Study On Fault Diagnosis Methods For Rolling Bearing Based On SVM-HMM

In the process of production,bearing is almost indispensable in the use of mechanical equipment.The corresponding equipment damage and production damage caused by faulty bearings are emerging in an endless stream.,resulting in huge economic losses.But the factors are various which lead to the bearings faults,so it is particularly impotent to accurately diagnose the faults of the bearing.In numerous rotating mechanical failure about one third is resulted from rolling bearing fault.Therefore,to handle the working status and the fault formation as well as development of rolling bearing is currently important topic in the area of mechanical fault diagnosis.So,by using theoretical derivation and numerical simulation analysis and experiment research,the fault diagnosis of rolling bearing were studied in this paper.In Chapter 1,the background and significance of the research were introduced,the failure modes and the research status and development trend of fault diagnosis for rolling bearing as well as the main methods to acquire the fault information were described in detail.Then,the main research contents of this dissertation were put forward.In Chapter 2,The structure and the basic classification of the rolling bearing and were introduced.Combined with the analysis of the mechanism kinematics and vibration of rolling bearings as well as the fault characteristic frequency.Finally,the research status of feature extraction and fault diagnosis of rolling bearing were analyzed.In Chapter 3,Introducing the empirical mode decomposition algorithm in the process of rolling bearing vibration signal decomposition,consider the aliasing problem,the idea of Complementary ensemble empirical mode decomposition and autoregressive model was put out in the feature extraction of rolling bearing fault diagnosis.further research was made on its basic theory and implementation process.In Chapter 4,based on the study in the principle and algorithm of Hidden Markov Model and Support Vector Machine,an HMM-SVM based method for rolling fault diagnosis was proposed.Finally,completed the construction of the hybrid model.In Chapter 5,Combined with the research of feature extraction and pattren recognition in the previous chapter.the simulation experiments had been carried out to verity the proposed scheme.In Chapter 6,the main research work of this dissertation was summarized and prospected.