Knowledge‐Based Tool for Decomposing Complex Design Problems

Abstract

A design problem is viewed as a complex system divisible into modules. For designs based on novel concepts, such as designs for space platforms, the determination of subsystems, their interactions, and the participating disciplines is an important task. Before the design of such a system can begin, much money and time are spent in determining the interactions among the modules and iterative loops. This is important because the design manager must know how to group the modules into subsystems and assign the subsystems to design teams so that changes in one subsystem have predictable effects on other subsystems. Determining these subsystems is not an easy task, and often important interactions are overlooked. Very few tools are available to aid design managers in determining the hierarchical structure of a design problem and to assist them in making these decisions. The purpose of this research effort is to develop a knowledge-based tool to act as an intelligent adviser for the design manager by identifying the subsystems of a new, complex design problem, ordering the subsystems into a hierarchical format, and marking the interactions among the subsystems to facilitate the use of multilevel tools. This tool was tested in the decomposition of an experimental space system. This test indicated that this type of approach could lead to substantial savings by organizing and displaying a complex design problem as a sequence of subsystems easily divisible among design teams.