基于滑动窗口的三维机动目标在线Kalman滤波研究

对机动目标的在线跟踪是军民领域中的重要研究课题。基于雷达数据设计了一个实现三维机动目标在线跟踪的模型。首先,针对一类三维机动目标的跟踪模型,利用窗口平滑的方法获得了运动方程噪声的功率谱密度矩阵;其次,通过构造目标的运动方程和观测方程,利用Kalman滤波给出目标的最优估计。数值例子表明,对于高机动目标,该方法能够提供比较好的跟踪效果。     

两种跟踪机动目标的方法研究

飞行目标一般沿圆周路径作机动飞行,因而出现了以圆周运动为基础的跟踪滤波器。本文讨论了飞行目标进行转变机动时的两种跟踪方法,即机动中心坐标系法和高阶多项式模型法,对二者给出了性能评价,并给出了目标在标准转弯进入方式下的仿真结果。     

密集杂波环境下一种新的机动目标跟踪算法

针对密集杂波环境下的机动目标跟踪问题,在传统方法基础上结合模糊推理的思想,提出一种模糊自适应交互多模型概率数据关联算法。该算法将目标运动模式分为机动和非机动两类区别对待,无需选取大量模型来覆盖机动目标的运动模式,而是动态调节模型集合中部分模型的参数,具有更强的自适应跟踪能力。最后,给出了算法的仿真分析,结果表明:该方法能够有效地跟踪密集杂波中的机动目标,并且跟踪性能较传统方法有所提高。     

基于反馈线性化及滑模控制的俯冲机动制导方法

以高超声速飞行器为研究对象,针对俯冲段精确制导及机动突防问题,基于反馈线性化与滑模控制研究了机动突防滑模跟踪制导方法。首先设计纵向俯冲及侧向机动弹道,其次利用反馈线性化将非线性运动方程转化为线性方程,基于该线性方程利用滑模控制对已设计的弹道进行跟踪,最终将线性跟踪制导律转换到非线性系统中获得非线性滑模跟踪制导律,该制导律完全基于飞行器当前运动状态,所需的相对运动信息大大减少。CAV-H飞行器制导实例仿真表明,该方法能够实现俯冲段精确制导及机动飞行,且对初始及过程偏差具有较强的鲁棒性,能够为高超声速飞行器俯冲段制导提供有益参考。     

机动目标跟踪的模式集自适应IMM算法的设计和比较

变结构多模型状态估计器方法适用于雷达跟踪空中机动目标。本文研究和设计了两种自适应时变模式集交互多模型(IMM)跟踪滤波器;切换网格(SG)IMM和自适应网格(AG)IMM算法。针对不同的飞行想定,通过Monte Carlo仿真方法对这些算法及相应的固定网格(FG)IMM滤波器的性能进行了评估和比较。仿真显示,对特定的机动目标跟踪问题,SGIMM和AGIMM跟踪滤波器在性能和计算量等方面要比固定结构多模型算法(FGIMM滤波器)有明显地提高。     

一种改进的当前统计模型自适应跟踪算法

针对"当前"统计模型算法跟踪弱机动目标时存在较大跟踪误差,同时对机动加速度极限值存在依赖的缺陷,在分析传统"当前"统计模型适用范围的基础上,提出一种改进的"当前"统计模型自适应跟踪算法。算法首先通过泰勒级数展开和忽略高阶项的方法得到模型加速度极值和均值的修正结果,使算法能更好地描述强机动和弱机动2种情况下目标的机动特征。之后,通过引入强跟踪滤波器,进一步增强模型对目标突发机动的自适应跟踪能力。最后,针对具有周期性运动特征的跳跃巡航目标,与IMM,CS算法仿真比对,验证了算法的可行性和合理性。     

