Classifying Component Function in Product Assemblies With Graph Neural Networks

A smart conflict resolution model using multi-layer knowledge graph for conceptual design

Reusing design knowledge of products is a useful way to solve the efficiency issue of complex product design. The design knowledge is tacit, empirical, and unstructured and there exists insufficient case matching and inefficient design reuse in complex products design process. Aiming at these problems, this paper presents an improved case-based reasoning methodology combining ontology with two-stage retrieval. Firstly, a knowledge domain ontology model of complex product design is constructed, and the technology of ontology-based data access is introduced to automatically generate a case knowledge base with semantic information. Then, a new two-stage case retrieval method integrated semantic query with similarity calculation is proposed. The case subset is selected by query statements. It has the characteristic of isomorphism with design problem. The retrieval mechanism is applied to compress the traversal space, reduce the redundancy of semantic similarity calculation, improve the retrieval efficiency, and fulfill the target of case reuse. Finally, a variant design of the chiller unit as an example is executed to illustrate the use of the proposed method, and experiments are organized to evaluate its performance. The result shows that the proposed approach has an average precision of 92% and high stability, outperforming existing methods.

Case representation and retrieval for complex product design based on case-based reasoning

A hierarchical assembly knowledge representation framework and microdevice assembly ontology

Knowledge representation and reasoning using fuzzy Petri nets: a literature review and bibliometric analysis

An approach to capturing and reusing tacit design knowledge using relational learning for knowledge graphs

Product design relies on various design knowledge, including explicit design knowledge and tacit design knowledge. This paper aims at issues that the current design knowledge representation methodology lacks scalability and flexibility, and presents a design knowledge representation framework combining CBR with knowledge graphs. Based on the framework, design cases are represented as a set of knowledge graphs. An ontology model for design case representation is proposed. A novel approach to retrieving design cases from knowledge graphs is presented, which uses a design problem query graph instead of keywords to match similar case subgraphs from knowledge graphs. A semantic similarity assessment method based on subgraph similarity is proposed. Finally a case study on knowledge representation for stamping die design is provided.

Jia Jia

Papers

An approach to capturing and reusing tacit design knowledge using relational learning for knowledge graphs

Knowledge Representation Framework Combining Case-Based Reasoning with Knowledge Graphs for Product Design

An approach to automatic boundary segmentation of solid models using virtual topology: toward reconstruction of design features

A graph database for life cycle inventory using Neo4j