Research On Time Frequency Analysis Method Of Rotor Rub Impact

Large rotating machinery is the key equipment of electric power,metallurgy,petrochemical and military departments.As the core component of rotating machinery,rotor system plays a very important role in the high-speed operation of rotating machinery.With the development of modern industrial technology and the improvement of process,the relationship between the equipment of production system is more and more close,which makes the probability of rotor rub impact fault increasing.The accidents caused by rotor rub impact fault seriously threaten the production safety.The diagnosis of rotor rub impact fault is of great significance to industrial safety.Rotor rub impact fault diagnosis includes qualitative analysis of rub impact and quantitative analysis of rub impact location and degree.Fault feature extraction is an important part of fault diagnosis,which is widely used in the qualitative analysis of rub impact fault.In this paper,acoustic emission detection technology is applied to rotor rub impact fault diagnosis,and the rub impact fault feature extraction method based on time-frequency analysis is studied.In this paper,a time-frequency analysis method for extracting fault features from nonlinear and non-stationary acoustic emission signals is studied,Compared with time domain analysis or simple frequency domain analysis,time-frequency analysis can reflect the frequency characteristics and corresponding time of fault at the same time.The common timefrequency feature extraction methods include short-time Fourier transform,wavelet transform,EMD and so on.According to the characteristics of acoustic emission signal of rotor rub impact fault,this paper studies the application of maximum correlation kurtosis deconvolution in feature extraction of rotor rub impact fault,and optimizes the parameter selection method of MCKD method;at the same time,it analyzes the characteristics of different time-frequency analysis methods,and studies the application of time-frequency analysis in signal processing of rotor rub impact fault acoustic emission,using MCKD,S transform and so on Methods the time-frequency analysis of the rotor rub impact AE signal was carried out,which proved that the time-frequency analysis method can effectively extract the time-frequency local features of the rotor rub impact AE signal.This paper mainly introduces the basic theory of time-frequency analysis methods such as maximum correlation kurtosis deconvolution and S-transform,and designs rotor rub impact experiments at three speeds,and collects acoustic emission signals of rub impact faults at various speeds.Through the simulation signals and experimental acoustic emission signals,the noise is removed by MCKD,and the fault pulse is enhanced,the same as the above According to the envelope autocorrelation function,the parameter selection principle of MCKD method is improved.The improved MCKD method is used to analyze the acoustic emission signal of rotor rub impact,and the fault frequency of rotor rub impact is obtained.At the same time,the time-frequency analysis based on S-transform is carried out on the acoustic emission signal collected from the experiment,and the time-frequency diagram of rub impact signal is obtained,so as to obtain the local time-frequency characteristics of rub impact fault signal.At the same time,the time-frequency diagram of the signal before and after processing by MCKD method is compared,and the effectiveness of extracting rub impact fault characteristics by MCKD and time-frequency analysis method is verified.