Annealing adaptive search,cross‐entropy,and stochastic approximation in global optimization

The Annealing Adaptive Search (AAS) algorithm for global optimization searches the solution space by sampling from a sequence of Boltzmann distributions. For a class of optimization problems, it has been shown that the complexity of AAS increases at most linearly in the problem dimension. However, despite its desirable property, sampling from a Boltzmann distribution at each iteration of the algorithm remains a practical challenge. Prior work to address this issue has focused on embedding Markov chain‐based sampling techniques within the AAS framework. In this article, based on ideas from the recent Cross‐Entropy method and Model Reference Adaptive Search, we propose an algorithm, called Model‐based Annealing Random Search (MARS), that complements prior work by sampling solutions from a sequence of surrogate distributions that iteratively approximate the target Boltzmann distributions. We establish a novel connection between MARS and the well‐known Stochastic Approximation method. By exploiting this connection, we prove the global convergence of MARS and characterize its asymptotic convergence rate behavior. Our empirical results indicate promising performance of the algorithm in comparison with some of the existing methods. © 2011 Wiley Periodicals, Inc. Naval Research Logistics, 2011     

一种基于相位校正的InISAR图像配准新方法

针对InlSAR三维成像中的图像配准问题进行研究,从理论上推导了同一基线上两干涉天线之间图像失配准的原因,并定量分析了图像失配准量对后续干涉处理的影响.研究表明,初始时刻同一基线上两干涉天线间的波程差引起两幅ISAR图像距离方向的亚像素级失配,而成像积累时间内两干涉天线波程差的变化量则引起两幅ISAR图像多普勒方向的像...     

The Naval Intelligence Underpinnings of Reagan's Maritime Strategy

Washington's so-called Maritime Strategy, which sought to apply US naval might against Soviet vulnerabilities on its maritime flanks, came to full fruition during the 1980s. The strategy, which witnessed a major buildup of US naval forces and aggressive exercising in seas proximate to the USSR, also explicitly targeted Moscow”s strategic missile submarines with the aim of pressuring the Kremlin during crises or the early phases of global war. Relying on a variety of interviews and newly declassified documents, the authors assert that the Maritime Strategy represents one of the rare instances in history when intelligence helped lead a nation to completely revise its concept of military operations.     

Shielding the ‘Hot Gates’: Submarine Warfare and Japanese Naval Strategy in the Cold War and Beyond (1976–2006)

The build-up of Japan's military apparatus in the 1990s and 2000s has been often regarded by security analysts as indicative of a departure from the country's Cold War strategic posture. Japan appears to be engaged in a process of militarisation that is eroding the foundations of its ‘exclusively defence-oriented’ policy. In the case of the archipelago's naval strategy, such assessments overlook the longstanding significance of a core feature of its defence policy, namely the surveillance of maritime crossroads delivering the wealth of the country. The paper reassesses the evolution of the Japanese strategy since the Cold War by examining the development of the Japanese Maritime Self-Defence Force's submarine force, one of the key components of the defensive shield for these crossroads. The paper argues that with the changes in the security environment of the 1990s, Japan already fielded a mature force with state-of-the-art submarines, and that the rise of a new naval competitor aiming at controlling key strategic points along Japan's sea lanes reconfirmed the critical importance of submarine operations to Japanese national security.     

Operation ‘enduring freedom’: a victory for a conventional force fighting an unconventional war

The track record of the US military in unconventional wars has not been good and there were fears that Operation ‘Enduring Freedom’ might suffer the same fate as previous campaigns. This contribution explores why the Taliban were defeated so easily by the US in 2001. It challenges the view that America's victory was due solely to changes in its modus operandi or that the outcome heralds a change in the fortunes of the US when fighting unconventional war. It also questions the idea that America's victory was a consequence of Taliban incompetence. Instead, it explains the defeat of the Taliban in terms of the prevailing political conditions within Afghanistan, which made them vulnerable to attack. The essay concludes that current political circumstances could, in the long run, permit the resurrection of the Taliban and undermine the US-led coalition's victory.     

