Research Initiation: Foundations of Quantitative Inference About Sets of Design Possibilities

研究启动：关于设计可能性集的定量推理基础

• 批准号: 9010393
• 负责人: Allen Ward
• 金额: $ 6.44万
• 依托单位: Regents of the University of Michigan - Ann Arbor
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1990
• 资助国家: 美国
• 起止时间: 1990-09-01 至 1993-04-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9010393&HistoricalAwards=false
• 关键词: Research; Initiation; Foundations; Quantitative; Inference

Most engineering mathematics is appropriate for representing a single object under a single operating condition. Mechanical designers, however, must reason simultaneously about many design possibilities; that is, sets of objects under sets of operating conditions. Previous work has produced a formalism for performing such inferences, called the labeled interval calculus. This formalism has been empirically validated in a mechanical design compiler, a program which for a wide range of design problems accepts mechanical schematics, specifications, and utility functions, and returns catalog numbers for optimal implementations. However, the tested version of the calculus is relatively complex, and has been only partially analyzed mathematically. The researcher seeks to simplify the calculus, develop precise definitions for all its terms and proofs for all its inferences, and clarify a number of issues surrounding it. This will enable the development of more powerful and reliable compilers, as well as furthering the basic understanding of design reasoning.

大多数工程数学都适用于表示 在单一操作条件下的单一对象。 机械 然而，设计师必须同时对许多设计进行推理， 可能性;也就是说，在操作集下的对象集 条件 以前的工作已经产生了一个形式主义，执行这样的 推理，称为标记区间演算。 这种形式主义 已经在机械设计编译器中进行了经验验证， 一种程序，它适用于各种设计问题， 机械原理图、规格和实用功能，以及 返回最佳实现的目录号。 但 测试版本的演算是相对复杂的，并已 只是部分数学分析。 研究人员试图简化微积分， 所有术语的定义和所有推论的证明， 并澄清围绕它的一些问题。这将使 开发更强大和可靠的编译器，以及 进一步加深对设计推理的基本理解。