Research On Maintenance Case Representation And Similarity Calculation For Complex Products Based On Ontology

With the development of manufacturing service,the profit source of enterprises has gradually changed from manufacturing to service.The long life cycle characteristics of complex products lead to a long period of use and maintenance.Its maintenance service is an interdisciplinary and multi-step complex process,which belongs to a typical knowledgeintensive service process.In reality,maintenance personnel often need to work with the help of past maintenance experience and knowledge.Therefore,it is particularly important to study the management mode of maintenance knowledge for complex products.Case-based Reasoning(CBR)has an outstanding performance in using existing knowledge to solve new problems.At the same time,the domain knowledge sharing characteristics of Ontology make it possible to make up for the deficiencies of CBR in knowledge sharing and reuse.Therefore,in the field of complex product maintenance,this paper uses the combination of ontology and CBR to study the knowledge representation and case similarity calculation of complex product maintenance cases.Based on the business process of complex product maintenance,this paper determines to formalize the knowledge of complex product maintenance cases by ontology on the basis of analyzing the elements of maintenance cases and the requirements of knowledge representation.In terms of knowledge representation,according to the type of knowledge and the purpose of application,the case ontology of complex product maintenance is divided into top-level ontology,domain ontology,task ontology and application ontology,which achieves the independent expression of declarative knowledge and procedural knowledge under the unified framework.In case similarity calculation,a two-stage similarity calculation model is proposed,which combines semantic similarity with attribute similarity.Firstly,the semantic similarity of ontology structure is used to reduce the scope of case retrieval,and then attribute similarity is used to rank the alternative cases.Finally,the concept of case fitness is introduced to select the most reusable cases that meet the similarity threshold.The combination of ontology and CBR could provide support for the reuse of maintenance knowledge.