Wavelet Analysis-based Aero-engine Vibration Fault Diagnosis

As an important part of aircraft, aero-engine has high rotation speed and complex structure. It has a mission to provide thrust and support for aircraft and its stability is directly related to the quality of the mission. With the development of modern aviation technology,the function of aviation engine is improved, increasing demands on the aero-engine failure.These cases reflect the importance of aero-engine fault diagnosis. Therefore, taking effective control measures to reduce engine vibration failure and developing reasonable measures to effectively eliminate vibration failure has become one of the focuses of current research on engine.Frequency analysis is one of widely useful methods in the aero-engine fault diagnosis.The FFT can be used directly to analyze the frequency for the vibration signal. But the defect of FFT is lower resolution. The wavelet package and harmonic wavelet have higher resolution and can be divided at any time and any frequency.Firstly, in this paper, the principle of the vibration and the failure mode of the aero-engine are studied and summarized because of the complexity of the vibration phenomenon and the characteristics that much force and factor can arose the aero-engine vibration. Secondly, it is inevitable to be interfered by a large amount of noise signal in the process of signal gathering. It’s a main step to denoise and extract original signal. In this paper, the results of some kinds of denoising methods for the same signal denoising simulation are compared and the best method for selecting the simulation results is used in this paper. And then, this paper studies the application of the wavelet analysis in aero-engine vibration signal feature extraction. The wavelet packet energy spectrum and the spectrum analysis of aero-engine vibration signal under different states are analyzed. In order to obtain better denoising effect, not only appropriate wavelet function but also the best decomposition layer must be chose. The results show that in some cases, the spectrum analysis can’t clearly draw the conclusion. Finally, according to the above questions, these kinds of the aero-engine fault feature are constructed based on wavelet analysis(discrete wavelet transform and continuous wavelet transform) which are normal operating conditions. The vibration fault ofa certain aero-engine is diagnosed by using these methods. The results show that these methods all can obtain a good vibration effect of the fault. The fault identification rate and the stability of the fault identification both meet the requirements. Therefore, the effectiveness of the fault diagnosis methods and process in the actual fault diagnosis are verified. Some practical experience and experience in the vibration fault diagnosis of aero-engine are obtained to choose the most suitable model for the fault diagnosis of this kind of engine.