Construction Of Knowledge Base On Steam Flow Excited Vibration Fault Of Turbo-generator Units And Application In Fault Identification And Diagnosis

Steam flow excitation vibration fault is one of the common fault in the operation of turbine generator units.The more high-parameter-units developing,the more serious influence of steam flow excitation vibration.Based on a large number of steam flow excitation vibration fault case,I deeply researches steam flow excitation fault.In order to build the knowledge base,I analysed the three essential factors of knowledge base,and developed the appropriate rules in the construction of steam flow excited vibration fault knowledge base.I Analysis knowledge acquisition from three aspects,including knowledge structure,diagnostic knowledge acquisition and heuristic knowledge acquisition,develop the knowledge acquisition of rules that is from the direct method to the indirect method,do pretreatment and finishing about diagnosis knowledge, then arrange diagnosis knowledge and find out the useful knowledge.Diagnosis knowledge representation is another factor of knowledge base,through to research the steam flow excited vibration characteristic,the rules of diagnosis knowledge representation is "Frame Pattern+Production rule Pattern".By the cases,I used the rulers to descripe the various components of factors of Steam flow excited vibration fault diagnosis.Because of diagnostic application of knowledge is the ultimate objective,using hybrid reasoning can realize.Based on "Frame Pattern+Production rule Pattern" and hybrid reasoning,decided the search mode is heuristic search mode.I formulate the framework of steam flow excited vibration fault knowledge base,The knowledge base includes the diagnosis knowledge,background knowledge,unit case, diagnosis knowledge includes rules,countermeasures library,library and dictionary,I analysis each component aeparately and give suitable for steam flow excited vibration fault diagnosis and reasoning process,then use actual case to analysis and verificate.