利用混合编程改善SMP机群上并行矩阵乘法的性能

针对SMP机群,探讨了分别利用单机优化、OpenMP与MPI从指令级、共享存储级与分布主存级三个层次上改善矩阵并行乘Fox算法性能的方法。并通过调用数学函数库与混合编程的方式,在深腾6800上进行了实验,取得了相当满意的数值效果。     

Optimization and mediated bartering models for ground delay programs

The Federal Aviation Administration (FAA) and the airline community within the United States have adopted a new paradigm for air traffic flow management, called Collaborative Decision Making (CDM). A principal goal of CDM is shared decision‐making responsibility between the FAA and airlines, so as to increase airline control over decisions that involve economic tradeoffs. So far, CDM has primarily led to enhancements in the implementation of Ground Delay Programs, by changing procedures for allocating slots to airlines and exchanging slots between airlines. In this paper, we discuss how these procedures may be formalized through appropriately defined optimization models. In addition, we describe how inter‐airline slot exchanges may be viewed as a bartering process, in which each “round” of bartering requires the solution of an optimization problem. We compare the resulting optimization problem with the current procedure for exchanging slots and discuss possibilities for increased decision‐making capabilities by the airlines. © 2005 Wiley Periodicals, Inc. Naval Research Logistics, 2006     

面向决策的一体化建模方法及其在兵力筹划中的应用

在对军事决策问题的主要特点进行分析的基础上,提出从军事概念模型到数学模型再到定性推理模型的一体化建模思想.以空军航空兵遂行直接航空火力支援任务为想定背景,根据EATI模板,利用基于UML的Rose工具依次对想定任务的实体、任务、交互等进行分析;对于兵力筹划中的结构化问题建立了机会约束多目标规划数学模型;对于非结构化问题则构造了推理机,推理机体现了定量与定性相结合的决策分析思想.最后,以一个实例验证文章思路的可行性.     

UNAMID: an African solution to a complex case of geopolitical dynamics

This article examines the geopolitical dynamics associated with the African Union (AU) and United Nations hybrid operation in Darfur (UNAMID) from the start of the Darfur conflict in 2003 until the time when UNAMID became fully operational in 2011. It provides an overview of the complex forces and geopolitical dynamics that affected the deployment of UNAMID and shaped its unique hybrid character. It mainly highlights those primary geopolitical factors that hindered the full deployment of UNAMID. It is concluded that this period was a showcase for the newly established AU and its support for its member states, as well as a new approach for the international community to maintain international peace and security in alliance with regional organisations.     

潜艇规避主动声纳浮标自主决策算法研究

主动声纳浮标可以实现对潜艇的精确定位,但易被潜艇感知采取机动规避措施,为了有效减少被主动声纳浮标发现的概率和暴露时间,提出了潜艇自主规避的决策算法,并针对一型主动声纳浮标阵探测进行仿真,分析探测概率变化规律,由此验证算法的有效性。     

The Blind Basinet: A New Interpretation of Royal Armouries IV.2

For many years a curiosity in the collection of the Royal Armouries, IV.2, or the ‘Blind Basinet’, is a fascinating example of an early 16th century great basinet not so much for what is present but for what is not. The lack of any means of seeing out of the helmet has led to a number of traditions regarding its provenance and purpose. A detailed examination of the helmet’s interior, as well as comparison with some other contemporary pieces, shows that it is an unfinished piece of armour. It was left at the stage where the sights and breaths would have been pierced, but they were left unfinished until some later modifications created a display piece, one which can still be seen at the Royal Armouries in Leeds.     

一种面向宽带复杂信号侦察的可配置计算应用框架

可配置计算系统的优势与系统设计过程的复杂性是一对矛盾,可配置计算框架为此提供了有益的思路。但受限于适用范围与通用性考虑,已有的特别是商业框架仅支持系统的静态重构。本文针对宽带信号侦察这一特定应用领域,提出了一种功能级可配置计算应用框架BSRRCS-RCAF,该框架将领域应用中具有共性的计算与功能模块抽取和集中起来,形成二进制比特流配置文件库,并提供构建此类系统的公共服务体系。基于公共服务体系以及配置文件库,当需要修改或者重新构建宽带信号侦察系统时,只需在现有框架下动态调用配置组件库,或者遵循框架的统一设计接口完善补充组件库中的特定功能部分,整个系统设计框架将得到方便复用,便于实现基于快速重构的多种系统功能。     

预防性维修工作组合优化

针对单部件的维修间隔期无法反映系统的维修最佳间隔期这一问题,对装备各部件在各修理级别上的预防性维修任务组合情况进行探索。通过建立费用模型、可用度和风险模型作为决策模型,用以组合某一修理级别各部件的维修工作,得出系统预防性维修间隔期,并结合部队实际优化从系统的角度优化等级维修中的维修工作间隔期,为确定大中小修工作提供科学、合理的理论指导。该算法在区分各修理级别的维修能力的基础上,能较好地对各修理级别的维修工作组合情况进行了优化,实例验证了模型的实用性和有效性。     

GPS接收信号中线性扫频干扰的抑制

研究了阵列接收的GPS接收信号中线性扫频(LFM)干扰的抑制问题。LFM信号瞬时频率的估计是抑制干扰的关键,而要准确估计信号瞬时频率,必须抑制接收信号时频分布中交叉项干扰。本文对接收信号矢量进行白化,得到信号矢量白化后的空时频分布(STFD)矩阵。文中推出一个与传统方法获得的判决变量不同的判决变量,用该判决变量可以更加清晰地选出LFM信号在时频分布中的自项。根据时频脊点的分布估计LFM信号频率参数以及信号瞬时频率,之后即可根据瞬时频率构建陷波器滤除干扰。仿真表明该方法能够将受交叉项严重干扰的接收信号时频分布映射为清晰的接收信号自项的时频分布,在数据快拍数满足一定要求时可以很好地抑制掉LFM干扰。     

基于CPU-GPU混合计算平台的RNA二级结构预测算法并行化研究

RNA二级结构预测是生物信息学领域重要的研究方向,基于最小自由能模型的Zuker算法是目前该领域最典型使用最广泛的算法之一。本文基于CPU GPU的混合计算平台实现了对Zuker算法的并行和加速。根据CPU和GPU计算性能的差异,通过合理的任务分配策略,实现二者之间的并行协作计算和处理单元间的负载平衡;针对CPU和GPU的不同硬件特性,对Zuker算法在CPU和GPU上的实现分别采取了不同的并行优化方法,提高了混合加速系统的计算性能。实验结果表明,CPU处理单元在混合系统中承担了14%以上的计算任务,与传统的多核CPU并行方案相比,采用混合并行加速方法可获得15.93的全局加速比;与最优的单纯GPU加速方案相比,可获得16%的性能提升,并且该混合计算方案可用于对其它生物信息学序列分析应用的并行和加速。