콘텐츠 본문
논문 해외 국제전문학술지(SCI급) Automatic Assembled Shape Generation using Genetic Algorithm in Axiomatic Design
- 학술지 구분 국제전문학술지(SCI급)
- 게재년월 2005-09
- 저자명 npyoung Jung, Kang-Soo Lee(교신) and Nam P. Suh
- 학술지명 Lecture Notes in Artificial Intelligence - 직접입력
- 발행처명 Springer
- 발행국가 해외
- 논문언어 외국어
논문 초록 (Abstract)
Axiomatic design is a powerful tool for in designing a product and describes functional requirements (FRs) and design parameters (DPs). In this paper, we extended axiomatic design by introducing shapes and databases. We used functional geometric features (FGFs) for the shape of DP. The databases store links of FR-DP-GE and their hierarchies. The retrieval of proper FGFs from the database is performed by matching a query FR with stored FRs linked to the corresponding FGFs by a lexical search based on the frequency of words and the sequence of the words in the FR statements using a synonym checking system. A computer algorithm automatically combines and assembles the retrieved FGFs. Genetic algorithm (GA) searches for the best matching of interface types between FGFs and generates the corresponding assembly sequences based on the codes of the chromosomes. Geometric interface-ability between FGFs is calculated as a value of INTERFACE_metric. The higher value of INTERFACE_metric means better the design solution for the given FRs that must be satisfied in the sense of language and geometric interface matching.
