中国科学院软件研究所机构知识库
Advanced  
ISCAS OpenIR  > 中科院软件所  > 中科院软件所
题名:
一种基于组织实体能力的软件过程Agent 模型研究
作者: 肖俊超
答辩日期: 2007-03-15
授予单位: 中国科学院软件研究所
授予地点: 软件研究所
学位: 博士
关键词: 组织实体能力 ; 过程Agent ; 过程Agent模型 ; 过程建模 ; 知识结构 ; 行为引擎
其他题名: Research on Organization-Entity Capability Based Software Process-Agent Model
摘要: 自60年代出现软件危机以来,世界各国政府、计算机软件研究机构和组织在软件工程化方法、技术和工具的研究、开发和实践方面投入了大量的人力、物力和资金。人们认识到,要高效率、高质量和低成本地开发软件,必须以改善软件生产过程为中心,实施过程指导的软件生产与质量管理。 软件过程建模方法和技术能够帮助软件组织实现以过程为中心的软件生产,对于软件组织保证软件产品质量,提高开发效率具有重要的理论和实践价值。它能够从根本上改变过去手工作坊式的软件生产,使软件开发形成真正的工业化产业。 软件开发过程是一种特殊的生产过程,它高度依赖于人的能力,同样的过程因执行者不同,会生产不同质量和数量的产品;同时,对于实现同样的目标,不同的执行者可能会采用不同的过程完成。现有的软件过程建模方法基于传统制造业过程建模方法的思想,它们在建立过程时仅仅关注过程和活动间的关系,忽视了人的能力对软件生产过程的影响,在过程中缺乏有效的资源描述和资源调度,也无法适应软件过程的动态多变特性。最终导致建立的过程缺乏良好的可预见性和稳定性。 针对这些问题,本文给出一种基于组织实体能力的软件过程建模方法,该方法将一个具有确定资源能力的组织实体,如它们具有确定的目标、技能、过程知识、经验等,定义为一个过程Agent(过程主体),过程Agent能够在目标驱动下主动自治地建立实现目标的项目软件过程,并在过程中分配合适的资源。过程Agent是建模方法的核心,它的定义应该使其具有建模所需要的各种知识以及主动自治的行为机制。本文围绕过程Agent模型展开研究。 本文首先提出一种基于组织实体能力的软件过程Agent模型,它包括知识结构和行为引擎。其中,知识结构是对过程Agent的特征、各种过程要素知识和经验知识的描述,它是过程Agent能力的基础;行为引擎是过程Agent行为机制的载体,能够基于知识实现目标驱动的自治建模行为。通过该模型定义的过程Agent 在建立软件过程时充分考虑其完成目标的能力,使建立的过程具有良好的可预见性,具备过程稳定执行的前提。 在过程Agent模型的基础上,本文接下来提出一种过程Agent模型的知识结构构成及组织方法,将知识结构分为描述性知识、过程性知识和经验库三部分。在描述性知识中,本文给出过程Agent的目标、技能、拥有人力资源状态的描述,用于说明过程Agent能够做什么;在过程性知识中,本文给出过程的描述方法和组织方法,用于说明过程Agent如何实现它的目标;在经验库中,本文给出过程Agent经验数据的定义以及经验数据的组织方法,这些经验数据可以用来估算过程Agent在实现目标时需要多少资源。通过上述知识的构成和组织,知识结构各部分知识的作用明确、容易构建、能够独立演化、并对过程Agent行为推理提供有效的支持。 然后,本文提出一种过程Agent模型的行为引擎结构并给出行为机制的实现,将过程Agent的行为引擎分为感知器、效应器、推理引擎、协商引擎、学习引擎和实施引擎等六个功能模块,以及信念、愿望和意图三个辅助状态集合。通过对行为引擎机制的介绍,说明过程Agent如何对给定的目标进行分析、协商、并最终为实现目标生成反映组织实体能力的、具有良好资源调度的、执行情况稳定和可预测的项目软件过程。 最后,在应用实例研究中,对本文提出的过程Agent模型有效性进行验证。该模型能够对过程Agent提供良好的定义,保证了基于组织实体能力软件过程建模方法的实现。
英文摘要: Since software crisis occurred in the 1960s, many governments, software research institutes and organizations have invested the massive manpower, resource and fund in the research, development and practice in areas such as software engineering methods, techniques and tools. It is increasingly realized that software development of high productivity, quality and low cost should focus on the improvement of software production process and should implement the process-centered software production and quality management. The software process modeling methods and techniques help enable the software organization to realize the process-centered software production, which is theoretically and practically valuable for software organizations to guarantee software product quality and enhance the development productivity. It has fundamentally changed the past manual workshop-like software production and has gradually made software development a real industry. Software development process is a special production process which is highly dependent on human capabilities. On the one hand, the same process will yield products of different quality and quantity if executors differ; on the other hand, different executors very likely adopt different processes in production when realizing the same goal. The existing software process modeling method is based on the thoughts of traditional manufacturing industry modeling method. The former merely considers the relationship between process and related activities in process modeling and ignores the powerful influence of human capabilities on software production process, lacks resource description and allocation, can not adapt to the dynamic changeable characteristics in software process. As a result, the constructed process lacks effective predictability and stability. Aiming at the above problems, we offer an Organization-Entity Capability based Software Process Modeling method. In this method, we define the Organization-Entity which has the certain resource abilities (such as definite goals, techniques, process knowledge, experiences, etc) as a Process-Agent. The Process-Agent can, driven by goals, actively and autonomously construct project software process to realize goals, and also appropriately allocate resources in the process. The Process-Agent is the core of modeling method. Its definition is supposed to include all kinds of knowledge that modeling needs and also the active, autonomous behavior mechanism. This paper mainly discusses about the research on Process-Agent model. Firstly, this paper introduces an Organization-Entity capability based software Process-Agent model. The model includes Infrastructure and Engine. In the model, Infrastructure functions as the basis of Process-Agent’s capability, and it describes the features, experience and various process elements of the Process-Agent; Engine is the behavior mechanism of Process-Agent, it can realize the goal-driven autonomous modeling behavior on the basis of Infrastructure. The Process-Agent defined under such a model can sufficiently consider its capability of goal realization in software process construction. Consequently, the execution of the constructed process will be well predictable and stable. Secondly, on the basis of Process-Agent model, the paper offers knowledge constitution and organization of Infrastructure in Process-Agent model. Infrastructure includes Descriptive Knowledge, Process Knowledge and Experience Library. Among them, the Descriptive Knowledge includes the depiction of Process-Agent's goals, skills and depiction of human resource states, and it aims to explain what Process-Agent can do; the Process Knowledge tells how to describe and organize the process, and it explains how Process-Agent realize its goals; the Experience Library introduces the definition and organization of experience data in Process-Agent, and these data are used to estimate how many resources Process-Agent needs to realize its goals. With the above knowledge constitution and organization, the knowledge of various parts in Infrastructure is explicit in function and easy to construct. It can also evolve independently and provides effective support for Process-Agent behavior reasoning. Then, the paper presents an Engine structure of Process-Agent model and also gives the realization of behavior mechanism. The Engine of Process-Agent is composed of six function modules, that is, Perceptor, Reactor, Reasoning Engine, Negotiation Engine, Learning Engine and Enactment Engine, and it also includes three auxiliary state sets, that is, Belief Set, Desire Set and Intention Set. The Engine mechanism illustrates how Process-Agent analyzes and negotiates the given goals and further generates project software process that reflects Organization-Entity capabilities, allocates resources effectively and executes both stably and predictably. Finally, in case study, the paper validates the effectiveness of Process-Agent model. The result is that the model is able to make a well definition of Process-Agent and thus guarantees the realization of Organization-Entity Capability based Software Process Modeling method.
语种: 中文
内容类型: 学位论文
URI标识: http://ir.iscas.ac.cn/handle/311060/6840
Appears in Collections:中科院软件所

