Perl-speaks-NONMEM (PsN)--a Perl module for NONMEM related programming

The NONMEM program is the most widely used nonlinear regression software in population pharmacokinetic/pharmacodynamic (PK/PD) analyses. In this article we describe a programming library, Perl-speaks-NONMEM (PsN), intended for programmers that aim at using the computational capability of NONMEM in external applications. The library is object oriented and written in the programming language Perl. The classes of the library are built around NONMEM's data, model and output files. The specification of the NONMEM model is easily set or changed through the model and data file classes while the output from a model fit is accessed through the output file class. The classes have methods that help the programmer perform common repetitive tasks, e.g. summarising the output from a NONMEM run, setting the initial estimates of a model based on a previous run or truncating values over a certain threshold in the data file. PsN creates a basis for the development of high-level software using NONMEM as the regression tool.     

支持COC的低层高性能通信库

COC(aClusterofClusters)是指由多个集群组成的大集群系统,COC中的各个集群可能是用不同网络协议连接的。设计基于COC的高性能通信库对于提高并行应用系统的可移植性等具有重要意义。AN是在低层实现的支持COC通信的高性能通信库系统,它支持以太网,Infiniband和FC通信协议,用AN开发的应用程序在这些网络环境下不用重新编译就能运行。由于AN的设计并没有牺牲各种网络的硬件特性,所以能实现高吞吐量低延迟的通信。在AN的基础上可以开发全局操作系统和MPI通信库等各种软件。测试表明,AN是一个适合集群计算的高性能通信库系统。     

NONMEMory: a run management tool for NONMEM

NONMEM is an extremely powerful tool for nonlinear mixed-effect modelling and simulation of pharmacokinetic and pharmacodynamic data. However, it is a console-based application whose output does not lend itself to rapid interpretation or efficient management. NONMEMory has been created to be a comprehensive project manager for NONMEM, providing detailed summary, comparison and overview of the runs comprising a given project, including the display of output data, simple post-run processing, fast diagnostic plots and run output management, complementary to other available modelling aids. Analysis time ought not to be spent on trivial tasks, and NONMEMory's role is to eliminate these as far as possible by increasing the efficiency of the modelling process. NONMEMory is freely available from http://www.uct.ac.za/depts/pha/nonmemory.php.     

A scientific workflow environment for Earth system related studies

Many separate tasks must be performed to configure, run, and analyze Earth system modeling applications. This work is motivated by the complexities of running a large modeling system on a high performance network and the need to reduce those complexities, particularly for the average user. Scientific workflow systems can be used to simplify these task and their relationships, although how to implement such systems is still an open research area. In this paper, we present a methodology to combine a scientific workflow and modeling framework approach to create a standardized work environment and provide a first example of a self-describing Earth system model. We then show the results of an example workflow that is based on the proposed methodology. The example workflow allows running and analyzing a global circulation model on both a grid computing environment and a cluster system, with meaningful abstractions for the model and computing environment. As can be seen through this example, a layered approach to collecting provenance and metadata information has the added benefit of documenting a run in far greater detail than before. This approach facilitates exploration of runs and leads to possible reproducibility.     

Run Placement Policies for Concurrent Mergesorts Using Parallel Prefetching

We study the performance of various run placement policies on disks for the merge phase of concurrent mergesorts using parallel prefetching. The initial sorted runs (input) of a merge and its final sorted run (output) are stored on multiple disks but each run resides only on a single disk. In this paper, we examine through detailed simulations three different run placement policies and the impact of buffer thrashing. The results show that, with buffer thrashing avoidance, the best performance can be achieved by a run placement policy that uses a proper subset of the disks dedicated for writing the output runs while the rest of the disks are used for prefetching the input runs in parallel. However, the proper number of write disks is workload dependent, and if not carefully chosen, it can adversely affect the system performance. In practice, a reasonably good performance can be achieved by a run placement policy that does not place the output run of a merge on any of the disks that store its own input runs but allows the output run to share the same disk with some of the input runs of other merges.     

A Windows computer application for long term data acquisition, storage and presentation of biological signals

We developed a windows application dedicated to recording, displaying and storing large amounts of biological signals on a standard PC. The application works in Matlab and uses an A/D-converter interface that enables a user to run the same software for signal acquisition, display and processing with different A/D-converter boards. The software program consists of three integrated modules, for setting the protocol for signal acquisition; for commands that can be accessed during data acquisition and for software routines that enable off-line viewing, pre-processing and processing of recorded data. It adapts to different computer hardware configurations by enabling the user to define file size, data resolution and file storage on local and network hard disks during signal acquisition.     

一种基于分簇的分布式无线传感器网络拓扑控制算法研究

拓扑控制是无线传感器网络的一个重要研究方向。无线传感器网络中一般节点数量大,分布范围广泛且不规则,难以进行集中式控制。本文提出了一种基于分簇的分布式无线传感器网络拓扑控制(CDTC)算法。利用分簇思想将网络划分为可重叠的簇,簇内各节点按照局部最小生成树算法思想确定邻居关系,调整发送功率,生成合适的网络拓扑。仿真实验证明运行CDTC算法后,网络中节点平均发送功率明显减少,平均节点度较低,节点间干扰较少。     

Simdist: a distribution system for easy parallelization of evolutionary computation

This article introduces Simdist, a software tool for parallel execution of evolutionary algorithms (EAs) in a master-slave configuration on cluster architectures. Clusters have become a cost-effective parallel solution, and the potential computational capabilities are phenomenal. However, the transition from traditional R&D on a personal computer to parallel development and deployment can be a major step. Simdist simplifies this transition considerably, by separating the task of distributing data across the cluster network from the actual EA-related processing performed on the master and slave nodes. Simdist is constructed in the vein of traditional Unix command line tools; it runs in a separate process and communicates with EA child processes via standard input and output. As a result, Simdist is oblivious to the programming language(s) used in the EA, and the EA is similarly oblivious to the internals of Simdist.     

