全语言思路对大学英语教学的启示

全语言思路是国外语言教学界颇有影响的一种教学理念,上世纪被介绍到我国,但是在大学英语教学方面实践较少。笔者通过对国内外有关文献的综述,给读者介绍了这一思路的基本观点和教学原则,并结合自身的体会,对目前争论激烈的大学英语教学改革提出了一些建议。     

军队院校图书馆在OA模式下的发展策略

OA是一种新兴的学术信息交流模式。它促进了学术交流，免费向公众开放学术信息。简要阐述OA的发展及其现状，分析OA运动对军队院校图书馆的意义，并探讨了OA环境中军队院校图书馆的发展策略。     

基于反步法的高超声速飞行器终端滑模控制

为了使高超声速飞行器在再入过程中完成姿态控制,在反步法的基础上将模糊自适应和终端滑模控制相结合设计控制器。对于姿态角动态,使用反步法和模糊自适应设计虚拟控制律;而对于角速率动态,利用终端滑模方法设计控制器,能够有效提高系统鲁棒性。稳定性分析说明所提方法可以保证系统状态的一致渐进有界性。     

利用状态机检测JAVA空指针引用故障

介绍了一种在控制流的基础上利用故障状态机对JAVA空指针故障进行检测的方法。该方法利用变量区间来表示状态的前提条件,对程序中不可达路径进行处理,从而达到减少误报的目的。     

有向无环图决策支持向量机和经验模式分解在轴承故障诊断中的应用

当旋转机械发生故障时,其振动信号常常表现出较为复杂的调制形式,经验模式分解能根据信号的真实物理意义完成自适应分解。支持向量机由于其出色的学习性能和良好的推广能力,使其在包括故障诊断在内的众多领域得到较为广泛的应用。利用经验模式分解结果提取频带能量特征向量,采用有向无环图决策支持向量机实现对轴承状态的判别,并基于留一法优化支持向量机的模型参数。最终的应用结果表明,基于EMD和有向无环图决策支持向量机方法可以有效实现对轴承的状态判别。     

The Anglo-American military relationship: Institutional rules,practices, and narratives

The Anglo-American military relationship is a vital yet neglected area of study. This article argues that the British military have actively cultivated a relationship with the U.S. military that has contributed to the longevity of the broader so-called “Special Relationship,” even in the Trump era. The article contends that the complexities of the military relationship can best be captured by the theoretical lens provided by Lowndes and Roberts that combines different strands of institutionalism to focus on rules, practices, and narratives. The intense linkages between the United States and United Kingdom have become routinized, enabling them to adapt their peacetime cooperation to conflicts, and thereby address post-Cold War security challenges. The article draws upon semi-structured interviews with senior British military officers as well as policy documents to explore how these patterns of collaboration have become ingrained in patterns of both thinking and behavior.     

Multilateral defense cooperation in the Indo-Asia-Pacific region: Tentative steps toward a regional NATO?

Some U.S. military leaders have asserted that the United States, Japan, Australia, and India and the Republic of Korea are developing multilateral defense cooperation to deter aggression and uphold norms much like North Atlantic Treaty Organization (NATO) has in Europe. Frequent military exercises and China’s threats to freedom of navigation (FoN) and North Korea’s nuclear missiles comprise the motive force for such cooperation. However, cooperation thus far has been trilateral and minimal, given divergent national interests and dispersed geopolitical locations. Cooperation among Japan, Republic of Korea (ROK), and the United States is increasing given the threat, but ROK’s public opinion is divided about Japan. Australia, Japan, and India have increased cooperation with the United States but are reluctant to conduct FoN operations with the United States to challenge China’s expansionism in the South China Sea. If China becomes more aggressive and blocks FoN or seizes territory, development toward an Asian NATO is possible.     

多空域窗射击技术研究

未来空域窗射击体制相对集火射击体制,提高了对机动目标的拦截概率,能够对付有限机动目标,然而依然难以对付具有高度机动能力的目标,尤其是有人驾驶飞机,如果在射弹飞行时间内进行拐弯、爬升、俯冲等强机动.为了进一步提高对目标的拦截概率,提出了多空域窗射击体制,即在传统火控系统中增加目标实时态势估计模块,实时预测目标在射弹飞行时间内的可能机动模型,并根据预测结果进行多个未来空域窗射击.     

面向癫痫脑电的简化深度学习模型

针对脑电信号随机性强、动态变化迅速等特点,提出了一种简化深度学习模型研究癫痫脑电识别问题。提出的模型以一维卷积神经网络为基础,在结构方面简化了卷积层、池化层等以提高模型效率,在整体框架方面应用了Keras框架,在训练优化算法方面采用RMSProp算法作为模型优化算法,通过预定义的目标函数来进行损失估计,模型设计上加入了批标准化层和全局均值池化层。基于所提模型,从三个方面研究了癫痫脑电识别问题,即:利用经验模态分解,分别选取前三阶、前五阶、前七阶、前八阶的本征模态函数分量,在简化模型上进行对比分析；利用提出模型所具备的深度学习特点,直接识别原始脑电信号而无须特征提取环节；增加了三种不同方法分别提取7类特征,对相同的脑电数据进行对比分析。性能分析结果表明:对于五类不同的脑电信号,前三阶的本征模态函数分量的识别率达到92.1%,比其他几种处理方式识别率高；前八阶的本征模态分量识别率不及原始信号,表明人工数据处理时会给数据带来噪声; 所提出的简化深度学习模型能高效处理癫痫脑电识别问题,具备较高效率和较好性能。     

基于EMD和神经网络的气阀机构故障诊断研究

运用经验模态分解(EMD)对柴油机缸盖振动信号进行分析,求得各内禀模态函数(IMF)的能量百分比;将能量百分比作为神经网络的输入进行网络训练和故障识别,实现了气阀机构的故障诊断,取得了较好的效果。