Research On Condition Monitoring Of Rolling Element Bearing Based On Information Scale Spectrum

As the core component of rotating machinery,the rolling element bearing(REB)plays a key role in the machine.Timely and effective monitoring of REB condition is an important method to ensure operation of the industry.The working environment of REB is complex and harsh.Weak fault is the main early failure form.This will produce weak fault feature in the vibration signal.However,the traditional method is not effective for weak fault monitoring in complex noise.In view of the above problems,this paper systematically research the condition monitoring method of REB in complex environment.The main research work and innovations are as follows:In order to solve the problem of the lack of early warning thresholds for health indicators in bearing degradation monitoring,an early warning threshold for the sum of weighted normalized square envelope spectrum(SWNSES)based on the statistical threshold of squared envelope spectrum under colored noise is proposed.The influence of REB fault component on the statistical threshold of squared envelope spectrum under colored noise is analyzed.The false peak caused by the fault components in the statistical threshold are effectively eliminated by removing the spectrum maxima.The early warning threshold of SWNSES is constructed through the improved statistical threshold.The proposed early warning threshold is effectively verified by the IMS system REB run-to-failure experimental platform.The proposed early warning threshold successfully detects the initial stage of the fault,and has strong adaptability and generalization ability.In order to solve the problem of REB fault feature attenuation caused by noise and transmission path,a novel fault feature enhancement method spectrum weighted deconvolution is proposed.The objective criterion is the sum of the weighted normalized envelope spectrum to enhance cyclostationarity in the signal.A constrained quadratic maximization model is established.The trust-region self-consistent-field like iteration algorithm is used to update the filter,which avoids the iteration from falling into local optimization.The proposed method is verified by various conditions simulation and experimental data.The results show that compared with other deconvolution filtering methods,the proposed method effectively enhances the cyclostationarity of fault signals and has stronger diagnostic performance under strong noise and long transmission path.To solve the problems that envelope spectrum-based method in bearing fault diagnosis has heavy attenuation of fault harmonic,and is interfered by strong rotational frequency.,which reduces the reliability of the the fault degree judgment,a novel bearing fault diagnosis method called information scale spectrum is proposed in the paper.First,the vibration signal is segmented,and the fluctuation feature is extracted through the feature operator.Then the binary threshold processing is used to classify the feature frame signal and extract the fault interval.The fault interval is compressed and transformed into feature pulse to realize fault signal energy enhancement.Simulation and experimental data are used to verify the proposed method.The results show that: compared with the envelope spectrum-based method,the proposed method solves the shortcomings of heavy attenuation of fault harmonic in the spectrum,eliminates the interference of rotational frequency,and has a stronger diagnostic performance in low signal-to-noise ratio.To sum up,aiming at the weak fault diagnosis of rolling bearing in complex environment,this paper puts forward a series of methods such as degradation monitoring and early warning,fault feature enhancement and fault diagnosis.The effectiveness of the methods is verified by experiments.