Research On Fault Diagnosis Based On Acoustic Emission Signal Of Rolling Bearings

Rolling bearings are key components of rotating machinery and equipment.Once they fail,they will not only affect production,but also bring immeasurable economic losses to enterprises,and may even cause personal casualties in severe cases.Traditional vibration signal analysis methods cannot effectively detect and distinguish early bearing faults,and it is difficult to effectively analyze the severity and development trend of bearing faults.This paper mainly takes the rolling bearing as the research object,and analyzes and studies the monitoring method of the bearing acoustic emission signal.The main research contents are as follows:This paper took rolling bearing of fault simulation test bench as the research object,the acoustic emission detection technology was used,the acoustic emission data acquisition platform was built,the acoustic emission detection and acquisition system explained and the acquisition plan was formulated.The time domain analysis,spectrum analysis and parameter analysis were carried out on the acoustic emission signal of the bearing,and certain results had been obtained.Aimed at the difficult problem of acoustic emission signal processing,an acoustic emission signal of rolling bearing based on optimized Maximum Correlation Kurtosis Deconvolution(MCKD)and Variational Mode Decomposition(VMD)combined with energy entropy was used as fault feature extraction method.Aimed at the parameter selection problem of MCKD algorithm,the kurtosis was used as the objective function of optimization,and the method of grid search with variable step size was used to optimize the filter length L in the MCKD algorithm.The VMD method was used to decompose the signal,and the energy entropy screening component was used to reconstruct the signal and then make the envelope spectrum,so as to realize the fault feature extraction of the acoustic emission signal.Through the analysis of the rolling bearing test bench data,the effectiveness of the proposed method was verified.Aimed at the acoustic emission signal of bearing,a fault identification method based on Complete Ensemble Empirical Mode Decomposition with Adaptive Noise,(CEEMDAN),Support Vector Machine(SVM)and combined with Permutation Entropy(PE)were used.Firstly,the collected acoustic emission signal was decomposed by the CEEMDAN method,and then the permutation entropy value of each component obtained by the decomposition was calculated,and the component with the permutation entropy greater than 0.6 was taken to form a eigenvector and input into the SVM model for training;compared with the set empirical mode decomposition(Ensemble Empirical Mode Decomposition,EEMD)and Empirical Mode Decomposition(Empirical Mode Decomposition,EMD)methods after SVM training,and comparing the SVM classification effect of the feature vector composed of a single component,the classification accuracy of the method proposed in this paper reaches 98.89%,which proves the effectiveness and advantages of the method in this paper.