首页> 外文学位 >Model-Based Detection of Emergent Behavior in Distributed and Multi-Agent Systems from a Component Level Perspective.
【24h】

Model-Based Detection of Emergent Behavior in Distributed and Multi-Agent Systems from a Component Level Perspective.

机译:从组件级角度研究基于模型的分布式和多代理系统中紧急行为的检测。

获取原文
获取原文并翻译 | 示例

摘要

Requirement elicitation is one of the most challenging and critical phases of the software development lifecycle. Many faults are introduced into the system as the result of incomplete requirements. An effective approach for the design of software systems is to describe system requirements using scenarios. A scenario, commonly expressed using a message sequence chart or a sequence diagram, is a temporal sequence of messages sent between system components. However despite their simplicity and expressive power, scenario-based specifications are prone to subtle deficiencies with respect to analysis and validation known as incompleteness and partial description. These deficiencies in scenario-based specifications are the prime cause of emergent behavior. Emergent behavior, also known as implied scenarios are behavior that the system exhibits but are not explicitly defined in its requirements. Emergent behavior is an important issue in the design of software systems; particularly ones with the lack of central control such as distributed and multi-agent systems. Detecting and removing emergent behavior during the design phase will lead to huge savings in deployment costs of such systems. In this thesis, a method for detecting emergent behavior in system requirements described using scenario-based specifications is proposed. The use of this methodology for a variety of different software systems such as distributed and multi-agent systems (MAS) is demonstrated. Furthermore this research contains methodologies for verifying the lack of existence of a particular emergent behavior in the software system. These methodologies have been demonstrated using various case studies such as distributed systems for a mine-sweeping robot and an online commerce application and a multi-agent system for a manufacturing system. Furthermore as this research aims to develop these methodologies into a software tool, the requirement and design documents as well as the prototype of this tool are presented in this thesis.
机译:需求激发是软件开发生命周期中最具挑战性和关键性的阶段之一。由于需求不完整,许多故障被引入系统。设计软件系统的有效方法是使用方案描述系统要求。通常使用消息序列图或序列图表示的场景是系统组件之间发送的消息的时间序列。然而,尽管基于场景的规范具有简单性和表达能力,但它们在分析和验证方面仍存在细微的缺陷,称为不完整和部分描述。基于方案的规范中的这些缺陷是紧急行为的主要原因。紧急行为,也称为隐含场景,是系统显示但未在其要求中明确定义的行为。紧急行为是软件系统设计中的重要问题。特别是缺乏集中控制的系统,例如分布式和多代理系统。在设计阶段检测和消除紧急行为将大大节省此类系统的部署成本。本文提出了一种基于场景的规范描述的系统需求中紧急行为的检测方法。演示了此方法在各种不同软件系统(如分布式和多代理系统(MAS))中的使用。此外,该研究还包含一些方法论,以验证软件系统中不存在特定的紧急行为。这些方法论已通过各种案例研究得到了证明,例如用于扫雷机器人的分布式系统和在线商务应用程序以及用于制造系统的多代理系统。此外,由于本研究旨在将这些方法学发展成为一种软件工具,因此本文提出了需求和设计文档以及该工具的原型。

著录项

  • 作者

    Moshirpour, Mohammad.;

  • 作者单位

    University of Calgary (Canada).;

  • 授予单位 University of Calgary (Canada).;
  • 学科 Engineering System Science.;Computer Science.
  • 学位 M.Sc.
  • 年度 2011
  • 页码 96 p.
  • 总页数 96
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号