基于修正转弯模型的交互多模型跟踪算法

临近空间高超声速飞行器具有极快的飞行速度和独特的机动模式,传统单模型跟踪方法不能对此类目标进行有效跟踪。提出了基于修正转弯模型的交互多模型跟踪算法,该算法修正了转弯模型,利用当前统计模型思想对目标角速度进行实时估计。首先分析了临近空间高超声速目标的特性,然后根据目标运动方程,对目标角速度进行实时估计,最后通过交互多模型实现对临近空间高超声速目标的跟踪。蒙特卡罗仿真结果表明,IMM跟踪算法能够有效跟踪临近空间高超声速目标,且跟踪的精确度、稳定性等明显优于单模型。     

自适应交互多模型算法在机动目标跟踪中的应用

针对多模型算法在机动目标跟踪中存在的问题,运用交互多模型算法(IMM)和自适应滤波理论,设计了一种自适应交互多模型算法(AIMM),结合目标运动模型对目标当前加速度和其方差进行估计,并在此基础上给出了AIMM中模型集和模型转移概率的设计方法,进行了计算机仿真.蒙特卡罗仿真结果表明,与标准IMM算法相比,该算法比IMM算法的跟踪性能有很大提高,跟踪复杂机动目标比IMM有更快的收敛速度,跟踪滞后问题得到较好的解决,跟踪目标的稳定性和精确性均优于IMM算法,有利于机动目标的实时跟踪.     

图像跟踪系统中机动目标预测的实现

针对装甲车上的图像跟踪系统,提出了一种机动目标预测方法.将目标的运动分解为两部分,即全局运动和局部运动,并根据各自的动态特性分别进行预测.以局部运动预测为重点,通过对机动目标跟踪技术的研究,建立起性能优良的跟踪滤波器,并将其扩展到预测滤波器中对目标的未来运动状态进行预测,从而提出了"交互式多模型预测算法".仿真结果表明,该算法具有预测精度高、自适应能力强的特点,同时给出了计算方法.     

密集回波环境下的一种交互式多模型算法

利用神经网络算法对基于机动目标“当前”统计模型的均值和方差自适应滤波算法进行修改 ,提高该算法的性能 ;然后与匀速模型交互作用 ,利用概率数据关联处理密集回波环境下机动目标的跟踪问题 ,这样既保持了对强机动目标的跟踪性能 ,同时又提高了对弱机动目标的跟踪精度。仿真结果证明该算法确为一种非常有效的方法     

Reviews

Douglas S. Derrer, We Are All the Target: A Handbook of Terrorism Avoidance and Hostage Survival. Annapolis, MD: US Naval Institute Press, 1992. Pp. x+135, notes, index. $14.95. ISBN 01–55750–150–5

Ian Knight, Zulu: Isandlwana and Rorke's Drift, 22–23 January 1879. London: Windrow & Greene, 1992. Pp.136, 150 illus., incl 8 colour plates, maps, biblio. £35. ISBN 1–872004–23–7

Ian Knight, By the Orders of the Great White Queen: Campaigning in Zululand through the Eyes of the British Soldiers, 1879. London: Greenhill Books and Novato: Presidio Press, 1992. Pp. 272, 17 illus., 1 map. £18.95. ISBN 1–85367–122–3

Manfried Rauchensteiner and Erwin A. Schmidl (eds.), Formen des Krieges: vom Mittelalter zum ‘Low‐intensity’ Conflict’. Graz : Verlag Styria, 1991. Pp.208. DM35. ISBN 3–22–12139–7

Harold J. Kearsley, Maritime Power and the Twenty‐First Century. Dartmouth: Dartmouth Publishing Company, Limited, 1992. Pp.xv + 203, 13 diagrams, index. £32.50. ISBN 1–85521–288–9     

Panjwai: a tale of two COINs in Afghanistan

The US Army has two approaches to counterinsurgency operations in Afghanistan. One is hard, or combat-focused, and the other is soft, or development-focused. This study examines two US Army task forces deployed to Panjwai District, Afghanistan from 2012 to 2013. CTF 4-9 and 1-38 offer a meaningful comparison because they pursued these contrasting approaches among the same population and against the same enemy at the same time and place. The study compares each unit’s approach and finds that neither approach was successful absent the other. The article concludes by recommending further research into combining the approaches at the operational level.     

