比例最小偏度单形UKF的RBF神经网络算法

为提高RBF神经网络的建模性能,提出一种基于改进无迹Kalman滤波(UKF)的RBF神经网络训练算法。在该算法中,首先将比例最小偏度单形Sigma点采样策略引入UT,以有效改进UKF,提升其计算效率,然后利用改进的UKF优化估计RBF神经网络的最优参数。仿真结果表明,改进的UKF比EKF具有更高的RBF神经网络模型训练精度,与传统UKF的模型精度大体相当,但速度更快,计算效率更高。     

大中亚区域禁毒警务合作初探——以“金新月”毒品问题国际化为视角

面对“金新月”毒品问题日益国际化,在中亚各国复杂多变的政治、经济、民族、宗教背景下,迫切需要各方共同努力,创设、发展出一种更加灵活、高效、有针对性的禁毒合作机制,以应对该地区越来越严峻的毒品扩散问题及其引发的相关社会秩序危害。阐述了“金新月”毒情发展及其对周边国家的影响,分析了大中亚区域禁毒警务合作的必要性,探讨了大中亚禁毒警务合作模式的选择,以创设出一种立体的、全方位的、有较强执行力的区域禁毒合作模式。     

索契冬奥会安保措施探析

奥运会是世界上规模和影响力最大的体育赛事,因其具有参与人员多、受关注程度高等特点,极易成为恐怖袭击的目标.分析索契冬奥会面临的主要安全威胁和成功举办的安保经验,探讨我国举办冬奥会等大型活动的安保措施。     

基于“当前”统计模型的机动目标自适应跟踪算法

针对"当前"统计模型对加速度极值及机动频率需人为设定的缺陷,提出了一种基于滤波残差的高斯型隶属度函数改进算法,该法在跟踪过程中对加速度极值进行自适应调整。引入一种基于残差的Sigmoid二次型隶属度函数,通过自适应改变机动频率,使模型趋于目标真实状态。理论分析和仿真结果表明,改进的模型避免了人为引入误差,提高了机动目标的跟踪精度,算法简单,易于工程实现。     

A note on the forecast performance of temporal aggregation

Earlier research on the effects of nonoverlapping temporal aggregation on demand forecasting showed the benefits associated with such an approach under a stationary AR(1) or MA(1) processes for decision making conducted at the disaggregate level. The first objective of this note is to extend those important results by considering a more general underlying demand process. The second objective is to assess the conditions under which aggregation may be a preferable approach for improving decision making at the aggregate level as well. We confirm the validity of previous results under more general conditions, and we show the increased benefit resulting from forecasting by temporal aggregation at lower frequency time units. © 2014 Wiley Periodicals, Inc. Naval Research Logistics 61: 489–500, 2014     

Critics gone wild: Counterinsurgency as the root of all evil

The Western interventions in Iraq and Afghanistan have produced a heated polemic concerning the merits and demerits of counterinsurgency – the operational approach underpinning both campaigns. The two books reviewed here provide a good summation of the arguments against counterinsurgency: it is not a strategy and will fail when mistaken as such; its theory does not make intervention and war significantly easier; and even the most successful counterinsurgency campaigns have been bloody, violent, and protracted. Yet as this review highlights, beyond these central points, criticism of counterinsurgency is too often off the mark in its approach and totalizing in its pretentions. There is much to criticize and an urgent need to learn from past campaigns, yet bold claims and broad generalizations can mislead rather than enlighten. The analysis is particularly unhelpful when the definition of the central issue at hand – counterinsurgency – is being unwittingly or deliberately distorted. In the end, these two books form a poor basis for the debate that must now take place, because they are too ideological in tone, too undisciplined in approach, and therefore too unqualified in what they finally say.     

应用径向基代理模型实现序列自适应再采样优化策略

针对径向基插值代理模型样本点预测误差为零时无法获得误差函数进行序列再采样优化的问题,将样本点分布约束引入序列再采样过程,利用潜在最优解加速收敛性,提出一种适用于径向基插值代理模型序列优化的再采样策略,该策略兼顾仿真模型的输出响应特性与样本点的空间分布特性。仿真结果表明,使用该再采样策略后,算法寻优效率和精度均优于传统基于代理模型的优化方法,在对最优解进行有效预测的同时,能显著减少原始模型计算次数。     

用紧支径向基函数传递固体域耦合数据的方法

针对热-电场耦合仿真计算中非匹配网格之间的数据插值问题,提出了一种基于紧支径向基函数点插值的数据传递方法。推导了数据传递矩阵,根据所提出的数据传递方法编制了非匹配网格之间数据传递计算程序,以若干组非匹配网格之间的温度插值为例进行了验证,并分析了不同参数对计算时间及精度的影响,结果表明,算法效率较高,计算误差较小,适用于以热-电场耦合为代表的固体域耦合非匹配网格间的数据传递计算。     

基于改进C-均值聚类算法的空中目标分类

空中目标的正确分类对防空作战效果有着重要的影响。在分析了空中目标类型和目标的表征方法之后,构建了动态聚类中的一种分类算法———改进C-均值算法。该算法改变了以前算法中只用一个聚类中心作为一类的代表的做法,而采用的是类核函数代替类心的方法。类核函数较一个聚类中心更能够表现出某一类的分布信息,从而提高分类的正确率。     

反函数的导数定理的一个注记

给出反函数的导数定理的改进形式:若f(x),x∈(a,b)与φ(y),y(A,B)互为反函数,x0∈(a,b),y0=f(x0),φ(y)在点y0处可导且φ′(y)≠0,f(x)在点x0处连续,则f(x)在点x0处可导,且f′(x0)=1/φ′(y0).并说明,f(x)在点x0处连续这一条件不可去掉。