基于贝叶斯网络的装甲目标战术企图推理模型构建

以贝叶斯网络理论为基础,分析基于贝叶斯网络的目标企图推理框架,从目标状态、目标性质、战场环境和战术特性4个方面提取节点,在分析这4个节点拓扑结构的基础上,构建了装甲目标战术企图推理的贝叶斯网络拓扑结构。     

KL散度多模块滑动窗口慢特征分析的故障诊断方法

针对传统分块方法根据经验划分子块导致变量特征信息无法充分利用,其单一的建模方式忽略局部信息以及离线模型无法适应时变特性的问题,提出了一种KL (Kullback-Leibler)散度多模块滑动窗口慢特征分析方法。在正常工况数据集中,利用KL散度来度量变量间的距离,同时引入最小误差平方和准则进行聚类,分成两个距离最小的子模块;在此基础上利用慢特征分析方法对每个子模块进行建模,结合滑动窗口对每次采样的数据进行更新,得到最优模型,分别计算监测统计信息,利用支持向量数据描述对故障监测结果进行融合,实现故障诊断。并将该方法应用于田纳西伊斯曼过程的监控中,得到了较高的故障检测率和较低的虚警率,验证了该方法的可行性和有效性。     

基于Huber权函数的Allan方差非相关差分抗差估计方法

原子钟频率稳定度计算时由于钟差观测异常会导致Allan方差出现较大的估计偏差。本文首先建立了Allan方差的差分估计模型,根据钟差的差分序列统计特性分析了相关差分估计的噪声识别与白化等计算复杂度问题;针对上述难点提出了一种基于Huber权函数的非相关差分抗差估计方法,建立了一种非相关差分序列的构造方法,有效避免了复杂的噪声识别及白化计算,给出了Allan方差的非相关差分抗差估计的推导,并对抗差过程引入的误差累积给出了一种抵消方法;最后给出了完整的抗差估计方案,并利用实测数据进行了实验验证,结果表明本文方法对相位单点跳变、相位阶跃跳变具有显著的抗差能力,抗差估计可使异常引入的相对偏差由近200%降至5%以内。     

基于稀疏贝叶斯学习的扩展目标雷达关联成像*

雷达关联成像不依赖于雷达与目标的相对运动,是一种高分辨凝视成像方式。传统的关联成像方法未考虑复杂扩展目标的结构信息,在高分辨成像时的应用受到限制,为此提出一种自适应结构配对稀疏贝叶斯学习方法。该算法在稀疏贝叶斯学习的框架内针对扩展目标建立一种结构配对层次化高斯先验模型,然后采用变分贝叶斯期望-最大化算法交替进行目标重构和参数优化。该算法将某一信号分量的重构与周围信号分量联系起来,并能在迭代过程中自适应地调整表征各信号分量相关性的参数。实验结果表明,该方法针对扩展目标可以有效地进行高分辨成像。     

多核贝叶斯优化的模型决策树算法

构造模型决策树时超参数较多,参数组合复杂,利用网格搜索等调参方法将会消耗大量的时间,影响模型性能的提升。提出了一种多核贝叶斯优化的模型决策树算法,该算法为应对不同分类数据特性,采用三种高斯过程建模寻优,利用贝叶斯优化技术,选出最优的参数组合。实验结果表明,所提算法在参数寻优上要优于传统的模型决策树寻优方法,并且能够在迭代次数不多的情况下找到全局最优参数值,在一定程度上提升了算法的分类性能,节省了大量的调参时间。     

EUROPEAN DEFENCE FIRMS: THE INFORMATION BARRIER ON PRIVATE FINANCE*

After the fall of the Berlin Wall, European governments adopted a hands‐off policy towards the defence industrial base, in an attempt to increase the sector’s efficiency and reactivity. In this context, one topical issue is how to motivate defence firms to apply for private rather than public finance. Since banks have no prior experience with European defence firms, a problem of asymmetric information may block this transition. The problem is analysed within the framework of a game between defence firms and banks. It is shown that the Bayesian Equilibrium might correspond to a situation where low‐risk firms prefer the state‐financed scheme; yet, in a perfect information set‐up, the same firms would apply for bank credit. In order to facilitate the transition to private finance, the government might decide to subsidize investors who agree on financing defence firms; the state aid should be made available during a transitory learning period.     

The effect of deployment on first‐ and second‐term re‐enlistment in the US active duty force

Why should deployment affect re‐enlistment? In our model, members enter the military with naïve beliefs about deployment and use actual deployment experience to update their beliefs and revise their expected utility of re‐enlisting. Empirically, re‐enlistment is related to the type and number of deployments, consistent with the learning model. Non‐hostile deployment increases first‐term re‐enlistment but hostile deployment has little effect except for the Army, where the effect is positive. Both types increase second‐term re‐enlistment. Interestingly, first‐term members with dependants tend to respond to deployment like second‐term members. In addition, deployment acts directly to affect re‐enlistment, not indirectly through time to promotion.     

基于压缩估计的单星观测定轨方法研究

分析了单星观测模式下的天基测控系统的可行性,并针对该观测模式下轨道确定中法矩阵的特点,提出了一种基于压缩估计的定轨方法,该方法对法矩阵进行变换,避免了法矩阵奇异造成的误差传递。证明了当满足一定条件时,该压缩方法的估计精度要高于传统的定轨方法。并结合单星观测的特殊性,提出了误差传递因子,设计了单星观测下的压缩估计定轨算法。最后以单星模式下的天基测控系统作为仿真背景进行了仿真试验。结果表明,该压缩估计可有效提高单星观测模式下的轨道确定精度。     

Waypoint analysis for command and control

Command and Control (C2) in a military setting can be epitomized in battles‐of‐old when commanders would seek high ground to gain superior spatial‐temporal information; from this vantage point, decisions were made and relayed to units in the field. Although the fundamentals remain, technology has changed the practice of C2; for example, enemy units may be observed remotely, with instruments of varying positional accuracy. A basic problem in C2 is the ability to track an enemy object in the battlespace and to forecast its future position; the (extended) Kalman filter provides a straightforward solution. The problem changes fundamentally if one assumes that the moving object is headed for an (unknown) location, or waypoint. This article is concerned with the new problem of estimation of such a waypoint, for which we use Bayesian statistical prediction. The computational burden is greater than an ad hoc regression‐based estimate, which we also develop, but the Bayesian approach has a big advantage in that it yields both a predictor and a measure of its variability. © 2004 Wiley Periodicals, Inc. Naval Research Logistics, 2004     

装备研制中的Bayesian网及其应用

针对装备研制过程中产生的大量试验和调试数据,提出采用Bayesian网挖掘各组成单元间的依赖关系,并对Bayesian网学习中基于信息论的方法进行了改进,使确定网络拓扑结构的过程更加客观.学习得到Bayesian网后,分析了其在失效源判定和发现设计缺陷等方面的应用.