恶劣天气下的仿生偏振定向方法

针对偏振光导航在恶劣天气下精度显著下降的问题,提出了一种可以在恶劣天气下基于大气偏振模式的定向算法。与现有方法相比,首次将三维块匹配与canny边缘检测结合的思想应用于修复被不同天气破坏的偏振角度图像中。具体而言,将偏振角度图像的修复分为噪声粗处理和边缘提取处理去噪两部分。在噪声粗处理阶段采用三维块匹配算法,在边缘提取处理去噪部分,利用canny边缘检测算法对偏振角度图像进行二次去噪。实验结果表明,该方法不仅能够提高晴朗天空下的定向精度,而且能够显著提升在阴天、沙尘、雾霾等恶劣天气条件下的导航定向精度,即使在偏振角图像对称∞模式被破坏的情况下,航向角精度仍可由9.4470°提高到1.6859°。     

一种采用行列聚类并行的星像坐标提取方法

为了提高星敏感器的数据更新率,必须减少星像坐标的提取时间。在提取星像坐标的过程中,现有基于FPGA的方法虽然避免了星图保存和读取过程,然而需要高性能的FPGA才能实现。为此提出了一种采用行列聚类并行的星像坐标提取方法,将逐个像素聚类的方式优化成像素队列聚类的方式,实现了对恒星星像坐标的实时提取,可以解决现有的实时提取星像坐标方法中像素输出时钟与处理时钟不匹配的问题,达到1个时钟周期处理完1个像素的目的,从而减少了星像坐标的提取时间,提高了星敏感器的实时性。最后把该方法嵌入到某星敏感器系统中,实验结果表明,与现有基于FPGA的实时提取星像坐标方法相比,在相同工作频率且不降低星像坐标提取精度的前提下,提取时间仅为现有方法的21.1%,数据更新率提高了近5倍。因此,该方法不但具有实时性,而且具有较好的鲁棒性。     

融合明度特征的V-SAD立体匹配算法

针对SAD算法难以对图像中纯色与弱纹理部分进行准确匹配的问题,提出了将HSV空间明度特征与SAD算法相融合的立体匹配算法,称为V-SAD算法.首先将图像从RGB空间转化至HSV空间,并根据H、S、V值将像素点按照颜色分为10类,同时得到明度特征图.然后结合SAD算法需要的灰度特征图计算匹配代价,得到初步的视差图.接着,根据HSV空间的颜色信息对图像进行分割,结合数学形态法求解分割区块中的独立连通域.再利用边缘生长法对每一个连通域的视差进行恢复.最后,使用左右一致性检测方法对视差图进行优化.实验结果表明,利用图像的明度信息衡量纯色与弱纹理区域的匹配点的相似度是有效的,该V-SAD算法有效改善了SAD算法在弱纹理和纯色区域的匹配效果,平均误匹配率下降了13.02％.     

Method for utilizing PIV to investigate high curvature and acceleration boundary layer flows around the compressor blade leading edge

Particle Image Velocimetry (PIV) is a well-developed and contactless technique in experimental fluid mechanics, but the strong velocity gradient and streamline curvature near the wall substantially limits its accuracy improvement. This paper presents a data processing procedure combining conventional PIV and newly developed Mirror Interchange (MI) based Interface-PIV for the measurement of the boundary layer parameter development in the blade leading edge region. The synthetic particle images are used to analyze the measurement errors in the entire procedure. Overall, three types of errors, namely the errors caused by the Window Deformation Iterative Multigrid (WIDIM) algorithm, the discrete data interpolation and integration, and the wall offset uncertainty, comprise the main measurement error. Specifically, the errors due to the discrete data interpolation and integration and the WIDIM algorithm comprise the mean bias, which can be corrected through the error analysis method proposed in the present work. Meanwhile, the errors due to the WIDIM algorithm and the wall offset uncertainty contribute to the measurement uncertainty. Computational fluid dynamics-based synthetic particle flows were generated to verify the newly developed PIV data processing procedure and the corresponding error analysis method. Results showed that the data processing method could improve the accuracy of PIV measurements for boundary layer flows with high curvature and acceleration and even with significant flow separation bubbles. Finally, the data processing method is also applied in a PIV experiment to investigate the boundary layer flows around a compressor blade leading edge, and several credible boundary flow parameters were obtained.     

