KADS(2): A knowledge aided design system for mechanical transmissions

William Mark McVea, Purdue University

Abstract

During the concept phase of a new design most designers review many ideas before they make their final selection. This iterative method, is inefficient for generating workable solutions, but often reveals the best features of various designs. A knowledge based system is one of the most common computer tools for assimilating and managing information. This tool can be used to expose a designer to multiple ideas, suggestions and the experiences of senior designers, while minimizing the amount of time required to select the most productive design path. This thesis presents a design methodology in a knowledge based architecture, to provide designers with multiple related design ideas and reduce the iterative nature of design. Design suggestions are derived from heuristics obtained from design experts and are presented in a context sensitive manner. The benefit to a designer is a reduction in the amount of time spent on concept iteration. Also optimal presentation and use strategy for common engineering design and analysis tools, will ensure that a user follows good design practices, has been developed and implemented within KADS$\sp2$. KADS$\sp2$ has separated the initial automated design process from the results review and design iteration functions. This functionality closely mimics the cognitive processes a designer, uses to generate and review new ideas. Separating creative activities from modification also provides a clear concept in the mind of the designer, in terms of the task he is engaged in.

Degree

Ph.D.

Advisors

Haghighi, Purdue University.

Subject Area

Agricultural engineering|Mechanical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS