Research On Aero-engine Fault Diagnosis Technology Based On Gas Path Parameters

Aero-engine is the "heart" of aircraft, the source of power for aircraft flying and a machine to translate the heat of combustion of fuel into mechanical energy called heat engine. Because of its complex structure and a long period of high temperature, high-pressure work environment has become the main source of the fault plane. Thus, real-time monitoring of aero-engine operation and timely and accurate determining fault location and the cause of the engine fault is particularly important. Aero engine maintenance, as appropriate, is the fundamental methods and the inevitable trend to improve maintenance efficiency, reduce maintenance costs and protect airline security. And the prerequisite of condition maintenance is accurate state analysis and effective fault diagnosis.Firstly, it describes the current domestic and international status of aviation engine condition monitoring and fault diagnosis technology and related concepts, engine fault diagnosis principle and the main diagnostic methods are briefly analyzed and in-depth study of the QAR data preprocessing methods.Secondly, it introduces the basic principles of support vector machines, bases QAR(quick access recorder) historical data and applies support vector machine regression SVR(Support Vector Regression) method on establishing the baseline model of aviation engines and being verified. The combination of twin differences monitoring, trend analysis and fingerprints Figure diagnostics, twin differences instead of the state amount error, provides a method of fault diagnosis. It avoids dependence on the baseline diagnostic methods, and achieves self-monitoring and diagnostics.Finally, using MATLAB engine interface to achieve VC++ and MATLAB mix programming, developing aero-engine fault diagnosis system providing a powerful help for aero-engine fault analysis.