线性核时频信号表示及其仿真

根据ｃｈｉｒｐ信号在模糊域的特点,本文提出了线性核时频表示方法。这种方法在提高分辨率、消除交叉项以及抑制噪声等方面都具有较高的性能。理论分析和实验结果都证实了这种方法的有效性     

Putin – the masked nemesis of the strategy of ambiguity

Recent events demonstrate the complex and adaptive approach employed by Russia to reassert influence in Europe. The changing face of Russia’s strategy commenced in 2007 when it launched a crippling cyber-attack against Estonia. This was followed by a large Russian conventional attack against Georgia in 2008, occupying two large areas of the nation. 2014 witnessed the Russian annexation of Crimea where in just a week, Russia seized control of Crimea “without firing a shot.” The annexation of Crimea was rapidly followed by a Russian inspired and led subversive war in eastern Ukraine. The common thread among these diverse Russian operations is its use of ambiguity to confound and confuse decision makers in the West.     

Forecasting a Hurricane: Israeli and American Estimations of the Khomeini Revolution

The surprising ‘Arab Spring’ raises the question as to what would enable national intelligence to provide high quality warnings prior to the eruption of popular revolutions. This article uses new sources of evidence to trace and explain Israel's success in comparison to US failure at correctly estimating the course of the Iranian Revolution in 1977–79. In explaining this variance, the article shows that it was mainly the result of the intimate acquaintance of Israel's representatives in Iran with the local language, history and culture, as well as the ability to communicate with locals – tools which the Americans completely lacked.     

关于教师继续教育培训模式的思考

继续教育作为终身教育的重要组成部分和基本形式,对教师的发展成长意义重大。继续教育工作开展十多年来,取得了一定的成绩,但还存在不少问题,如学习动机较为被动,管理认识不到位,教学内容陈旧空泛,教学方法呆板机械等。随着课程改革的不断深入,继续教育必须采取相应的改进措施,树立以人为本,服务至上的观点,教学内容的前瞻性与实用性并重,教学人员进行整合,教学方法灵活多样,评价考核规范科学等,以此激活受训教师内在的学习动机,更大程度地提高继续教育培训的实际效果。     

用DirectShow实现网络视频

目前市场上的网络监控技术一般用昂贵的硬件设备来支撑,投资较大。在深入研究微软的DirectX9.0的基础上,提出如何用DirectShow软件技术和普通的数字摄像头来实现网络视频的监控和实现技术,内容涉及通信协议、视频的编码与解码技术、视频的编码格式和COM组件编程技术等,最终的实验结果令人满意,且费用低廉,技术可靠,为在高速局域网内构建视频监控系统提供有效的技术支持。     

Interpreting the Ribchester Helmet

In 1796 a 13-year-old boy playing on wasteland behind his father’s house in Ribchester, Lancashire discovered an assemblage of over 30 Roman artefacts, including a decorated brass helmet. Although partially corroded the Ribchester helmet is a fine example of a first-century AD cavalry sports helmet. This article considers how the interpretation of the helmet has evolved since its discovery. The multi-faceted iconography of the helmet with its mural crown diadem is that of an elite soldier and is paralleled in other examples of cavalry sports helmets of the Ribchester type. The long-neglected imagery on the helmet bowl is reinterpreted and seen not as a haphazard collection of randomly placed combatants but as a highly structured, symmetrical composition specifically designed to be viewed from different directions.     

Desert insurgency: lessons from the third Tuareg rebellion

This article provides an in-depth examination and analysis of the 2006–2009 Tuareg rebellion in Mali and Niger. It identifies the underlying reasons behind the rebellion, explores contrasting counter-insurgency (COIN) strategies employed by the two governments, and presents some lessons learned. While both COIN approaches ultimately produced similar peace settlements, the article argues that the Malian strategy of reconciliation combined with the selective use of force was far more effective than the Nigerien iron fist approach at limiting the size and scope of the insurgency and producing a more sustainable peace. It concludes by looking at the role of external actors, particularly the United States, and how the failure to internationalize the conflict was actually more beneficial to the local COIN effort, as well as to the longer strategic interests of the United States in the region.