Research On Rotating Machinery Fault Diagnosis Based On Vibration Signal Processing And SVM

With the development of science and technology and the continuous expansion of the scale of enterprises,the use of rotating machinery in industrial production is more and more extensive,and the role played by it is becoming more and more important.Nowadays,the automation and complexity of rotating machinery are increasing,and the attendant failure rate is gradually increasing.In many cases,once the rotating machinery fails,the company will face large property losses and casualties.Therefore,it is very necessary to monitor the operating state of the rotating machine and to perform an effective fault diagnosis.This has a very high practical significance for ensuring safe and smooth operation of the industrial site.This paper mainly studies the vibration signal based on rotating machinery and Support Vector Machine(SVM)to achieve effective fault diagnosis.Through the time-frequency domain analysis of the vibration signal,the effective feature vector is extracted,and then the support vector machine is used as the classifier to complete the fault identification.In this paper,the common faults in the rotating mechanical system,including: rotor imbalance,rotor misalignment,rotor crack,oil film whirl and oil film oscillation,friction and mechanical looseness,etc.,are analyzed and introduced in detail,revealing the causes and characteristics of these faults in depth.In terms of signal processing,the vibration signal is analyzed by the Empirical Wavelet Transform(EWT)method.Aiming at the problem of excessive band boundary points and band rupture in the classical scale space method,it is proposed to use frequency energy to filter the band boundary points,which effectively avoids the band rupture phenomenon and improves the quality of signal reconstruction of the empirical wavelet transform method.On this basis,the kurtosis,crest factor,margin factor,pulse factor,singular value entropy,energy entropy,permutation entropy and sample entropy are composed of 8 indicators to form the feature vector as the input of the support vector machine classifier.When training the support vector machine,the K-Cross Validation(K-CV)method is used to optimize the parameters of the support vector machine,which effectively improves the reliability and stability of the algorithm.Finally,based on Lab VIEW software,a host computer fault diagnosis system was developed.The program and requirements analysis of the system are introduced in detail,the software function module is designed reasonably,and the methods of experience wavelet transform and vector machine are integrated into it.The test results show that the software system can realize the fault diagnosis of the rotating machinery,the diagnosis result is accurate,and the operation is convenient and concise.