Relational Programs

关系型程序

• 批准号: 9900922
• 负责人: Farokh Bastani
• 金额: $ 18.26万
• 依托单位: University of Texas at Dallas
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1999
• 资助国家: 美国
• 起止时间: 1999-10-01 至 2002-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9900922&HistoricalAwards=false
• 关键词: Relational; Programs

CCR-99000922 Bastani, Farokh B. The University of Texas at DallasThe increasing automation of critical applications is imposing increasing demands on software capabilities and assurance levels. To handle these challenges, several methods have been proposed to decompose requirements specifications into disjoint pieces that can be implemented independently. However, most of these methods do not directly facilitate the assessment of the resulting system. This research focuses on IDEAL (Independently Developable End-user Assessable Logical) components, i.e., components that can be analyzed, developed, and evaluated independently by the end-user. To compose a system of IDEAL components, each component is relational (returns all possible outputs for a given input). This approach has severalsignificant advantages, including high reliability and safety assurance, automated composition, and support for multiparadigm control strategies. A formal foundation for the existence of relational programs for several classes of applications will be developed. A library of heuristic rules, data types, operations, and coordination schemes will be developed to facilitate the writing of efficient relational programs. Methods will be developed to infer the properties (performance, reliability, safety, and stability) of a system of IDEAL components from the properties of its components. These will result in a rigorous and practical approach for achieving certifiable high assurance for critical applications.

CCR-99000922 Bastani，Farokh B.得克萨斯大学达拉斯分校的关键应用程序的自动化越来越多，对软件能力和保证水平的需求增加。 为了应对这些挑战，已经提出了几种方法将需求规格分解为可以独立实施的不相交部分。 但是，这些方法中的大多数并不直接促进对所得系统的评估。 这项研究着重于理想的（独立开发的最终用户可评估的逻辑）组件，即可以由最终用户独立分析，开发和评估的组件。 为了组成理想组件的系统，每个组件都是关系的（给定输入的所有可能输出）。 这种方法具有不大的优势，包括高可靠性和安全性保证，自动组成以及对多载体控制策略的支持。 将开发为存在几类应用程序的关系计划的正式基础。 将制定启发式规则，数据类型，操作和协调方案的库，以促进有效的关系计划的撰写。 将开发方法来推断其组件属性系统系统系统的属性（性能，可靠性，安全性和稳定性）。 这些将导致一种严格且实用的方法来实现对关键应用的可认证高保证。