Research On Aero-Engine Fault Diagnosis Based On Its Whole Body Vibration Signals

An aeroengine is a machine with high rotation speed, its reliability directly affects the flight safety of airplane. With the development of science and technology, the thrust, rotation speed, dynamic intensity of aeroengine and so on increase gradually, the working processes and structural forms of aeroengines also become more complex, and then the examples of vibration destruction become more remarkable. The condition monitoring and fault diagnosis of aeroengine is being more and more important.It is an important means for determining the conditions and making fault diagnosis of the aeroengine by measuring its whole body vibration and then mastering vibration characteristics accordingly.In this paper the whole body vibration test and signal analysis technology of aeroengines are presented in detail. Three methods for aeroengine whole body fault diagnosis are studied systemically, which are based on spectrum analysis, nerve network (NN) and fault equation respectively. By analyzing the whole body vibration signal in time and frequency domain, vibration faults of aeroengine are diagnosed based on the spectrum characteristics. The NN model for the aeroengine whole body vibration fault diagnosis is built, and was trained with a mount of samples to make it have the ability of identifying the vibration fault patterns. The fault equation is built based on the main component model of fault diagnosis of aeroengine. The vibration faults of aeroengine are diagnosed by applying the optimization estimation and the main component analysis method to calculate the fault equation.In this paper, the aeroengine whole body vibration fault diagnosis system was also set up according to the above three fault diagnosis methods based on MATLAB and Visual C++. Furthermore, the whole body vibration faults of aeroengine are diagnosed. It proved that the above three fault diagnosis methods presented in the thesis can efficiently diagnose faults of aeroengines. The vibration fault diagnosis effects are improved by using the synthetical fault diagnosis method. Methods used in the diagnosis and system designed in this paper are valuable for aeroengine whole body vibration fault diagnosis.