一种节能的无线Mesh网络分簇路由协议

无线Mesh网络部分节点由电池供电,当大规模应用于无线接入时,节能是一件很有意义的工作。本文提出一种节能的无线Mesh网络分簇路由协议(ECRPWMN),对网络先分簇,簇间运行表驱动路由;簇内节点承担路由工作的多少决定于节点剩余能量,从而达到能量消耗均衡。仿真结果表明,ECRPWMN协议可有效节约能量,平衡网络负载,延长网络生存时间。     

多模态粒子群集成神经网络

提出一种基于多模态粒子群算法的神经网络集成方法,在网络训练每个迭代周期内利用改进的快速聚类算法在权值搜索空间上动态地把搜索粒子分为若干类,求得每一类的最优粒子,然后计算最优个体两两之间的输出空间相异度,合并相异度过低的两类粒子,最终形成不但权值空间相异、而且输出空间也相异的若干类粒子,每类粒子负责一个成员网络权值的搜索,其中最优粒子对应于一个成员网络,所有类的最优粒子组成神经网络集成,成员网络的个数是由算法自动确定的.算法控制网络多样性的方法更直接、更有效.与负相关神经网络集成、bagging和boosting方法比较,实验结果表明,此算法较好地提高了神经网络集成的泛化能力.     

Perl语言在群机系统中的应用

Perl语言是实用性很强的一门程序设计语言。本文介绍了Perl语言的特点、功能，以及Perl语言的效率问题。另外，还讨论了该语言在群机系统领域，特别是系统管理软件中的重要应用。     

考虑知识溢出下产业集群创新网络协同知识创造决策分析

协同知识创造对于企业维持和获取竞争优势具有重要作用,但是知识外溢会对企业的竞争力造成威胁,从而使企业在创新网络协同知识创造决策时陷入两难困境.鉴于此,通过构造由一个核心企业和n个从属企业组成的创新网络在协同知识创造过程中知识投入决策Stackelberg博弈模型,引入知识内溢系数、私有知识、知识外溢损失等变量,探讨知识溢出下,网络成员的知识投入决策和知识产出分配问题.研究结果表明:核心企业与从属企业的知识投入比和知识投入总量与协同知识创造绩效和知识内溢收益之和有关,而与私有知识外溢的损失无关;集群创新网络协同知识创造机制得以运行的条件是核心企业的知识创造收益(即知识创造绩效与知识内溢收益之和)足够大;核心企业的知识投入比与其边际收益和知识内溢系数正相关,与从属企业的边际收益和知识内溢系数负相关;协同知识创造过程中从属企业会根据其边际收益和知识内溢系数按比例分配知识的投入比.     

Perl扩展到JVM的移植

Perl扩展作为Perl语言与其他语言的接口,使得Perl能够解决更加复杂的问题。该文从Perl扩展的创建入手,深入分析了Perl解释器的动态链接技术以及Perl扩展的工作原理。然后分析了Perl扩展到Java虚拟机移植的可行性和难点,最后给出了移植的几个实现方案及其评估。     

Formal component-based modeling and synthesis for PLC systems

Programmable Logic Controllers (PLCs) are widely used in industry. PLC systems are reactive systems which run cyclically. In each cycle, the system state is checked and the program is executed once to determine the system behavior for a single cycle. Development of PLC systems conventionally follows the V-model, but increasing demand for efficiency and reliability requires a new rigorous and rapid design flow. In this paper, we propose a component-based formal modeling and synthesis method for cyclic execution platforms and apply it to PLC. Our method consists of three main phases: modeling, verification and code synthesis. In the modeling phase, the BIP (Behavior–Interaction–Priority) framework which is flexible and expressive is used as the modeling language. Real-time behavior, which is intensely concerned in PLC systems, can be modeled as well. In the verification phase, the system model is translated to timed automata and checked by Uppaal. Verification helps to ensure correctness of the model and further increases reliability of the implementation. In the code synthesis phase, the software part of the system model is extracted and synthesized to cyclic code. Although the PLC software runs cyclically, the software model is not necessarily given in a cyclic manner. We propose an algorithm which can generate high-performance cyclic code from a model which describes the business work-flow. This feature significantly simplifies program development. A set of tools is implemented to support our design flow and they are applied to an industrial case study for a PLC system that controls dozens of physical devices in a huge palace.     

集群Web服务器预取机制中用户会话聚类的研究与实现

本文研究了使用集群环境下的用户访问日志数据生成用户会话聚类的方法：编制Perl脚本从用户访问日志中生成用户会话,以新的相似度度量取代欧几里德距离改进Leader算 法对用户会话集合进行聚类,并计算聚类的内部距离和间隔距离来验证算法的有效性。实验结果表明,这种实现能有效地对用户访问日志进行聚类,并能满足服务器预取机制
制在线分析的时间、空间要求。