Research On Fault Diagnostic Expert System Of Steam Turbine Based On SVD Transforming And Fuzzy Inferencing

Steam turbine is the core equipment of many industrial manufacturers,the occurrence of any failure would bring huge losses, therefore, the daily maintenance and troubleshooting of the steam turbine require a high level.Most of the steam turbine fault can be found from its vibration state problems, many countries have put a good deal of effort to research the steam turbine vibration failure mechanism.Expert system is a computer simulating system of the expert decision-making ability, which can replace a large number of mental labor in many occasions, so, the development of expert system is gaining more and more attentions in many countries and industry.Steam turbine vibration fault diagnosis is a complicated multi-factor comprehensive reasoning process, research the use of expert system in steam turbine fault diagnosis has very important theoretical significance and engineering application value.At present, the traditional fault diagnosis expert system still exist some deficiencies, such as symptom information utilization rate is not high, knowledge expression is quite wooden, the efficiency of inference method is not high.This article aimed at these shortcomings, further study of the steam turbine fault mechanism, proposed combine the traditional expert system, database theory, fuzzy diagnosis technology, and SVD(singular value decomposition) technology, building fault diagnostic expert system of steam turbine based on SVD transforming and fuzzy inferencing.The system mainly developed by Lab VIEW and MATLAB, using My SQL database and SQL language to realize the input, modify and delete of diagnosis knowledge, as well as automatic fault diagnosis.This article effectively solved the problem of the uncertainty of diagnosis knowledge by organic combination the fuzzy diagnosis rules, information subordinate and credibility theory, the knowledge stored in knowledge base;Using mix reasoning control mode, can not only based on symptom acquisition module automatic acquisition vector to fuzzy diagnosis inferencing, but also through the way of man-computer dialogue in the process of diagnosis to obtain other graphic signs information to make comprehensive diagnosis, if the fault confidence level is not high enough,the system can provide SVD analysis method to detailed analysis the fault data and confirm the fault.Finally, through two steam turbine fault diagnosis example to validate the design of the steam turbine fault diagnosis expert system in steam turbine fault diagnosis is scientific and practical.Compared with the traditional fault diagnosis method of rotating machinery, this paper studies steam turbine fault diagnosis expert system is not only effective use of the experience of numerous experts, and the efficiency and accuracy in the diagnosis of aspects have greater improvement.At the same time,by changing the domain knowledge in knowledge base and fuzzy relationship matrix element, the fault diagnosis expert system in the automotive, aerospace, medical and other fields also has a certain practical value.