Measurement-driven Gauss-Hermite particle filter with soft spatiotemporal constraints for multi-optical theodolites target tracking

Multi-Optical Theodolite Tracking systems (MOTTs) can stealthily extract the target’s status information from bearings only through non-contact measurement. The constrained MOTTs are partially compatible, yet many existing research works and results are based on the known model, ignoring its discrimination with the target maneuvering behavior pattern. To compensate for these mismatches, this paper develops a Measurement-driven Gauss-Hermite Particle Filter (MGHPF), which elegantly fuses the spatiotemporal constraints and its soft form to perform MOTT missions. Specifically, the target dynamic model and tracking algorithm are based on the target behavior pattern with the adaptive turn rate, fully exploiting the spatial epipolar geometry characteristics for each intersection measurement by a minimax strategy. Then, the center of the feasible area is approximated via the analytic coordinate transformation, and the latent samples are updated via the deterministic Gauss-Hermite integral method with the target’s predictive turn rate. Simultaneously, the effects of truncation correction and compensation feedback from the current measurement and historical estimation data are adaptively incorporated into the PF’s importance distribution to cover the mixture likelihood. Besides, an effective causality-invariant updating rule is provided to estimate the parameters of these soft spatiotemporal constrained MOTTs with convergence guarantees. Simulated and measured results show good agreement; compared with the state-of-the-art Multi-Model Rao-Blackwell Particle Filter (MMRBPF), the proposed MGHPF improves the filtering accuracy by 7.4%-34.7% and significantly reduces the computational load.     

基于单目视觉的三维重构相机技术研究

针对飞行器执行在轨服务获取空间多维图像的需求，提出了基于单目视觉的三维重构相机技术，利用基于亮度法体素着色原理实现空间目标的三维重构。搭建地面试验，设计模拟圆轨道和椭圆轨道多种试验，验证了在200 m范围内伴飞状态下,实现对目标的高分辨率三维重构图像，重构精度为0.105 m，为识别空间目标特征部位奠定技术基础。     

Evolution of turbulent boundary layer over a three-dimensional bump

A bump is typically used in the inlet system of an aircraft engine to compress the incoming airflow and to reduce boundary layer thickness developed over fuselage. In this work, the turbulent flow over a three-dimensional bump is experimentally studied. The bump model is mounted in a closed return wind tunnel operated at the nominal velocity 10 m/s, corresponding to a friction Reynolds number of 2300. The flow field upstream the bump, along the bump centerline and at two different spanwise plane...     

基于条件生成对抗模型的遥感卫星影像重建技术研究

为了快速侦察未知区域的地貌信息，遥感卫星可对特定区域进行扫描以获取遥感卫星影像。当卫星经过国外未知区域时，部分卫星无法针对某特定区域进行长时间的驻留扫描，本文提出一种基于条件生成对抗网络模型（Conditional Generative Adversarial Network，CGAN）进行网络训练，前期将某方法获取的区域轮廓地形信息作为CGAN网络的生成网络和鉴别网络中的条件约束信息，通过网络生成器与判别器在训练过程中互相博弈产生特定的输出集，有效地实现由单张电子轮廓图像到对应卫星遥感图像的端到端的非线性映射。本文通过原真实卫星遥感图像与生成卫星遥感图像进行四种对比误差计算，平均误差、均方误差与结构相似度均高于99%，峰值信噪比高于30 dB，生成的图像与原图像之间具备高相似度，实现了在获取坐标定位轮廓信息的先验条件下，对特定区域进行遥感卫星影像内容重建技术。     

Numerical simulations of a continuously injected relativistic electron beam relaxation into a plasma with large-scale density gradients

In this paper the influence of large-scale decreasing and increasing gradients of the density of magnetized plasma on the relaxation process of a continuously injected relativistic electron beam with an energy of 660 keV ($v_b=0.9c$) and a pitch-angle distribution is studied using particle-in-cell numerical simulations. It is found that for the selected parameters in the case of a smoothly decreasing gradient and in a homogeneous plasma the formation of spatially limited plasma oscillations of large amplitude occurs. In such cases, modulation instability develops and a long-wave longitudinal modulation of the ion density is formed. In addition, the large amplitude of plasma waves accelerates plasma electrons to energies on the order of the beam energy. In the case of increasing and sharply decreasing gradients, a significant decrease in the amplitude of plasma oscillations and the formation of a turbulent ion density spectrum are observed. The possibility of acceleration of beam electrons to energies more than 2 times higher than the initial energy of the beam particles is also demonstrated. This process takes place not only during beam propagation in growing plasma density, but also in homogeneous plasma due to interaction of beam particles with plasma oscillations of large amplitude.