基于遗传算法的协同多目标攻击空战决策方法

多机协同多目标攻击是未来空对空作战的一种重要形式。首先建立了多机协同空战的自主优势矩阵 ,并依据多人冲突决策理论构造了空战的总体优化指标向量 ,然后针对其他优化算法的不足 ,提出用遗传算法优化该指标向量 ,实现多机协同多目标攻击空战决策 ,最后对 2∶ 2空战进行了仿真。仿真结果证明了上述思想的正确性。     

An interactive procedure for selecting acceptable alternatives in the presence of multiple criteria

In this paper, we consider the multiple criteria decision‐making problem of partitioning alternatives into acceptable and unacceptable sets. We develop interactive procedures for the cases when the underlying utility function of the decision maker is linear, quasiconcave, and general monotone. We present an application of the procedures to the problem of admitting students to the master's degree program at the Industrial Engineering Department, Middle East Technical University. © 2001 John Wiley & Sons, Inc. Naval Research Logistics 48: 592–606, 2001.     

不同热解温度石英杂化酚醛复合材料渗透率测试

为得到不同温度下石英杂化酚醛材料的渗透率,自主搭建了材料气体渗透率测试平台,提出基于Darcy定律的复合材料气体渗透率的测试方法。对不同温度下石英杂化材料进行研究,测量渗透过程中试验件上下表面气体压力变化和流过试验件的气体流量,进而得到复合材料的渗透率。结果表明:该实验平台可以用来测量复杂孔隙复合材料的渗透率。石英杂化酚醛材料渗透率与热解温度呈正相关,热解温度越高,复合材料的渗透率越大,材料渗透率和热解温度满足关系式K=9.5×10-15T-6.32×10-12。673 K热解温度下,复合材料渗透率为10-13 m²量级;873 K和1 073 K热解温度下,其渗透率在10-12 m²量级。本试验结果丰富了该类树脂基复合材料的基础物性数据,为材料渗透和热质扩散特性分析提供了依据。     

无人机先进地面站关键技术综述

具有沉浸显示、智能辅助、自然化人机交互等先进控制能力的新型无人机地面站已成为当前无人机控制领域的研究热点。为分析其中的技术脉络,系统性地梳理国内外一系列无人机先进地面站的功能要点及设计理念,在此基础上从无人机地面站指挥控制的观察—判断—决策—行动回路出发,归纳提炼了其技术体系构成,分析指出了其中的任务环境构建、战场态势沉浸式显示、智能化辅助决策和自然化人机交互等关键技术,并对各项技术的主要研究方法进行了深入剖析,还对无人机先进地面站目前存在的挑战和未来发展趋势进行了研判。该研究对新型地面站的研制具有指导和借鉴意义。     

专家系统和多目标决策理论在抗沉决策系统中的应用

应用专家系统理论建立了抗沉规则库,确定了方案产生逻辑,同时运用多目标决策理论中的简单加性加权法建立了数学模型,对多方案抗沉效果进行了综合评估．成功地解决了舰艇生命力指标间的相互制约、相互影响的问题,为抗沉优化决策提供了新的思路和方法．     

异构混合存储的软硬件协同数据放置策略

分析比较当前大数据中心典型混合存储架构,针对其不能综合运用存储管理系统和存储设备优势的问题,提出软硬件协同的数据放置策略,同时考虑软件层混合存储管理系统和硬件层混合存储设备的特点,根据应用特性为数据选择合适的存储管理系统和设备。面向不同应用场景,提出运行前规划存储路径的静态放置模式和运行中规划存储路径的动态放置模式。基于存储管理系统和设备性能参数建模,采用仿真方法实现各数据放置策略,并运行实际应用中的三类负载进行实验测试,结果表明采用软硬件协同的数据放置策略相比只考虑存储管理系统和存储设备的数据放置策略,性能可以提高近30%。     

利用复单脉冲比的平面阵列雷达群目标检测方法

针对平面阵列雷达群目标检测问题,以矩形阵为例,详细分析方位向和俯仰向复单脉冲比虚部的统计特性;建立平面阵列雷达群目标检测模型,根据恒虚警检测原理推导复单脉冲比虚部的椭圆判决域,并利用复单脉冲比实部辅助判决,综合提出基于复单脉冲比的平面阵列雷达群目标检测算法。针对双目标情形进行仿真实验,结果表明:在较高信噪比、较大角度间隔以及幅度相当且相位差较大的条件下算法能够有效检测群目标。     

From Westphalian Failure to Heterarchic Governance in MENA: The Case of Syria

Abstract

The problematic export of the Westphalian system to MENA is examined, taking Syria as exemplar. The export model is juxtaposed to actual non-lineal trajectories, semi-sovereignty and hybrid or failing states. This is manifested in post-uprising Syria in failing statehood, fragmented and overlapping governance, permeable and collapsing borders, the loss of sovereignty to trans-state movements, “competitive regime-building” between the Asad regime and jihadist warlords, and “competitive interventionism” by external powers filling the governance vacuum with their own proxies. The result is heterarchic zones of limited statehood in which state sovereignty is contested by both international (supra-state) penetration and sub-state fragmentation.     

Impacts of sequential acquisition,market competition mode,and confidentiality on information flow

We investigate information flow in a setting in which 2 retailers order from a supplier and sell to a market with uncertain demand. Each retailer has access to a signal. The retailers can disclose signals to each other (horizontal information sharing), while the supplier can solicit signals by offering retailers differential payments as incentives for signal disclosure (vertical information acquisition). In the base setting, market competition is in quantity, and a retailer can fully infer the signal that the other retailer discloses to the supplier. We show that the supplier prefers to sequentialize the procedure for information acquisition. Moreover, vertical information acquisition by the supplier is a strategic complement to horizontal information sharing between the retailers to establish information flow. In the equilibrium, the retailers have no incentive to exchange signals, but system wide information transparency can be realized through a combination of information acquisition and inference. We further study the signaling effect, whereby the supplier utilizes wholesale pricing as an instrument to affect the retailers' inference of the shared signals, and price competition to explore their impacts on the supplier's preference for sequential acquisition and the sustainability of information flow.     

集中控制混合网络中基于流的资源分配算法

集中控制混合网络中,异构化网络内部的流量具有一定的规律和特性,如果使用原有的单一的离散式最大极值和无状态的网络资源调度算法,忽略了异构网络规律,会造成网络利用率较低、易震荡、部分网络流延迟等问题。通过分析由集中控制网络和普通网络组成的混合网络的拓扑结构,对混合网络结构中常见问题如流闪现、不能估计的流、路径堵塞或连接震荡场景进行分析,并提出基于期望和状态的流量评价资源规划算法POS和POS-FME。算法考虑混合网络的运行状态,对系统可用资源进行评估,为系统中各种流匹配对应可用资源,并具有一定的预测作用,从而避免混合网络出现运行效率低下的场景。通过实验,POS算法和POS-FME算法相对传统算法,利用率提高了10%～30%,并降低了震荡和平均延迟。