Font Size: a A A

Simulation Research And Intelligent Fault Diagnosis For Boiler Four-tube Leakage Faults

Posted on:2015-05-26Degree:MasterType:Thesis
Country:ChinaCandidate:N S WangFull Text:PDF
GTID:2272330434457772Subject:Control engineering
Abstract/Summary:PDF Full Text Request
The boiler four-tube leakage faults refer to the working substance leakage in thetubes of superheater, reheater, economizer and waterwall inside the boiler,caused byoverheating, wear, corrosion, subpar welding quality,et al. Due to the long-termoperating environment of high temperature and pressure, four-tube leakage faults are themost common faults in a coal-fired boiler,which leads to not only abnormal shutdown,economic loss but also possible life-threatening to operating personnel. Because thesteam tubes are connected to each other and the operating parameters are highlycoupled,the phenomenon of the four leakage faults are with certain similarity. Thephenomena of a same fault will be different under different operating modes. So the tubeleakage inside the boiler is very difficult to be located accurately. In order to locate theleakage failure position depending on the steam temperature、steam pressure and otheroperating parameters, it is of great necessity for us to understand the changing rules ofthe four-tube leakage faults under different operating modes.With the help of a full-scope simulator, detailed fault simulation tests were carriedout for the four-tube leakage faults of a600MW supercritical boiler unit under differentcoordinated control modes. By selecting characteristic parameters that are mostlyrelevant to the faults, the fault knowledge set for the boiler’s four-tube leakage faults wasbuilt through parameter pretreating. and then a BP neural network fault diagnosis modelwas trained off line. An intelligent fault diagnosis method, which combines artificialneural network with symptom zoom technology, was applied to realize online diagnosisof the four-tube leakage faults. Fault diagnosis program was developed with MATLABsoftware. By communicating with the simulator, detailed fault diagnosis simulation testswere carried out,which showed that this method can recognize and locate the four-tubeleakage faults correctly with good engineering practicability.
Keywords/Search Tags:supercritical boiler, four-tube leakage, simulation tests, artificial neuralnetwork, fault diagnosis
PDF Full Text Request
Related items