Research On Fault Diagnosis Of Rolling Bearings Based On Full Vector Spectrum Features

Rolling bearing is one of the most important parts of rotating machinery,and it is also the most easily broken mechanical component.Its running state is directly related to the safety of industrial production.Therefore,analyzing the vibration signals of rolling bearings,monitoring the running status of bearings,and carrying out fault diagnosis of rolling bearings have important theoretical research value and practical application value.The dynamic behavior of the rolling bearing has a non-linear nature.The vibration signal generated by it has obvious nonlinearity and non-stationarity,and the vibration signal is affected by strong noise,which greatly increases the difficulty of fault diagnosis of rolling bearings.Due to the low reliability of spectrum analysis methods based on single channel information,the spectrum amplitude is not obvious.Therefore,this paper focuses on noise reduction,feature extraction and fault identification of rolling bearing vibration signals.Solving the key problems and technical difficulties in signal processing and status recognition in the status monitoring of rotating machinery,and realize the monitoring and fault diagnosis of the running status of the rolling bearing.The study is as follows:(1)In order to solve the problem of low recognition reliability and unpredictable spectral amplitude caused by fault diagnosis using only single channel information,this paper proposes a feature extraction method combining EEMD and Full-Vector Envelope Spectrum.Firstly,EEMD is used to decompose the two-channel vibration signal,then the effective IMF component reconstruction signal is selected according to the kurtosis criterion.Finally,the Full-Vector Envelope Spectrum fusion dual channel reconstruction signal is used to complete the bearing fault feature extraction.The method can extract the feature frequency completely and accurately.The experimental results show that the proposed method has higher reliability than the single channel spectrum analysis method.(2)Choosing the IMF component according to the kurtosis will have the problem of uncertain component selection.In order to solve this problem,a feature extraction method combining the EEMD and Full-Vector Envelope Spectrum selected from multiple features is proposed.This method is on EEMD and Full-Vector Envelope Spectrum,Combining the kurtosis value with the RMS value and constructing it as a multi-feature index for selecting effective components.This method avoids the instability caused by the single index component selection method.This method effectively solves the problem of unstable IMF component selection.The experimental results show that this method has higher stability and reliability than the single index component selection method.(3)The vibration signal of the rolling bearing is affected by the strong noise,and it is difficult to extract the characteristics of the fault.In order to solve this problem,a feature extraction method combining the MCKD-EEMD and Full-Vector Envelope Spectrum is proposed.The method is based on EEMD and Full-Vector Envelope Spectrum selected from multiple features,MCKD as a prefilter of EEMD enhances the fault impact component in the vibration signal,reducing the impact of strong noise.The experimental results show that the proposed method extracts the fault features,and has superiority compared with the EEMD and the full vector envelope spectral methods based on multiple feature selection.This paper takes the rolling bearing in the rotating machinery as the research object,studies the problems of vibration signal noise reduction,feature extraction and fault identification,enriches the research content of the rolling bearing fault diagnosis theory,and has certain application value.