Research On Acoustic Emission Signal Denoising Method Of Aero Engine Based On Wavelet Analysis And Artificial Intelligence

As the power source of aircraft,aero engine is the most sophisticated and complex power equipment.Since the aero engine is in a high-speed,heavy-load working state for a long time,the rotor system is prone to mechanical failure.Acoustic emission technology has a high sensitivity to early failures such as fatigue cracks and microdamage caused by foreign impact.It can also continuously monitor the status of structural defects,which is suitable for condition monitoring and fault diagnosis of aero engines.Because of the extremely high structural complexity of the aero engine,the acoustic emission signals of each component are coupled,resulting in the fault characteristic information is often under the background noise,which brings difficulties to the extraction and identification of fault information,the acoustic emission signal needs to be denoised before fault identification.Based on the above analysis,this paper carries out the noise reduction research of aero engine acoustic emission signal based on artificial intelligence algorithm and wavelet noise reduction.To study the selection of optimal wavelet parameters,according to the nature of wavelet basis function and the characteristics of acoustic emission signals,pick out the eligible wavelets.The object of research is the acoustic emission simulation signal that simulates metal cracks,different wavelet families are used to reduce the noise of the noisy simulation signal under different decomposition layers,and the wavelet parameter with the highest signal-to-noise ratio is selected as the optimal parameter suitable for processing the acoustic emission signal.This method overcomes the shortcomings of traditional empirical selection of parameters to achieve the best noise reduction effect.To study the problem of threshold selection in wavelet threshold noise reduction,this paper proposes a method to optimize wavelet threshold using the improved dung beetle optimization algorithm.According to the theoretical model of the original dung beetle(DBO)algorithm,the original algorithm is improved in combination with the position update strategy,and the improved dung beetle optimization(TLRDBO)algorithm is proposed to improve the convergence speed and solution accuracy of the optimization algorithm.The TLRDBO algorithm and wavelet threshold noise reduction are used to reduce the noise of the simulated signal and the real aeronautical acoustic emission noise signal.Several noise reduction methods are used in comparative experiments to verify the effectiveness of the noise reduction methods mentioned in this article.The intermediate bearing of aero-engine rotor system is prone to failure,carried out acoustic emission monitoring experiments of bearing defects.Obtained bearing fault signals and noise signals under different working conditions through static experiments and dynamic experiments,multiple sets of noisy bearing fault signals are formed with different signal-to-noise ratios.The noise reduction results show that the noise reduction effect of the wavelet threshold noise reduction method using the optimization threshold is significantly better than that of the traditional threshold reduction method under different noise conditions.Indicating that the noise reduction method proposed in this paper can better achieve effective noise reduction for noisy emission signals.