1990 Fiscal Year Final Research Report Summary
Knowledge Information Processing System for Conceptual Design Support Based on Function - Structure Analyses
| Project/Area Number |
63550308
|
|---|
| Research Category |
Grant-in-Aid for General Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Research Field |
計測・制御工学
|
|---|
| Research Institution | Kyoto University |
|---|
Principal Investigator |
KATAI Osamu Kyoto University, Faculty of Engineering, Associate Professor, 工学部, 助教授 (50089124)
|
|---|
| Project Period (FY) |
1988 – 1990
|
| Keywords | Conceptual Design / Value Engineering / Functional Analysis / Explanation-Based Learning / Knowledge Acquisition / Machine Learning / Axiomatic Design Approach / Operationality of Knowledge |
|---|
| Research Abstract |
This research has been done with the purpose of constructing Knowledge information processing systems to support conceptual design tasks.
Main results are summarized as follows :
1. By introducing the functional analysis methods in Value Engineering, a way of analyzing designed objects from the teleological, causal and economical rationalities, which result in multilayered modeling of designed objects.
2. Based on this method of analysis and a machine learning method called Explanation-Based Generalization, a way of acquiring varied design knowledge from designed objects is introduced and implemented using a logic programming language.
3. The operationality of the acquired knowledge is analyzed in relation to the mode of knowledge acquisition, by which we can find the most suited phase of design for each piece of the acquired knowledge.
4. Based on the idea of Axiomatic Design Approach, the above knowledge acquisition system is extended so that it can also extract strategic design knowledge from the analyses of designed objects, by which design processed are efficiently guired.
5. A way of searching and indexing for the above acquired knowledge is introduced by referring to the methods of Explanation-Based Generalization and Case-Based Reasoning. Also, a way of modifying and combining designed objects is introduced in relation to design requirements.
6. In order to build effective human interfaces for the support system, a representational method of modification and combination of design models is introduced, by which a designer (user) can grasp the essential effects of modification and the consistency of the combination.
|
