Research And Application Of Fault Diagnosis Method For Rolling Bearing Of Mine Main Fan

As the key component supporting the operation of the mine main fan,the rolling bearing has a complex working environment and is easy to cause fault.In the early stage of the failure of the rolling bearing of the mine main fan,the fault feature of the rolling bearing is weak,exist interference component and is difficult to extract.In order to ensure the safe operation of main fan,the rolling bearing of the axial-flow mine main fan is taken as the research object,and its failure mechanism,feature extraction and pattern recognition method are studied.Finally,the condition monitoring and fault diagnosis system of the rolling bearing of the mine main fan is established.The main research contents and conclusions are as follows:According to the fault type and vibration mechanism of the rolling bearing of the mine main fan,the vibration form caused by internal factor such as rolling bearing structure,machining and assembly error and fault is analyzed.The surface damage fault provokes a series of periodic impact component.The single damage point theoretical model of the outer ring,the inner ring and the rolling element are established and its vibration form are analyzed.Each model produce a series of periodic impact component with different amplitude and period.Aiming at the problem of weak and difficult extraction of fault feature for the rolling bearing of the mine main fan in the initial stage,the resonance sparse decomposition and spectral kurtosis method are applied to the fault feature extraction of rolling bearing respectively.It is found through simulation analysis that the resonance sparse decomposition can extract the fault feature of rolling bearing,and the frequency band of the periodic impact component can be accurately located by the spectral kurtosis.However,the periodic impact component characterizing fault feature,which is obtained by using the resonance sparse decomposition and the spectral kurtosis respectively,exist a certain interference component and is not prominent in the early stage.Aiming at the above problem,a fault feature extraction method based on resonance sparse decomposition and spectral kurtosis is proposed.It can eliminate the interference component in the low resonance component and highlight the periodic component through simulation analysis.Finally,it is proved to be suitable for the extraction of fault feature of the rolling bearing of the mine main fan through experiment.It can extract the periodic impact component annihilating in the interference component,filter out interfering component and highlight fault feature.The feature vector characterizing the rolling bearing fault state is composed of parameters such as kurtosis,pulse factor,margin factor,sample entropy and permutation entropy of the periodic impact component.It can be used as the input of the support vector machine to realize the fault recognition of the rolling bearing of the mine main fan.It is found through experiment that the classification accuracy of the rolling bearing fault support vector machine classifier of the mine main fan based on random selection parameter of(C,σ)is greatly affected,and it is difficult to find the optimal parameter combination.In order to improve fault recognition rate of the rolling bearing of the mine main fan,the PSO-SVM is applied to the fault diagnosis of the rolling bearing of the mine main fan.It can achieve global optimization through PSO optimizing parameter of(C,σ).It is proved by experiment that the rolling bearing fault PSO-SVM classifier of the mine main fan is simple to implement,fast convergence speed and high classification accuracy.It can correctly identify the normal,inner ring fault,outer ring fault and rolling element fault of the rolling bearing of the mine main fan.In order to apply the resonance sparse decomposition and spectral kurtosis and PSO-SVM to the fault diagnosis of the rolling bearing of the mine main fan,LabVIEW is used as the development platform to design the condition monitoring and fault diagnosis system of the rolling bearing of the mine main fan.It realize the signal acquisition,storage,query,frequency domain analysis,feature extraction and fault diagnosis.The experiment result indicate that the system has good stability,easy operation and high fault recognition rate.It is integrated into the FAMC-2 fan centralized console and the operation result show that the system is reliable,high sensitivity and accuracy.It provides an important guarantee for the safe and efficient operation of the mine main fan.