TY - JOUR
T1 - Knowledge and engineering parameter mapping technology supporting product conceptual design
AU - Mo, Zhenchong
AU - Gong, Lin
AU - Ye, Fan
AU - Fu, Tie
AU - Zhu, Degang
AU - Cui, Haoran
AU - Xie, Jian
N1 - Publisher Copyright:
© 2022 The Authors.
PY - 2022
Y1 - 2022
N2 - In the fuzzy front stage of product conceptual design, requirements are often expressed briefly as the expectation of product effect. However, the mapping relationship between this kind of effect expectation requirement and product design knowledge is difficult to explore and establish, and it is more difficult to guide designing process objectively. In order to expand product conceptual design ideas, professionally search the design solution space, and improve user satisfaction, this article proposes product design knowledge and engineering parameter mapping technologies that support conceptual design: First, construct a domain knowledge graph based on patent text. Extract entities-such as effect knowledge, functional knowledge, technical knowledge, and structural knowledge-and inter-entity relationships from patent texts through BERT technology to realize the construction of domain knowledge graph; Second, in view of the fuzziness and incompleteness of the front stage design requirements, by introducing General Engineering Parameters (GEPs), construct the domain product-engineering parameter-structure (PGS) mapping relationship library. On the one hand, the library realizes the professional representation and completion of the product design requirements from the engineering perspective; on the other hand, it improves the mapping relationship between the difficult-to-identify relationships between design requirements and their structural solutions in the domain knowledge graph; Third, a product conceptual design process model based on PGS is proposed. The model takes the professional and complete requirements expressed as engineering parameters as input, which matches patent candidate sets through effect knowledge in the domain knowledge graph. Then select and construct a structural solution morphology matrix through computational evaluation methods to support the generation of product design conceptions; Finally, an example of manipulator design verifies the effectiveness of the knowledge library and the technical solution proposed by this research.
AB - In the fuzzy front stage of product conceptual design, requirements are often expressed briefly as the expectation of product effect. However, the mapping relationship between this kind of effect expectation requirement and product design knowledge is difficult to explore and establish, and it is more difficult to guide designing process objectively. In order to expand product conceptual design ideas, professionally search the design solution space, and improve user satisfaction, this article proposes product design knowledge and engineering parameter mapping technologies that support conceptual design: First, construct a domain knowledge graph based on patent text. Extract entities-such as effect knowledge, functional knowledge, technical knowledge, and structural knowledge-and inter-entity relationships from patent texts through BERT technology to realize the construction of domain knowledge graph; Second, in view of the fuzziness and incompleteness of the front stage design requirements, by introducing General Engineering Parameters (GEPs), construct the domain product-engineering parameter-structure (PGS) mapping relationship library. On the one hand, the library realizes the professional representation and completion of the product design requirements from the engineering perspective; on the other hand, it improves the mapping relationship between the difficult-to-identify relationships between design requirements and their structural solutions in the domain knowledge graph; Third, a product conceptual design process model based on PGS is proposed. The model takes the professional and complete requirements expressed as engineering parameters as input, which matches patent candidate sets through effect knowledge in the domain knowledge graph. Then select and construct a structural solution morphology matrix through computational evaluation methods to support the generation of product design conceptions; Finally, an example of manipulator design verifies the effectiveness of the knowledge library and the technical solution proposed by this research.
KW - Knowledge graph
KW - conceptual design
KW - engineering parameters
KW - patent analysis
UR - http://www.scopus.com/inward/record.url?scp=85133505950&partnerID=8YFLogxK
U2 - 10.1016/j.procir.2022.05.264
DO - 10.1016/j.procir.2022.05.264
M3 - Conference article
AN - SCOPUS:85133505950
SN - 2212-8271
VL - 109
SP - 368
EP - 374
JO - Procedia CIRP
JF - Procedia CIRP
T2 - 32nd CIRP Design Conference, CIRP Design 2022
Y2 - 28 March 2022 through 30 March 2022
ER -