Files in This Item:
File Name/ File Size Content Type Version Access License
10001_200318015003103肖俊超_paper.pdf(1555KB)----限制开放-- 联系获取全文

Recommended Citation:
肖俊超. 一种基于组织实体能力的软件过程Agent 模型研究[D]. 软件研究所. 中国科学院软件研究所. 2007-03-15.
Service
Recommend this item
Sava as my favorate item
Show this item's statistics
Export Endnote File
Google Scholar
Similar articles in Google Scholar
[肖俊超]'s Articles
CSDL cross search
Similar articles in CSDL Cross Search
[肖俊超]‘s Articles
Related Copyright Policies
Null
Social Bookmarking
Add to CiteULike Add to Connotea Add to Del.icio.us Add to Digg Add to Reddit
所有评论 (0)
暂无评论
 
评注功能仅针对注册用户开放,请您登录
您对该条目有什么异议,请填写以下表单,管理员会尽快联系您。
内 容:
Email:  *
单位:
验证码:   刷新
您在IR的使用过程中有什么好的想法或者建议可以反馈给我们。
标 题:
 *
内 容:
Email:  *
验证码:   刷新

Items in IR are protected by copyright, with all rights reserved, unless otherwise indicated.

 

 

Valid XHTML 1.0!
Copyright © 2007-2017  中国科学院软件研究所 - Feedback
Powered by CSpace