Application of ART neural network to development of technology for functional feature-based reference design retrieval

Author(s): Chen Y.M. | Chen, Y.N. | Chu, H.C. | Wang, C.B. |

Year: 2005

Citation: COMPUTERS IN INDUSTRY Volume: 56 Issue: 5 Pages: 428-441

Abstract: Engineering design is a knowledge intensive process. The execution of each task in the process requires various aspects of knowledge and experience. Therefore, organizing, storing and retrieving product design information, design intents and underlining design knowledge is one of the most important tasks in engineering knowledge management. This study develops a novel scheme for functional feature-based reference design retrieval using adaptive resonance theory (ART I) neural network to provide engineering designers with easy access to relevant design and other knowledge. This retrieval process includes the steps of functional feature-based query, case searching, and case ranking. The technology involves a binary code-based representation for functional features, ART I neural network for functional feature-based case clustering, functional feature-based case similarity ranking, and a case-based representation for designed entities. The objective of this study can be achieved by performing the following tasks: (i) designing a functional feature-based reference design retrieval process, (ii) developing a functional feature representation, (iii) investigating ARTI neural network, (iv) implementing a functional feature-based reference design retrieval mechanism, and (v) experimenting with functional feature-based case clustering.

Topics: Machine Learning, Applications: Industrial Control, Models: ART 1,

PDF download

Cross References