基于移相器控制电压的“动中通”波束形成改进算法

在波束形成技术中,基于同时扰动随机梯度估计算法计算接收功率对移相器控制电压的梯度,研究了同时扰动随机梯度估计算法中扰动量和电压更新方程中步长的取值问题。仿真结果表明,随迭代次数变化的扰动量较固定扰动量能有效地提高算法的收敛性能,在同时扰动随机梯度估计算法中,将步长改为随阵元空间位置变化,能有效地提高电压更新方程的收敛速度。研究了与变扰动量有关的参数取值,指出了参数的选取准则。MATLAB仿真结果验证了所改进算法的正确性和有效性。     

改进的混合蛙跳算法在雷达网部署中的应用

针对雷达组网优化部署的原则,建立了区域雷达组网优化部署的数学模型,提出了一种基于改进的蛙跳算法的雷达组网优化部署方法。在阐述了改进方法的基础上,与先前方法进行了对比分析。最后,通过计算机仿真试验进行了验证,结果表明:改进的混合蛙跳算法能够快速得到多种优化部署方案,且操作性强,能较好地应用于雷达组网优化。     

Planning for the next Takur Ghar

The USA is currently facing an increasingly diverse range of threats, including non-state actors, particularly violent extremist organizations attempting to do harm. The national leadership has made it clear that for the near future it will employ Special Operations Forces (SOF) as the tool of choice in short-duration, high-intensity operations that have more in common with the battle of Takur Ghar, than those with a larger footprint. It is critical then that the lessons of the past are learnt now and are applied before employing SOF in the future. This becomes especially important given the short lead-time in concert with the short-duration/high-intensity tempo of their operations. In the Battle of Takur Ghar, during Operation Anaconda, rigid adherence to the joint functions, especially clearer and more robust Command and Control (C2) and Intelligence, Surveillance, and Reconnaissance (ISR) in the planning and execution of Operation Anaconda would have yielded a better operational outcome and saved lives. With the benefit of hindsight, many things could have been done differently, but several key issues stand out. The C2 structure must include all components from the outset to synchronize effects and rapidly re-task ISR. There must be multiple communications redundancies and adequate bandwidth. Finally, all SOF, conventional and interagency capabilities must be synchronized under one task force commander with a sufficiently robust C2 infrastructure to direct the entire effort. As our nation calls on SOF at an ever-increasing rate, we must take those lessons learned and apply them to future operations.     

How and why Boko Haram blossomed: examining the fatal consequences of treating a purposive terrorist organisation as less so

Boko Haram (BH) is an insurgent group that operates mainly in northeastern Nigeria. Its stated aim is to establish an Islamic state, and it employs terrorism as its strategy. Earlier interests of security analysts centred on the drivers of BH uprising and the possibility of its internationalisation. Today's concerns relate more to why the rebellion has lasted this long. In toeing the same line, this article demonstrates that BH is a purposive terror group against which the government has evolved no viable strategy. It examines some gains of BH over the past six years and how it benefitted from the government's underestimation of its capacity and determination. It concludes that to effectively engage BH, the Nigerian government must revaluate its threat and sincerely pursue an expanded strategy beyond the current military-centric approach. To be effective, government's response must be packaged in such a way as to enhance human security in the region.     

基于保障时间窗的油料调拨运输模型

通过引入保障时间窗,同时考虑油料保障的时间约束和运力约束,建立了基于保障开始时间最早,并尽可能满足保障需求量的调度模型。针对问题的多目标性,基于理想点法将初始模型转化为单目标优化模型。采用粒子群算法对模型进行求解,并设计了算法编码和求解步骤。通过算例验证了模型和算法的可行性及有效性。     

针对应急救援路径规划的一种改进蚁群算法

针对不返回起始点、多个待救援点的应急救援路径规划问题,提出了一种应急救援路径规划的改进蚁群算法,设计了一种新的路径构造方法,为蚁群算法求解该类问题打下了基础。为提高收敛性,改进了信息素更新规则,构造了一种与蚁群算法有效结合的局部搜索算法,提高了算法快速寻优的能力。仿真结果表明:改进蚁群算法能够快速找到一条从救援中心到多个待救援点的优化路径,且收敛速度和最短路径较同类算法更优。     

喷气燃料中特征真菌Amorphotheca resinae的生物学性状研究

通过提供不同的碳源对喷气燃料中检测出的特征真菌Amorphotheca resinae(A.resinae)进行培养,观察菌落和孢子形态,研究其生物学性状;在液体培养基中对其进行振荡培养时出现了油包水现象,从而对特征真菌污染过的油、水相进行表面张力测定。研究结果表明:A.resinae在固体培养基中生长缓慢且出现角变现象;孢子直径仅约1μm;被特征真菌污染过的水相表面张力降低,推测培养A.resinae时可能产生了生物表面活性物质。     

基于改进粒子群优化算法的微电网孤岛选址定容

分析了在重要孤立系统中采用微电网供电的必要性,在此基础上提出了采用由非支配排序法和拥挤距离排序法改进的粒子群算法进行微电网选址定容,其目的是改进多目标微电网规划问题的优化结果。通过改进算法与基本算法算例结果的对比分析,验证了改进算法在求解孤岛选址定容问题上具有较强的寻优能力、较高的收敛精度和较稳定的最优解,为进一步研究微电网孤岛选址定容问题提供了思路。     

科氏流量计信号相位差测量新算法

对科氏流量计信号进行处理时,传统相位差测量算法存在较大误差。为更好地抑制频谱泄露,提高相位差测量精度,在传统DTFT算法的基础上,通过矩形窗和汉宁窗卷积运算构造出一种新窗——RHC窗。计及负频率成分的影响,提出了一种基于RHC窗的DTFT相位差测量新算法。详细阐述了新算法的原理,并与传统DTFT算法进行了对比仿真,验证了新算法的有效性。     

多核数字信号处理器矩阵乘卷积算法性能评测

矩阵乘卷积算法能够为各种卷积配置提供高性能基础实现,是面向给定芯片进行卷积性能优化的首要选择。针对国防科技大学自主研制的飞腾异构多核数字信号处理器(digital signal processor, DSP)芯片的特征以及矩阵乘卷积算法自身的特点,提出了一种面向多核DSP架构的高性能并行矩阵乘卷积实现算法ftmEConv。该算法由输入特征图转换、卷积核转换、矩阵乘以及输出特征图转换这四个均运行在通用多核DSP上的并行化部分构成,通过有效挖掘通用DSP核中功能单元的潜力来提升各个部分的性能。实验结果表明,ftmEConv实现了高达42.90%的计算效率,与芯片上的其他矩阵乘卷积算法实现相比,获得了高达7.79倍的性能加速。