Research On Weak Fault Diagnosis Method Of Rolling Bearing Based On Time-frequency Analysis Under Noise Background

As a key component of the rotating machinery equipment,the rolling bearings directly affect the safe operation of the equipment,and the rolling bearing is one of the components with the highest frequency of failures.Therefore,the study of rolling bearing failures is of great significance.The weak fault signal of rolling bearing has the characteristics of low signal-to-noise ratio and non-stationary nonlinearity,which is a difficult point in the study of rolling bearing faults.In order to realize the diagnosis of weak faults of rolling bearings as soon as possible,this paper studies the weak fault diagnosis of the rolling bearing based on time-frequency analysis under the background of noise.The main research contents are as follows.Firstly,for the problem that weak fault signals are greatly affected by noise and are not easy to be found,the noise reduction technology based on singular value decomposition(SVD)algorithm is studied.The noise reduction principle based on SVD algorithm is analyzed,and an improved SVD algorithm is proposed.The method is used to improve the signal-to-noise ratio of the fault signal and reduce noise interference.The algorithm uses the average decline rate speed of the singular value as the noise reduction order index,calculates the difference of the singular value accumulation,and uses the principle that the difference is the largest at the critical point between the effective singular value and the noise singular value to determine the effective singular value.The comparison experiment of singular value accumulation method,singular value difference spectrum and singular value curvature spectrum proves that this method can effectively reduce the interference of noise to the signal.Secondly,a method based on singular value accumulation method noise reduction and fast spectral kurtosis algorithm for diagnosing weak faults of rolling bearings is proposed.Fast spectrum kurtosis,as a method for determining the optimal frequency band of a filter,is an important method for studying mechanical fault signals.This method determines the center frequency and bandwidth of the fault characteristic signal by finding the frequency band corresponding to the maximum kurtosis value.The noise reduction performance of the singular value accumulation method can reduce the interference of noise on the maximum kurtosis value and improve the accuracy of the fast spectral kurtosis.The optimal frequency band is used to determine the parameters of the band-pass filter,the frequency band of the fault frequency is extracted from the mechanical vibration signal,and the envelope demodulation analysis is performed to obtain the characteristic frequency of the fault,so as to realize the diagnosis of the weak fault of the rolling bearing.The characteristic frequency corresponding to the fault signal is accurately obtained from the experimental results of the measured signal,which shows the feasibility of the diagnosis method in weak fault diagnosis.Finally,in view of the non-stationary and non-linear characteristics of the rolling bearing fault signal and the low signal-to-noise ratio of the weak fault signal,a timefrequency analysis method based on the idea of synchroextracting transform is proposed,which is the general synchroextracting chirplet transform.The algorithm introduces SET to post-process the chirplet transform results,extracts the instantaneous frequency trajectory most relevant to the characteristics of the fault signal,and improves the time-frequency readability and noise robustness.At the same time,the algorithm supports the reconstruction of signal components to obtain the time-domain characteristics of fault signal components.Experiments on actual and simulated signals have proved that the algorithm can handle nonstationary and non-linear signals with emphasis.Effective in diagnosing weak faults of rolling bearings.