军用飞机PHM系统一体化设计架构分析

军用飞机故障预测与健康管理（PHM）系统设计与工程应用存在设计标准缺失、复杂度高、准确率低等问题。通过对PHM系统数据、结构、功能维度的设计需求分析,提出三维多相迭代设计准则;基于三维多相设计准则对PHM系统设计流程进行梳理;提出基于软硬件协同加速的健康管理实时化、基于知识图谱的知识逻辑可视化、基于MBSE的PHM系统模型化的三化一体设计架构,并通过多领域模型联合仿真进行验证。结果表明：该设计架构实现了5.63倍的加速计算效果与82.7%的计算功耗缩减,提升了PHM系统知识与设计需求管理的可追溯性。     

NASA的智能协同系统

文章首先概述了智能协同系统在美国航空航天局(NASA)的应用情况,并以NASA普遍使用的两个典型软件工具PMT和NX为例,阐述了智能协同系统的基本功能和技术实现方法,为国内的智能化协同系统软件方面的研究提供参考。     

基于领域知识图谱的卫星故障辅助诊断与处置方法

本文重点针对传统故障诊断方法能力不足的缺点，开展基于领域知识图谱的卫星故障辅助诊断与处置方法研究，突破卫星故障领域知识图谱构建关键技术，关联固化卫星系统多维专家知识，构建卫星故障领域知识图谱，进而实现结合卫星故障知识信息的卫星故障推理分析和辅助诊断处置，提升复杂故障场景下的卫星故障诊断效率与处置能力。本文提出的方法为保障大气环境监测卫星的在轨长期稳定运行，提供了一种新的故障辅助诊断与处置措施。     

Formation flying for magnetic sails around artificial equilibrium points

This paper investigates the problem of magnetic sail-based spacecraft formation control around the artificial equilibrium points (AEPs), which can eliminate the requirement of the propellant. The thrusts are achieved by utilizing the interaction between the solar wind and the artificial magnetosphere generated by superconducting current coil onboard. The circular restricted three-body problem (CRTBP) of magnetic sail is discussed including the allowed regions and linear stability of AEPs, the locations of collinear AEPs and the possibility of existence of periodic orbits around the collinear AEPs. Next, the dynamical models of magnetic sail formation around the collinear AEPs are established. A novel fast fixed-time nonsingular terminal sliding mode controller (FFNTSM) based on fixed-time disturbance observer (FTDO) is developed to account for external disturbances. Several numerical simulations are conducted to substantiate that spacecraft formation can be precisely controlled by the proposed propellantless propulsion method in the presence of external disturbances.     

基于知识图谱的遥感图像检索定位方法

为了快速、准确地在海量遥感图像库中检索出所需的遥感图像资源，提出了一种基于知识图谱的遥感图像检索定位方法。为了充分利用遥感图像中所蕴含的地理语义信息，方法中使用知识图谱节点代表遥感图像，根据图像间的相似程度和地理位置在节点间建立关系。通过知识图谱缩小图像检索空间，加快检索速度，准确地定位遥感图像所示区域的地理信息，并返回最相似的图像。将此方法应用于卫星自主导航领域中，与传统的线性检索相比，图像检索的时间缩短了50%，准确率也得到提高。     

面向星上目标提取的卷积神经网络优化技术

针对卷积神经网络规模庞大、参数数量众多、在资源受限的在轨场景中难以应用的问题,提出了一种基于知识蒸馏的剪枝压缩改进方法。该方法对训练好的网络进行基于权重和基于通道的混合参数剪枝,在保留网络重要连接的同时剔除冗余信息;采用知识蒸馏法,用原始网络学到的知识指导剪枝后网络的再训练过程,以恢复损失的精度;在遥感数据集上对VGG-16分类网络进行实验。结果表明:所提方法可以实现16～18倍的压缩效果,并且网络精度下降不到1%。这使得卷积神经网络的在轨应用成为可能,具有理论及现实意义。     

A new interpretable fault diagnosis method based on belief rule base and probability table

It is vital to establish an interpretable fault diagnosis model for critical equipment. Belief Rule Base (BRB) is an interpretable expert system gradually applied in fault diagnosis. However, the expert knowledge cannot be utilized to establish the initial BRB accurately if there are multiple referential grades in different fault features. In addition, the interpretability of BRB-based fault diagnosis is destroyed in the optimization process, which reflects in two aspects: deviation from the initial expert judgment and over-optimization of parameters. To solve these problems, a new interpretable fault diagnosis model based on BRB and probability table, called the BRB-P, is proposed in this paper. Compared with the traditional BRB, the BRB-P constructed by the probability table is more accurate. Then, the interpretability constraints, i.e., the credibility of expert knowledge, the penalty factor and the rule-activation factor, are inserted into the projection covariance matrix adaption evolution strategy to maintain the interpretability of BRB-P. A case study of the aerospace relay is conducted to verify the effectiveness of the proposed method.     

基于点云深度学习的对称结构空间目标相对位姿测量

针对对称结构空间目标相对位姿解算过程中点云误匹配带来的误差问题，提出一种基于点云深度学习的对称结构空间目标相对位姿测量方法。首先设计空间目标点云特征提取网络及关键点回归网络，将位姿测量问题转换为空间目标点云关键点回归问题，通过两个并行的回归网络分别输出空间目标平移向量和具有固定标签的目标点云三维边界框角点；其次利用具有连续稳定标签的角点求解目标姿态，可有效解决目标的对称结构导致的点云误配准问题；最后通过仿真数据集的实验表明，该方法相比于传统的点云配准方法有更高的准确率，能够精确求解具有对称结构的空间目标相对位姿。     

面向视觉着陆的高精度结构化约束特征点研究

为提高视觉着陆过程中无人机的相对定位精度,选取视觉图像中的直线交点作为结构化约束特征点,设计了基于梯度一致性的边缘检测算法,并结合Shi-Tomasi角点检测算法进行结构化约束特征点的粗定位。对LSD直线检测算法进行改进并设计了亚像素角点定位精度改进算法,在结构化约束特征点粗定位的基础上,将其精度提高到亚像素级。基于实际场景中固有约束的结构化约束特征点具有鲁棒性、旋转和尺度不变性,抗干扰能力更强,其高精度定位有利于提高视觉着陆相对定位的精度与可靠性。     

基于可重构标定板的激光与视觉联合标定方法

目前,激光雷达与视觉传感器的联合标定方法包含动态在线标定与静态离线标定两大类.动态在线标定对标定的环境有较高的要求,且标定结果不稳定;静态离线标定通常采用标定板,因而对标定板的要求较高,手动选择与线拟合都易引入或放大误差,故提出了一种基于可重构标定板的激光雷达与视觉传感器的联合标定方法.首先,对传统的标定板进行拆分与重...