Title
Research on the information transfer characteristics of dimensions in the product variant design process.
Abstract
Product variant design, as one of the key enabling technologies of mass customization, is the transfer of variant information among mating parts from the perspective of informatics. A dimension constraint network (DCN) among mating parts carries on the task of transferring variant information. What are the information transfer characteristics of dimensions in a constraint network is a fundamental issue to plan the product variant design process reasonably. We begin by showing the natural dynamics of the DCN from two aspects: structure and uncertainty. The information efficiency of the DCN was proposed based on its simple path to specify the information transfer capability of the network. Based on this, the information centrality of the dimension was developed by measuring the efficiency loss of the DCN after the removal of a dimension from the network, which describes the information transfer capability of this dimension. Further, the information centrality of a part was derived. Using a spherical valve subassembly, we calculated the information centrality of the dimensions in a constraint network. We determined that the information centrality of dimension is highly correlated to its out-degree. An approach to plan the sequence of the part variant design according to its information centrality was proposed. We calculated the uncertainties of the DCN and its cumulative uncertainties under different sequences of the part variant design. Results indicate that part variant design under the descending information centrality of the parts minimizes the uncertainty of the DCN. This suggests a new method of planning the sequence of part variant design.
Year
DOI
Venue
2018
10.1017/S0890060417000233
AI EDAM-ARTIFICIAL INTELLIGENCE FOR ENGINEERING DESIGN ANALYSIS AND MANUFACTURING
Keywords
Field
DocType
Dimension Constraint Network,Information Centrality,Product Variant Design,Uncertainty
Information transfer,Systems engineering,Engineering,Design process
Journal
Volume
Issue
ISSN
32
1
0890-0604
Citations 
PageRank 
References 
0
0.34
12
Authors
3
Name
Order
Citations
PageRank
Xinsheng Xu100.68
Tianhong Yan243.47
Yangke Ding300.34