多级集群式供应链跨链库存合作及鲁棒优化算法

以四级两单链的集群式供应链为研究对象,在两个单链相巨竞合条件下,建立了两单链问两零售商跨链问库存协作的时变鲁棒控制系统模型,以及牛鞭效应的弱化方法,并用H∞优化算法来求解动态最优决策.通过实例分析表明,集群式供应链跨链间协调及鲁棒优化能有效抑制牛鞭效应、降低库存水平,在减小订货波动的同时可提高对集群市场的响应速度.     

基于扩散激活模型的无线自组网分簇方法

传统的分簇方法很少同时考虑安全因素及其对网络性能的影响。针对此问题,提出了一种基于信任关系的分簇方法,该分簇方法结合人类记忆的扩散激发模型思想,能够根据有限的局部信息自动地对整个网络进行分割。试验结果表明,该文所提出的分簇方法在精确度方面与集中式的分簇方法非常接近。因此,在提高Ad Hoc网络性能的同时,还可提高其安全性。     

移动Sinks无线传感器网络的能量效率分析

能量效率是无线传感器网络的关键性问题,无线传感器网络通过移动sinks节点延长了网络的生命周期。讨论一些关键性要素对能量效率的影响,如簇的大小、移动sinks的速度、数量和数据包生命周期等。分析基于能量效率的单跳、多跳路由模式,提出最优跳数的簇路由模式,并以仿真实验验证移动sinks无线传感器网络的系统性能。     

一种伪随机自治服务可生存调度算法

可生存调度算法是服务可生存性研究中的重要内容,在服务器集群系统中引入随机策略,使服务在各个服务器之间的轮转产生不确定效果,可增加攻击者攻击系统的难度,增强系统可生存性。介绍可生存性研究的层次划分,指出可生存调度算法在可生存性研究中的作用,分析现有可生存调度算法的优缺点,提出一种伪随机自治服务可生存调度算法,并分析算法的性能。     

集群系统中的FP-Growth并行算法

针对FP-Growth算法面临大规模数据库时空效率不高的问题,提出一种面向计算机集群的并行算法。采用投影方法直接寻找频繁项的条件数据库,将挖掘条件数据库的工作分化成若干独立的子任务,分配到集群中的节点上并行实现,由中央节点汇总结果并输出。结果证明,该算法不仅能够提高计算速度,解决数据库规模过大时内存溢出的情况,且具有良好的延展性。     

Using Similarity Analyses to Scale Out Research Findings Across Andean Watershed Basins

Abstract

Strategic research in agriculture and natural resources carried out by international research centers is deemed a public good and should, sooner or later, be put into the hands of development, governmental and non-governmental organizations. However, this research is usually done at specific pilot sites; there is a greater need to know how representative those sites are in relation to the diversity of contexts in other locations. Such is the case with the Challenge Program on Water and Food (CPWF), a global initiative in water research promoted by the Consultative Group on International Agricultural Research (CGIAR), which is developing and implementing strategic research in nine basins located in the tropics of Africa, Asia and South America. Given that resources are not available to collect data from the whole of the region, pilot sites are needed. It is hoped that research outputs obtained in the selected pilot sites can be the basis for scaling out solutions to similar situations in neighbouring or adjacent areas in same or different basins.

In order to contribute to the scaling-out process, different classification methodologies were applied to determine how specific watershed basins are representative of larger areas. The Andean eco-region served as a case study but the methods can easily be applied in other regions. The spatial diversity of biophysical and social conditions across the Andes requires careful site selection. Two methods, a combination of Weight of Evidence (WofE) and Logistic Regression (LR) methods and Fast Cluster analysis, were used to determine the similarity of selected sites with those excluded. A 1-km study resolution covering most of the Andes eco-region included annual rainfall, elevation, length of growing period, land cover, roads and population density as the key variables. Results showed complementarities between the two methods in presenting a probability surface of similarity across the Andes and a clustering of similar sites inside and outside the pilot basins. The output information forms a strong basis for devising plans to scale out research findings from the pilot basins to the whole region.     

基于三角不等式原理的TTSAS聚类加速算法

顺序聚类算法是一种非常直接和快速的算法，并且不需要提前确定聚类个数。但是当处理海量数据时，时间效率仍然有待提高。TTSAS算法是两个阈值的顺序聚类算法，在此基础上，该文应用三角不等式原理提出了TI_TTSAS算法，该算法避免了冗余的距离计算，实验结果证明，相对于TTSAS算法，TI_TTSAS在速度上有很大程度的提高，数据规模越大，改进效果越明显。并且聚类效果保持了TTSAS算法的准确性。     

用Python+GTK开发机群网络通信监控系统

文中介绍了Python语言、GTK图形包、Glade图形界面生成器以及用Python GTK开发GUI程序的方法。同时描述了机群网络通信监控系统的功能和结构，并对开发过程中的几个关键技术进行了详细阐述。     

嵌埋于Ag基质中纳米Co团簇结构及其磁性的研究

基于溅射－气体－聚集（SGA）形成原子团族的原理，提出了蒸发－气体－聚集（EGA）共沉积制备嵌埋团簇新方法，并在方华膜电镜载网和单晶硅片上成功地制备了嵌埋于Ag基质中纳米Co团簇样品。TEM／ED分析表明，样品中Co团族被Ag原子所包裹，有效地阻止了团簇的表面氧化与自聚集，其结构为fcc－Co和fcc－Ag多晶共存形态。VSM磁性测量表明，样品呈现出超顺磁性，随着测量温度的降低，其饱和磁化强度逐渐升高．     

Parallel global optimization with the particle swarm algorithm

Present day engineering optimization problems often impose large computational demands, resulting in long solution times even on a modern high-end processor. To obtain enhanced computational throughput and global search capability, we detail the coarse-grained parallelization of an increasingly popular global search method, the particle swarm optimization (PSO) algorithm. Parallel PSO performance was evaluated using two categories of optimization problems possessing multiple local minima-large-scale analytical test problems with computationally cheap function evaluations and medium-scale biomechanical system identification problems with computationally expensive function evaluations. For load-balanced analytical test problems formulated using 128 design variables, speedup was close to ideal and parallel efficiency above 95% for up to 32 nodes on a Beowulf cluster. In contrast, for load-imbalanced biomechanical system identification problems with 12 design variables, speedup plateaued and parallel efficiency decreased almost linearly with increasing number of nodes. The primary factor affecting parallel performance was the synchronization requirement of the parallel algorithm, which dictated that each iteration must wait for completion of the slowest fitness evaluation. When the analytical problems were solved using a fixed number of swarm iterations, a single population of 128 particles produced a better convergence rate than did multiple independent runs performed using sub-populations (8 runs with 16 particles, 4 runs with 32 particles, or 2 runs with 64 particles). These results suggest that (1) parallel PSO exhibits excellent parallel performance under load-balanced conditions, (2) an asynchronous implementation would be valuable for real-life problems subject to load imbalance, and (3) larger population sizes should be considered when multiple processors are available.