Failure Warning Method Of Power Generating Equipment Based On Similarity Modeling

Thermodynamic cycle system failure occupies a very large proportion in the powergenerating equipment failure. Currently, the thermodynamic cycle system failure warningsystem has fewer actual application. So it has very practical significance to carry outthermodynamic cycle system failure warning research. Therefore, this paper based on therich thermodynamic cycle systematic observation parameters of incentive-type simulator,study on the high-pressure heater failure warning by using similarity-based modelingnonlinear state estimation techniques (NSET). Based on the Similarity-based modelingmethod this paper provide a performance degradation prediction method for solvingthermal power generation equipment failure warning and maintenance plan. Maincontents are as follows:(1) Research on the high-pressure heater system simulation, building the simulationmodels based on LN simulator system to access the large amounts of data required forfailure warning.(2) Through the in-depth analysis of high-pressure heater, the observation parameters forthe failure warning method can be confirmed.(3) Based on NSET solving the complex thermodynamic cycle system state estimationproblem. Give a NSET state estimation method based on historical data.By in-depthstudy on its principles and specific method, the process of memory matrix and theselection of simulation operator have been optimized.(4) Take the high-pressure heater as example. By using the above method of powergeneration equipment failure warning to handle the selected observations parameters wecan get the residuals of the failure system. And then through the sliding window and thethreshold test method, and proposed method of determining similarity we can get thesystem failure warning. This example illustrate the method described herein can be agood solution of the problem generating equipment failure warning.(5) Study on the relationship between the actual NSET state estimation residuals andsystem degradation. A method has been given to predicte performance degradation, andwhat’s more to overhaul planning guidance.