Research On Ontology-based Knowledge Modeling For Rotating Machinary Fault Diagnosis

Now,for these shortcomings of the current method of fault diagnosis knowledge representation, which is versatility not strong in the description of field problems, terms and their relationship, and the lack of reasoning capacity and flexibility about the complex concepts, etc. The fault knowledge representation as one of the cores of fault diagnosis expert system, it is premise to improve the accuracy of fault diagnosis. So, it is necessary to use more perfect knowledge representation method to improve the level of knowledge management, to promote the reuse and sharing of machinery fault knowledge. Ontology, as an expression method, can describe knowledge model on semantic and knowledge level, provides a formal description of the concept, and lays a strong foundation for knowledge sharing.Combining with the NSFC project, Hunan Provincial Education Department project, Hunan University of Science and Technology Graduate Innovation Fund project, this paper takes ontology knowledge representation method as the research object, building rotating machinery fault diagnosis field knowledge ontology. By the application of Ontology theory in the field of knowledge engineering, and makes in-depth research and analysis from the following contents:(1) Researching the method of ontology building and the principles of ontology building, discussion the complexity, multi-source, correlation, hierarchy of rotating machinery fault diagnosis field knowledge. Proposed ontology as the method of rotating machinery fault knowledge representation, building rotating machinery fault diagnosis field knowledge system model, which includes rotating machinery fault, vibration characteristic, sensitive parameter, fault reason, control measure, etc.(2) Exploring a general use ontology building method which is semi-automatic in the field of rotating machinery fault diagnosis, designing rotating machinery fault diagnosis field knowledge ontology. Use of knowledge editing tool Protégé, build the ontology-based rotating machinery fault knowledge base which is classified according to classes, attributes, relations, axioms, instances, which have reaserched. The system acquires knowledge in accordance with the cencepts, attribute format which is appointed; The method which is based on the property and standard, and combined with the semantic environment is used to implement the ontology evolution.(3) Design the inheritance algorithms between the classes and the disjoint relation algorithm between the classes, it proved that there is no cycle inheritance and relation conflict in the ontology which we built, that is to say the knowledge is consistent in the ontology. Defined transitivity, associative these axioms restriction, implement the reasoning of rotating machinery fault knowledge ontology and the rotating machinery fault diagnosis which is based ontology.