窄带回波信号的循环统计特性,在此基础上提出了基于循环互相关变换的多普勒频移估计算法,该方法能有效地估计多普勒频移参数。最后利用计算机进行了仿真实验,验证了该算法的有效性。     

基于单目视觉的非合作目标相对位姿测量

为实现对空间姿态翻滚航天器的在轨服务与维护以及对空间碎片的清理,需对其进行精确的相对位姿测量。针对相对位姿测量问题,提出了基于单目视觉与卡尔曼滤波的相对位姿测量方法。通过对特征点匹配算法进行调查,采用了具有尺度不变性与旋转不变性的尺度不变特征变换算法(SIFT)和加速稳健特征算法(SURF)的特征点提取方法,并对二者进行了对比,得到了二者分别适用的工况条件。通过对Kalman滤波算法进行研究,引入了相机偏置矩阵,设计了Kalman滤波器,解决了单目相机的距离模糊问题,估计得到了非合作目标的相对位姿、主惯量比以及特征点位置信息。经过仿真,姿态角度估计误差在稳定后低于0.3°,相对位置估计误差在稳定后低于0.5m,相较于真值,误差小于1.67%,主惯量比估计误差在稳定后低于0.01,特征点位置误差在稳定后低于0.005m。在引入相机偏置条件后,滤波状态变量均收敛,并得到具有足够精度的估计,成功解决了单目相机深度信息缺失问题。