Research On The Technology Of Turbofan Engine Starting Fault Diagnosis And Fault Location

The ground starting of turbofan engine has always been a serious problem for advanced fighters to take off successfully in complex climate conditions. If the engine starting failure, it may result in the turbine blades burn down even the entire engine damage, which will cause irreparable damage. Therefore, it is necessary to investigate the technology of online fault diagnosis and fault location.The first part of this thesis is devoted to establish nonlinear mathematical model of turbofan engine under idle speed. The problem of selecting the initial guess value is analyzed. The influence of interpolation calculation method on the engine model real time performance is analyzed. The correction technology of starting model is summarized, including total pressure recovery coefficient, combustion efficiency, thermal inertia model and temperature sensor. The starting model influenced by different factors is simulated, which lays the foundation for the study of starting fault diagnosis and location.Moreover, five faults are simulated by engine model and the fault characteristics are analyzed according to the simulation results and the test data. The typical characteristics of engine are summarized to investigate the above fault; In addition, the mathematical method of dividing the threshold range is studied. According to the characteristics of starting failure mode at different starting stages, the diagnosis logic of each fault mode is determined and the accuracy of this method is verified by the test data and model data.Finally, the causes of fault are studied by analytic hierarchy process(AHP) and the fault tree is established. Moreover, the knowledge base is designed, the method of establishing a fault diagnosis expert system is studied. Taking the fault of over temperature for example, the process of simulating fault location on computer is studied. The method of combining location method with engine starting diagnosis system is studied, and the method of real-time on-line fault diagnosis and location is discussed.