During product lifecycle management (PLM), product information from CAD/CAE applications regularly needs to be exchanged and shared between the various applications. However, these applications often have different product data semantics and corresponding representations. The interoperability problem caused by the heterogeneous semantics and data representation is critical and needs to be addressed and automated. Recent research has focused on integration frameworks for CAD/CAE applications in order to improve interoperability. There are fundamental problems that still need to be addressed.; We identified the following important roadblocks and sought to address these specifically in our work: (1) The need for an adequate product knowledge representation of engineering design/analysis, which is easily expandable, and customizable for traditional and non-traditional (e.g. virtual prototyping) design information systems that also allows the sharing of product data semantics across all these heterogeneous systems to support distributed, collaborative engineering capabilities; (2) The need for a way to generate product data semantics by using engineering design/analysis knowledge to interpret actual product data; (3) The need for a way to reconcile the differences in the different product semantics by finding underlying similarities between different knowledge representations that are from different viewports and reconcile, and use these similarities to then translate product data semantics correctly.; This dissertation proposes an ontology-based approach for a semantic level exchange and integration to improve interoperability, which includes an ontology building tool, ontology mapping tools and custom tools to associate ontologies to prduct data. For the purpose of semantic level integration, a way of representing engineering design/analysis knowledge using an engineering ontology is proposed. A layered structure is used for building knowledge into engineering ontologies so as to improve the scalability and composition adaptivity. Based on the knowledge, a semantic layer is built upon product data to use concepts/relations in ontologies to describe actual product data, which can be used to represent understandings about a product design from different perspectives. To enable translating different understandings (product data semantics) using different ontologies, an ontology mapping method is proposed to find matching concepts between different ontologies, based on three basic relation types between concepts: composition, inheritance and attribute.; A scenario is explained to describe the working mechanism of the system and to demonstrate the concept of semantic level integration framework for a simple example. A sample assembly is designed and simulated in different software packages and an integrated process is made to exchange information between them. The scenario successfully demonstrates the ontology based approach.
展开▼
