动态威胁环境下单无人机测向定位航迹优化算法

为解决单架无人机在动态战场环境下的测向定位问题,提出了一种基于动态窗口法的单机测向定位航迹优化算法。以最大化Fisher信息矩阵行列式为测向定位评价准则,在由动态探测雷达和静/动障碍构成的动态战场环境中,基于动态窗口法思想,将测向定位航迹优化评价准则由传统的单步最优原则扩展到对多步预测航迹的评价,同时考虑雷达探测和静/动障碍环境对预测航迹的影响,通过滚动时域方法控制无人机最优航向。仿真结果表明,所提方法能够使无人机在有效逃避雷达探测威胁以及规避环境中静/动障碍的条件下保证对目标的高精度测向定位,为解决动态战场环境下的单架无人机测向定位问题提供了新思路。     

激光辐照下45#钢板响应特性

采取实验和数值模拟相结合的方法对激光辐照45#钢板的响应特性进行研究。基于连续光纤激光发射系统,开展钢板的激光辐照实验,得到温度场分布及靶板的烧蚀穿孔特性;基于ANSYS建立靶板有限元模型,模拟激光辐照下靶板的热响应过程,得到了靶板厚度、激光功率和光斑直径等因素对靶板温度分布及靶板穿孔特性的影响规律。     

【专题】基于深度学习及FPGA的装备目标检测研究

本文应用深度学习技术实现海天背景下基于可见光、红外方式成像的舰船及角反、烟幕干扰的目标检测,这也是反舰导弹作战使用的关键技术之一。采集的可见光与红外成像目标检测数据集涵盖实施典型干扰下的态势场景,贴近实战;结合四种不同的目标检测机制,选取YOLOV3、Faster R-CNN、SSD及CenterNet四种典型模型分别进行训练与验证,通过对比分析进一步提高弱小目标、复杂干扰态势的的检测,可以实现端到端的高精度装备目标检测模型。在确保精度的前提下基于现场可编程门阵列（FPGA）进行软硬件协同设计,通过对比分析选定基于Vitis AI的实施方案,经过模型的量化、编译与优化,可在保证检测效率的前提下快速实现模型的小型化部署,便于进行装备移植。研究结果表明,该研究内容可有效提高现役反舰导弹目标检测的准确率。     

Rigid elliptical cross-section ogive-nose projectiles penetration into concrete targets

The elliptical cross-section ogive-nose projectile (ECOP) has recently attracted attention because it is well suited to the flattened shape of earth-penetrating weapons. However, the penetration performance of ECOPs has not been completely understood. The objective of this study was to investigate the pene-tration performance of ECOPs into concrete targets using a theoretical method. A general geometric model of ECOPs was introduced, and closed-form penetration equations were derived according to the dynamic cavity-expansion theory. The model was validated by comparing the predicted penetration depths with test data, and the maximum deviation was 15.8％. The increment in the penetration depth of the ECOP was evaluated using the proposed model, and the effect of the major—minor axis ratio on the increment was examined. Additionally, the mechanism of the penetration-depth increment was inves-tigated with respect to the caliber radius head, axial stress, and resistance.     

Blast resistance evaluation of urban utility tunnel reinforced with BFRP bars

To improve corrosion-resistance of shallow-buried concrete urban utility tunnels(UUTs),basalt fiber reinforced polymer(BFRP)bars are applied to reinforce UUTs.As the UUT must have excellent survival capability under accidental explosions,a shallow-buried BFRP bars reinforced UUT(BBRU)was designed and constructed.Repetitive blast experiments were carried out on this BBRU.Dynamic responses,damage evolutions and failure styles of the BBRU under repetitive explosions were revealed.The tunnel roof is the most vulnerable component and longitudinal cracks develop along the tunnel.When the scaled distance is larger than 1.10 m/kg1/3,no cracks are observed in the experiments.When the BBRU is severely damaged,there are five cracks forming and developing along the roof.The roof is simplified as a clamped-supported one-way slab,proved by the observation that the maximum strain of the transverse bar is much larger than that of the longitudinal bar.Dynamic responses of the roof slab are predicted by dynamic Euler beam theory,which can consistently predict the roof displacement under large-scaled-distance explosion.Compared with the UUT reinforced with steel bars,the BBRU has advantages in blast resistance with smaller deflections and more evenly-distributed cracks when the scaled distance is smaller than 1.260 m/kg1/3 and the steel bars enter plastic state.Longer elastic defamation of the BFRP bars endows the UUT more excellent blast resistance under small-scaled-distance explosions.     

雷达空间目标识别仿真系统设计

设计了基于雷达观测的空间目标识别仿真系统,分析了主要功能组成、仿真系统技术关键,介绍了ISAR成像技术及雷达空间目标RCS预估系统的组成和方法,仿真实践证明,该系统具有一定的实用价值。     

基于U-D分解滤波的交互多模型算法

简要回顾了基于相控阵雷达的三维空间中机动目标建模与自适应跟踪的机理,针对交互多模型算法在其跟踪应用过程中可能会出现的滤波发散问题,提出了一种基于U-D分解的交互多模型目标跟踪算法.该算法可以有效地限制交互多模型算法滤波计算过程中的滤波发散的可能性,从而使得IMM算法更加稳定.     

目标打击价值的灰色决策分析

目标价值分析是舰艇指控系统进行目标分配和火力通道组织的基础性工作.为了评估目标价值大小,从分析影响目标价值大小的主要因素着手,将其归结为多目标决策中的方案排序问题,用灰色关联分析法建立了战场多目标打击顺序的评估模型.实例计算分析表明,所提出的方法较好地解决了评估中的多指标问题,计算结果与实际经验判断非常吻合,弥补了指挥员主观判断上的某些缺陷,为舰艇指控系统适时进行目标打击价值分析提供了一种有效的参考处理方法.     

激光成像雷达空间目标探测与识别技术

研究了激光三维成像雷达在动能拦截器空间目标探测与识别方面的应用。首先介绍了激光3D成像雷达系统基本原理,然后设计了激光3D成像雷达目标识别的工作流程,最后提出了基于旋转图像转换的3D目标识别算法。进一步探索了红外+LADAR主被动复合光学成像导引头在空天防御拦截武器目标探测与识别的应用,分析了复合光学导引头的关键技术,给出了相应的成像结果。     

高斯马尔可夫融合算法在卡尔曼滤波中的应用

针对目前常用的按标量、对角阵、矩阵加权融合方法过程较复杂,计算负担较重的问题,结合三维目标跟踪模型提出一种高斯马尔可夫融合方法来融合卡尔曼滤波估计值,不需要计算局部稳态滤波误差互协方差,只需要知道传感器的观测噪声方差就可以了。计算方法和操作相对简单,省略了很多繁琐步骤也能达到不错的滤波融合效果,通过仿真证明了此方法的可行性和高效性。