| 软件过程的建模、分析和演化技术研究 |
| 其他题名 | Research on Software Process Model ing, Analysis and Evolution Technology
|
| 陈诚
|
| 专业 | 计算机应用技术
|
| 2002
|
| 学位授予单位 | 中国科学院软件研究所
|
| 学位 | 博士
|
| 学位授予地点 | 中国科学院软件研究所
|
| 关键词 | 软件过程
过程建模
过程模型分析
过程模型演化
软件工程环境
|
| 摘要 | 对软件开发和维护过程的管理能够帮助软件项目的控制和协调,提高软件的质量和生产效率.软件过程技术通过描述、运作、监控、度量和演化有针对性的过程模型,对用户实际进行的软件过程提供了指导和控制,是支持软件开发、项目管理、质量保证的关键技术,已经成为软件工程领域的研究重点之一.论文研究了其中软件过程建模、过程模型和动态分析、以及分析和处理过程模型动态演化的影响这几个方面,主要的贡献和创新如下：1.论文提出了一个能够满足不同用户需求的、可扩展的、形式化的软件过程建模语言.2.论文采用软件验证中的有限状态验证技术,给出了一个过程模型动态分析的方法,避免了过程模拟技术中由于取样限制而可能得到不准确的分析结果的缺点.论文中将软件过程模型的行为和需要验证的性质表示成Buchi自动机的形式,证明了分析过程模型的动态语义相当于判断某个Buchi自动机的非空性的问题.3.论文提出了分析和处理过程模型动态演化造成的影响的方法.其中,为了防止用户在演化中随意地修改过程模型,论文提出了"目标一致性",即演化后的过程模型能够通过继续运作达到预定的功能性目标,作为一般情况下判断演化是否合理的标准.论文还研究了各种演化方式对过程运作状态的影响,支持对过程运作状态进行适当交换以适用于演化后的过程模型,并在此基础上证明了可以判定演化前后的过程模型是否满足目标一致性. |
| 其他摘要 | Managing the software development and maintenance process can help to control and coordinate software projects, and further improve software quality and productivity. Software Process Technology is a key technique to support software development, project management, and quality assurance, which can be used to guide and control actual software process by modeling, enacting, monitoring, measuring, and evolving specified process model. It has been one of the focal points in the domain of software engineering. The thesis investigates the aspects involving the modeling of software process, analyzes the dynamic property of process model, and studies the methods of managing the process model evolution. The main contribution and innovation of the thesis are as follows: Firstly, in this thesis, an extensible and formalized process modeling language is defined to meet various user requirements. This language provides multiple levels of abstraction, which achieves both usability and semantic richness. The high level visualized abstraction can be readily used to describe main aspects of software process. The low level language can be used by experts to describe details of process behaviors by specifying more comprehensive semantics such as rules. Furthermore, it is extensible by supporting user-defined notations, so users can customize their preference. Additionally, it introduces patterns to specify constraints over global behaviors. Users can choose to execute any operations in process enaction that don' t violate specified patterns. This mechanism makes the language appropriate to deal with dynamic and non-determinant characteristics of software process. Secondly, the formal method of finite state verification is used for dynamic analysis of process model, which avoids the weakness of process simulation techinques that may cause inaccurate results due to the limitation of sampling. The thesis uses Btlchi automata to describe the behavior of process model and the properties that need to be verified, and proves that analyzing the dynamic property can be reduced to judging the non-emptiness of a Btichi automaton. Thirdly, methods for analyzing and managing the impact of dynamic process model evolution are proposed. The criteria of goal consistency to judge the rationality of evolution is defined to prevent ad-hoc modification of process model, which requires that the evolved process model still reach the functional goal after resuming its enaction. Also, on investigating the impact of each process evolution mode, the method of adjusting process enaction status is proposed to make process enaction status consistent with evolved process model, which can be used to determine whether the goal consistency can be maintained after evolution. |
| 页数 | 111
|
| 语种 | 中文
|
| 内容类型 | 学位论文
|
| URI标识 | http://ir.iscas.ac.cn/handle/311060/6372
|
| 专题 | 中科院软件所_中科院软件所
|
推荐引用方式
GB/T 7714 |
陈诚. 软件过程的建模、分析和演化技术研究[D]. 中国科学院软件研究所. 中国科学院软件研究所,2002.
|
除非特别说明，本系统中所有内容都受版权保护，并保留所有权利。
修改评论