The Zimbabwe people’s revolutionary army military operations in Makonde district and the attack on Salisbury’s fuel storage tanks, 1965-1979

ABSTRACT

This article analyses the Zimbabwe People’s Revolutionary Army operations in Mashonaland West province with particular reference to Makonde District during the country’s war of liberation. Despite the growing literature on ZAPU and ZPRA in Zimbabwe’s war of liberation there is little that has been written concerning the operations of ZPRA guerrillas in Mashonaland provinces. Guided by evidence drawn from secondary and primary sources the article further asserts that ZPRA adopted a four pronged strategy in Makonde District. This strategy in part was hinged on the politicization of the population on the one hand and annihilation of the enemy forces, attrition and manoeuvre warfare on the other. Again, the article argues that ZPRA guerrillas attempted to capture Salisbury (Harare) in order to conquer the Rhodesian Security Forces’ centre of gravity. It was hoped that the capture of the capital city would Makoni District was of strategic significance to the conduct of ZPRA operations and arguably hastened Ian Smith’s decision to attend the Lancaster House Conference negotiations that led to the independence of Zimbabwe in 1980.     

信息推荐系统中的朋友关系预测算法设计

伴随着互联网规模的不断扩展,信息过载问题越来越突出。信息推荐系统被视为解决信息过载问题的最有效方法。然而目前的方法大多数仅考虑用户独立的反馈,而忽略用户的社会属性对推荐的重要作用,这对信息推荐系统的性能会造成巨大的影响。为此,本文提出了基于朋友关系预测的信息推荐算法,将用户的社会关系预测引入信息推荐过程中,分别基于用户的拓扑信息及历史交互信息建立用户社会关系的存在性判定及关系类型判定,并利用线性回归分析方法和逻辑回归分析方法实现了基本特征的融合。最后,通过在Epinions和Slashdot真实数据集上的实验证明,本方法能够有效提高用户社会关系预测的准确性。     

L1-analysis稀疏重构在阵列信号恢复及波达角估计中的应用

通过适当的空域稀疏化构造了可对阵列接收信号进行冗余稀疏表示的阵列流形矩阵,建立了相应的L1-analysis稀疏重构模型用于恢复阵列接收信号,重点证明了该流形矩阵是满足L1-analysis 稀疏重构条件的紧框架,从理论上保证了将L1-analysis 稀疏重构用于阵列接收信号恢复及波达角估计问题的合理性,并推导出信号恢复误差的理论上界。利用在微波暗室环境中采集的实测数据,结合MUSIC算法进行实验验证,结果表明基于L1-analysis 稀疏重构的信号恢复对提高低信噪比环境下的波达角估计性能是有效的。     

PHEV电池电荷状态估计设计及仿真

电池管理系统(battery management system,BMS)是混合动力汽车(parallel hybrid electric vehicle,PHEV)能量管理系统中的核心组成部分,而其中电池电荷状态(state of charge,soc)则是PHEV控制策略中的重要参数.针对PHEV动力电池组SOC系统高度非线性和复杂性的特点,提出了一种基于改进的BP神经网络的HEV动力电池组的实时SOC估计,并对网络的收敛性进行了证明.利用大量PHEV动力电池组在行驶过程中充放电的数据样本,对神经网络进行网络训练并且进行仿真.结果表明,与传统离线SOC估计方法相比,能够有效地减小误差,提高电池SOC的精度.     

海军舰艇编队防空导弹协同作战技术研究

主要以信息化条件下如何提升海军舰艇编队防空导弹作战能力为背景,重点描述协同指挥和协同制导作战的概念内涵、国内外研究现状、作战模式、主要实现的功能,给出不同指挥层次下各指挥节点的体系结构和功能定位,进而从信息场和功能界面等方面说明协同指挥和协同制导2个功能层次的区别.