Fault Diagnosis Of Turbine Generator Sets Based On Ontology And Case Reasoning

As the critical equipment in electric power system,turbine generator sets is a large and complex rotating machine with rigorous requirements on the reliability,safety and life,and it works for long time in the bad external environmental conditions.Once faults occur,they will bring huge losses and harms.Therefore,in order to ensure the safety and reliability of the operation of turbine generator sets,it is urgently demanded to diagnose the abnormal operation of turbine generator sets accurately.With the trend of intelligent manufacturing,realizing the integrated utilization of existing knowledge of fault diagnosis for turbine generator sets has become an important part of intelligent fault diagnosis.In view of the advancement of ontology in the field of knowledge representation,semantic ontology technology and turbine generator sets fault diagnosis technology is combined to conduct the thorough and systematic research in this thesis.The main contributions and innovative achievements of this thesis are summarized as follows:(1)A suitable knowledge model for the fault diagnosis of turbine generator sets is established.Aimed at the problem that the fault diagnosis domain terms of turbine generator sets are complex,heterogeneous and difficult to represent and share,the advantages of ontology in the field of fault diagnosis knowledge representation are considered,and the traditional method named seven-step is improved as the construction method of fault diagnosis ontology of turbine generator sets,which makes up for its defects in ontology evaluation and tracking update.According to the principle of ontology construction,the ontology of fault diagnosis for turbine generator sets is established by Protégé,which provides a clear formal representation for fault diagnosis knowledge.The detailed procedures include: knowledge acquisition,knowledge storage,improvement of construction method,selection of editing tool and description language of ontology,definition of class,relationship,attribute,axiom and individual in ontology and so on.(2)The feasibility and effectiveness of the fault diagnosis ontology for turbine generator sets are verified.In view of the inconsistency that may exist in the ontology,the algorithm for consistency based on Tableau algorithm is designed to check the ontology.The SQI mechanical fault simulation test bench is used to simulate the different fault types of turbine generator sets,and the acquired fault information is collected and analyzed.Finally,the ontology knowledge is tested by reasoning.(3)A fault diagnosis method for turbine generator sets based on ontology and case reasoning is proposed.The advantages of the ontology in the field of case representation areclarified,and the main components of the case representation are analyzed.The mathematical models of semantic distance,semantic depth and semantic density are quantified.The semantic similarity algorithm is simplified and improved,and an O-CBR hierarchical retrieval model based on semantic and case attribute similarity algorithm is established.Last,the method of fault diagnosis is verified through case studies.(4)The intelligent fault diagnosis system of turbine generator sets is designed and developed.The development framework and operation process of system is studied,including the creation of ontology knowledge base and O-CBR search module,and the design of main function interface of system.The development of fault diagnosis system is realized by integrating Protégé,Visual Studio C# and SQL Server.It improves the human-computer interaction of the whole fault diagnosis process,and makes the operation convenient and efficient.Finally,it is verified that the system can provide decision support for the fault diagnosis of turbine generator sets efficiently and accurately.