Research On Knowledge Graph Of STEP Design Intent Based Feature Construction Method

With the development of model-based definition(MBD)technology,it is urgent to realize the semantic exchange of MBD model including PMI information.Traditional product information exchange focuses on geometric data exchange.Since it does not contain semantic information such as design history,constraints and features,semantic information exchanging and sharing cannot be realized.Aiming at the requirement of design intent mining in product information exchange,knowledge graph associated with geometric and non-geometric information is constructed,to realize the extraction and exchange of STEP design intent.It provides a new method for exchanging design intent.The key technologies such as semantic modeling for design intention,STEP knowledge extraction for knowledge graph construction and feature-based STEP design intention knowledge map construction are mainly studied.It effectively remedies the defect that traditional ontology modeling methods cannot fully mine the hidden semantic knowledge in STEP information.The research work is as follows:1.Aiming at the ambiguity and abstraction of design intent,a multi-granular design intent model is proposed.Design intent is granulated and analyzed from different levels of granularity.With the product information semantic as the basic unit,the specific definition of the design intent model is given.2.Based on the semantic,a multi-semantic feature extraction is performed on product design intent.An improved NBA algorithm is used to realize the recognition of design features,solved composite features recognition to some extent.Besides,we define the block feature to recognize non-standard features.At the same time,PMI contained in STEP AP242 is extracted.Then construct the knowledge graph of reducers,and realize multi-dimensional design intent modeling.3.Elaborate the system realization.The knowledge graph visualization system is realized using HTML5 and D3.js,taking a STEP AP242 model as an example to introduce process and application of knowledge graph.