直接耦合干扰的自适应两级补偿方法

针对水下电磁探测系统对直接耦合干扰的补偿精度受限于硬件采样分辨率的问题,研究了基于自适应技术的直接耦合干扰两级补偿方法。采用高速 DSP 芯片实现直接耦合干扰补偿电路,完成直接耦合干扰的粗调补偿;提出了基于可变遗忘因子的 QRD-LS 算法,并采用基于该算法的自适应陷波器实现了直接耦合干扰的微调补偿。实验研究验证了两级补偿方法对直接耦合干扰的补偿性能。结果表明：基于可变遗忘因子的QRD-LS 算法稳定性好,收敛速度快;基于该算法的自适应两级补偿方法对直接耦合干扰的补偿精度达到10－4倍。     

水下爆炸载荷作用下夹芯圆柱壳的动态响应

采用有限元数值模拟方法,对夹芯圆柱壳在水下爆炸载荷作用下的动态响应进行了分析.在夹芯防护层质量给定的情况下,通过对不同夹芯层相对密度下内部圆柱壳体获得的最大加速度进行比较,得出了最优的夹芯层相对密度.同时,还对内部圆柱壳体的等效应力和夹芯层的变形进行了分析.研究结果表明:夹芯防护层对爆炸冲击波有较好的衰减作用,且对里面的圆柱壳体起到了较好的保护作用,具有优异的抗冲击防护特性.     

水下航行体惯性类水动力系数的计算方法

为了系统地计算水下航行体全套惯性类水动力系数,提高计算效率和计算精度,通过基于无粘模型的计算方法对水下航行体的运动进行了预报;通过UDF及动网格技术,对匀速运动和匀加速运动的SUBOFF模型进行了分析.设计了单方向速度线性变化的匀加速直线运动和匀加速回转运动,并通过换算和数值拟合处理得到了潜艇所受的惯性力和惯性类水动力系数.该系数与试验误差保证在6％之内,验证了所提方法的可行性及准确性.     

The impact of operational decisions on the design of salesforce incentives

When facing high levels of overstock inventories, firms often push their salesforce to work harder than usual to attract more demand, and one way to achieve that is to offer attractive incentives. However, most research on the optimal design of salesforce incentives ignores this dependency and assumes that operational decisions of production/inventory management are separable from design of salesforce incentives. We investigate this dependency in the problem of joint salesforce incentive design and inventory/production control. We develop a dynamic Principal‐Agent model with both Moral Hazard and Adverse Selection in which the principal is strategic and risk‐neutral but the agent is myopic and risk‐averse. We find the optimal joint incentive design and inventory control strategy, and demonstrate the impact of operational decisions on the design of a compensation package. The optimal strategy is characterized by a menu of inventory‐dependent salesforce compensation contracts. We show that the optimal compensation package depends highly on the operational decisions; when inventory levels are high, (a) the firm offers a more attractive contract and (b) the contract is effective in inducing the salesforce to work harder than usual. In contrast, when inventory levels are low, the firm can offer a less attractive compensation package, but still expect the salesforce to work hard enough. In addition, we show that although the inventory/production management and the design of salesforce compensation package are highly correlated, information acquisition through contract design allows the firm to implement traditional inventory control policies: a market‐based state‐dependent policy (with a constant base‐stock level when the inventory is low) that makes use of the extracted market condition from the agent is optimal. This work appears to be the first article on operations that addresses the important interplay between inventory/production control and salesforce compensation decisions in a dynamic setting. Our findings shed light on the effective integration of these two significant aspects for the successful operation of a firm. © 2014 Wiley Periodicals, Inc. Naval Research Logistics 61: 320–340, 2014     

Some comparisons of residual life of coherent systems with exchangeable components

Most of the research, on the study of the reliability properties of technical systems, assume that the components of the system operate independently. However, in real life situation, it is more reasonable to assume that there is dependency among the components of the system. In this article, we give sufficient conditions based on the signature and the joint distribution of component lifetimes to obtain stochastic ordering results for coherent and mixed systems with exchangeable components. Some stochastic orders on dynamic (or conditional) signature of coherent systems are also provided. © 2014 Wiley Periodicals, Inc. Naval Research Logistics 61: 549–556, 2014     

基于人工蜂群算法优化的改进高斯过程模型

高斯过程(GP)的非线性特征导致其对大样本的训练时间复杂度过高,而且其超参数的选取是否适当直接影响高斯过程回归模型的预测精度。提出采用人工蜂群(ABC)算法优化改进GP以减小时间复杂度和提高预测精度。改进GP通过选取训练样本的子样本进行模型学习,以降低训练过程的时间复杂度。ABC通过优化改进GP的超参数,提升预测精度。选取训练样本的子样本构建改进GP回归(GPR)模型,采用ABC算法搜寻改进GPR的最优超参数,并用得到的超参数构建最优的改进GPR模型,输入测试样本进行预测并输出预测精度。将该模型应用于解决海上远程精确打击(LPSS)体系作战效能评估问题中,通过MATLAB仿真实验,与常见的多种优化方法相比较,验证了该模型的有效性。     

基于数据站场的系统效能评估决策支持方法

随着信息技术的发展,系统信息量的提升给军事系统效能评估带来了新的问题。针对系统效能分析数据非线性、高维数、强耦合的特点,提出了一种基于数据站场技术的效能评估决策支持方法。通过数据恢复、约减、求精与重构操作,对评估数据进行了处理,并搭建了数据站场立方体,降低了数据间的依赖性,满足了不同层面系统分析的需求。以面向任务的网络化火控系统的为例,进行了系统效能评估、时效性评估和指标分析,从三个层次验证了方法的有效性与可行性。将数据挖掘技术应用到仿真系统效能评估决策中,获得了更广泛的应用。     

Counterinsurgency American style: Considering David Petraeus and twenty-first century irregular war

This article examines the complex legacy of David Petraeus who was a key figure in the emergence of the US military shift towards counterinsurgency doctrine in the years after 2006. Although Petraeus has been perceived by critics as a publicity seeker, he can be credited with laying the foundations for a more serious commitment to COIN involving in particular in integrating conventional and Special Forces in arenas like village stability operations. The article looks a Petraeus's role in both Iraq and Afghanistan: it concludes that, in the case of Afghanistan, it is too early to assess whether counterinsurgency has had a decisive impact of the outcome of the war against the Taliban.     

基于OMAPl510单兵通信系统中头盔子系统的实现

OMAPl510是德州仪器TI公司的双核架构处理器,整合了一个控制性能强的ARM925处理器、一个低功率消耗的高效能TMS320C55X数位信号处理器DSP。文中首先描述单兵通信系统的组成,并给出以头盔子系统为核心的基本工作原理图。然后介绍了0MAPl510的硬件架构和软件架构,并较为详尽地给出实现头盔子系统的蓝牙主控程序的主要代码。最后针对此头盔子系统蓝牙无线通信距离的局限性,提出与头盔共形的微带阵列天线的改进方案。