The Research On Intelligent Diagnosis Of Equipment In The Thermal Plant And Its Application

For some decades now,the thermal plant has been shortof equipment condition monitoring and diagnosis system of maturely applied.With the comprehensive promotion and application of the supervisory information syste m of the plant level(SIS),the real-time database as the basic software of SIS has solved the problem of DCS data storage volume,which can store 3-5 years of historical production data in the thermal plant by choosing the advanced data compression techno logy.The generalized application of real-time database and the development of data mining technology undoubtedly has provide very good technology basis for the research of the subject of equipment condition monitoring and intelligent diagnosis in the ther mal plant.This article has created equipment condition prediction model based on the past healthy sample of the equipment and the fuzzy clustering method of Gaussian mixed model,which calculates the estimated value curve of every monitoring parameter of the equipment,and gives the main degradation parameter influencing the similarity degree in high dimension space through computing the similarity value between the real condition and estimated condition,and automatically suggests the suspected equipment failure modes.This article has built spin axis monitoring model with respect to the turbine axis monitoring objective,choosing the correlation parameters and past normal sample,start model training and researching the suspected failure judging methods and helping building the equipment failure knowledge base.This article has given the introduction of the framework design,function design,database design and the features,main function and application effects of the system.The equipment intelligent diagnosis system being designed and developed has well integrated traditional equipment fault diagnosis theory with data mining theory,which makes fuzzy clustering analysis on the mass historical data of the equipment in normal operation,and automatically cl assifies the typical conditions of the equipment,and calculates the expected condition data respect to the real condition,owning the features of adapting the demands of the consumers,displaying briefly,using conveniently.The system has provided software platform for the generalization of the mode of equipment condition maintenance,and has helped the plant realize online-monitoring and evaluation and precaution handling process of equipment condition,which provides data infrastructure for the maintenanc e decision of equipment condition.