Condition Monitoring And Fault Diagnosis Of Auxiliary Equipment In Power Plant

The condition monitoring and fault diagnosis of power plant auxiliary equipment are very important to ensure the safety of power production and reduce the cost and enhance the economic efficiency.In this paper,MPS medium-speed coal pulverizer is taken as an example,and a curve similarity The information fusion fault diagnosis method,which is the probability of fault occurrence,is characterized by strong fault identification ability,and can diagnose the occurrence of faults in a short time,so as to prompt the operators to take measures to prevent further development and reduce fault losses.In this paper,various parameters of MPS medium-speed coal pulverizer and common fault types are summarized and analyzed.The causes of various faults,the change of relevant parameters during fault development and the general methods of dealing with faults are analyzed.The following provides the theoretical basis for selecting condition monitoring parameters for fault diagnosis.Then,the field operation data of the power station are collected,the mechanism and data of the MPS mediumspeed coal pulverizer are modeled,the unknown coefficient of the model is identified by genetic algorithm,the coal mill model is established and the model output data is used to compare the field data.,Which indicates that the model of coal mill can well reflect its dynamic characteristics and has good consistency.On the basis of modeling,the model is added PID controller,simulate the failure,simulate three types of faults: grinding coal,grinding coal,grinding spontaneous combustion.By studying the variation of various parameters,it is shown that the model is accurate and effective in simulating the fault,and can reflect the changing characteristics of each parameter.The trend of the parameter can be used as a typical sample of fault diagnosis.The fault simulation shows that there are enough fault information between different monitoring parameters when different faults occur.The problem of fault diagnosis is transformed into the problem of comparing the similarity between the curve of the field data and the typical curve of the fault at the same time scale.First of all,the curve similarity is proposed,which includes two aspects: the trend of the curve and the distance between the curves.The Pearson correlation coefficient is used to characterize the change trend of the curve.The credibility function based on the normal distribution probability function is used to measure the distance between the typical sample curve and the field data in the same time scale.Then the similarity function.The DS evidence theory is used to diagnose the multi-evidence fusion fault,and the algorithm is validated by the data of the coal failure at the scene.It is shown that the similarity function Reasonable feasibility,accuracy and high sensitivity,high resolution,failure can be found in the early failure of the occurrence of timely warning.The research method in this paper belongs to the general method and has strong expansibility.