Online monitoring and fault diagnosis of steam turbine is an important subject in the field of steam turbine. Vibration signal was paid close attention because it includes much information about machinery state and can reflect the variation of many parameters and many happening faults . By now, the analytic methods of vibration signals are relatively perfect. Though it is, it was felt that it is not enough only by vibration signals to assess machinery state fully and diagnose each typical fault.-More-state parameters should-be collected, processed, and integrated. Integrated diagnosis methods, including thermodynamic parameter diagnosis beside vibration signal diagnosis, should be taken. In recent years, thermodynamic parameter diagnosis for plant steam turbine was proposed. Online thermodynamic parameter diagnosis of industry steam turbine is more of characteristics, and corresponding studies were taken recently, and many problems are pressing to be solved. So, some problems about online monitoring and fault diagnosis of industry steam turbine were discussed in detail in this paper.More accurate and quick calculation methods about steam property and steam turbine's off-design characteristic are need for steam turbine thermodynamic performance assessment is the basis of thermodynamicparameter diagnosis. So, two steam property calculation formulas------IFCand BTH were both used in this paper------high pressure zone BTH andother zones IFC, and a mixture calculation method, which is fit to computer calculation and easy to understand, was proposed for detail off-design calculation in this paper. An expected performance calculation and assessment method of running steam turbine was also proposed in the paper. By using methods above, a steam turbine set thermodynamic performance online monitoring and assessment system was developed for Fushun Ethlyene Co., and this system was used for optimum running, producing management and fault prevention.Performance forecasting is an important method for steam turbine fault diagnosis, and it can supply a science basis for predictive maintenance as well. In this paper, two forecasting methods, Gray system and Markov chains, were compared, and used in steam turbine's performance forecasting. Actual examination shows that, the Markov forecasting method based Gray system...
|