A Knowledge-Based System Methodology for Conceptual Design of Mechanical Systems

Abstract

This paper presents a knowledge-based system methodology, called Domain Knowledge Transformation (DKT), that can be used at the conceptual design stage of mechanical systems. The DKT model is a two-levels process that reveals the insights of the methodology. In the first level, defined as the preconcept level, a list of design specifications is developed through the interaction between an intelligent system and the users or customers. In the second level, defined as the concept level, design solutions are generated using a functional requirements mapping (FRM) method. Design using FRM is characterized by the ability to structure hybrid sets of design knowledge, relevant information and customer needs. FRM method define results of current design efforts, future tasks and their priorities, various metrics, and a presentation format for decision making. After briefly introducing the DKT methodology, the paper shows how design requirements are generated and stored by questions and answers interface, objective tree guided structured interview and the dynamic knowledge base in the preconcept level. The paper then shows, through a real world design example, how FRM can be used to provide a detailed association between statements of design requirements (customer needs) and the physical entities proposed as solutions. The paper concludes that DKT methodology could be a useful concurrent engineering tool in the early design stage.