Study On The Fault Diagnosis Of Steam Turbine Flow Passage Based On Intelligent Algorithm

Now most of the coal-fired power plant steam turbine has the characteristics of high parameter and large capacity, but it is easy to cause steam turbine flow passage failure at run time, so in order to ensure the safe and economic operation of the unit, the fault monitoring and diagnosis is necessary.Based on SOM neural network, BP neural network, fuzzy theory, rough set theory and support vector machine intelligent methods such as analysis, combined with the advantages of various theories and technology, the steam turbine flow passage fault diagnosis done some research, mainly completed the following several aspects work:(1)On the basis of researching fault diagnosis of steam turbine flow passage of domestic and foreign, analyzes its fault diagnosis mechanism, the reason for the fault generation,prevention measures and their impact have been summarized, summed up the common faults and fault symptom of steam turbine flow passage the relationship between the signs, flow passage of knowledge base for fault diagnosis is established.(2)Since the complexity of steam turbine flow passage, between the fault coupling is strong, there is fuzziness between fault symptoms, fuzzy theory concept established fuzzy SOM neural network and SOM-BP composite neural network model, the flow passage of steam turbine fault diagnosis.(3)For characteristics of redundancy information the presence of fault data, using the concept of rough set theory attribute reduction of steam turbine flow passage scaling data preprocessing, and support vector machine classification method, the regulator valve fouling,adjusted level scale, high-pressure cylinder level group scaling, fouling the group stage low-pressure cylinder fault diagnosis.