Study On Fault Diagnosis Of Rolling Element Bearings Based On Resonance-based Sparse Decomposition

Rolling bearing is widely used in mechanical equipment,it is an important part of rotating equipment,and is also a major source of the fault,whether its working condition is normal or not directly affects the stability and safety of mechanical equipment.Therefore,in order to find fault in time and reduce the economic loss,rolling bearing condition monitor and fault diagnosis is of great significance.Based on the intensive study of the structure and vibration characteristics of rolling bearing,this paper studies the resonance-based sparse decomposition,and proposes its optimization method aiming at the parameter selection problem,which obtains better signal decomposition results.In addition,pattern recognition methods are studied and its optimization method is proposed,which can effectively recognizes the category of fault signal.The main contents in this paper are as follows:The research background and significance of rolling bearing fault diagnosis is elaborated,development process of fault diagnosis technology is summarized,research status of fault feature extraction and pattern recognition of rolling bearing is introduced systematically.The fault form and fault diagnosis method of rolling bearing are studied,the formula of fault characteristic frequency is obtained according to the structure and vibration mechanism of rolling bearing,and the basic steps of fault diagnosis of rolling bearing based on vibration signal are presented.The basic principle of resonance-based sparse decomposition method is deeply studied,aiming at the problem of parameter selection,the PSO algorithm is proposed to optimize the process of determining the quality factor.In order to improve the global optimization capabilities,the simulated annealing algorithm and the method of adjusting the inertia weighting factor are introduced to improve the PSO algorithm,which obtains the resonance-based sparse decomposition based on improved PSO algorithm.Different methods are used to decompose and spectrum the analog signal,which obtains the fault frequency,the validity of the proposed method is verified by the comparison of the decomposition results.The classification principle of support vector machine is studied,aiming at the limitation of support vector machine in dealing with large sample size problem,the least square support vector machine classification method is proposed,its parameters are optimized by the improved PSO algorithm.The Wine data is classified by the optimized classification method,which proves the effectiveness of the optimized classification method.The fault feature extraction method and the pattern recognition method proposed in this paper are experimentally verified by using the fault vibration signals of rolling bearing.Decompose rolling bearing fault signal with resonance-based sparse decomposition,on the one hand,extract fault characteristic frequency by analyzing spectrum of low resonance component;on the other hand,the coefficients corresponding to the low resonance components are used as the input of support vector machine for fault pattern recognition.The rolling bearing signal is decomposed by different methods,the comparison of the results of fault feature frequency extraction and pattern recognition classification show the superiority and robustness of the proposed method.