The Fault Diagnosis Of Rolling Bearing Based On Wavelet Analysis

As one of the most prone to occurring fault part of rotary machine, the rollingbearing state directly affects the operation of the entire system. In order to protect themachinery which is always in a safe and stable running state, and to avoid unnecessarymechanical accidents and economic losses and even casualties, it has great practicalsignificance to carry rolling bearing fault diagnosis.This paper has a research on fault diagnosis technology of N205EM rolling bearing,which can help people distinguish the type of fault. Through dedicated experimentalplatform (QPZZ-Ⅱsystem) of rolling bearings, which can provide the relevant stateinformation and the data of the state of fault., characteristic parameter of rolling bearingis confirmed and preprocessing of test data is completed. On the base of in-depth analysisof the test data and the characteristics of the data, there are two methods finishing faultidentification of rolling bearing in this paper. First of all, in order to solve the problem oftoo many data in the actual production, contrasted with traditional wavelet decompositioncombined with envelope analysis, this paper puts forward the fast wavelet decomposition.This method can complete fault diagnosis by observing the fault characteristic frequency.Secondly, in order to solve the problem of real-time nature in the actual production, themethod of recursive wavelet decomposition combined with SOM neural network hasbeen put forward. The neural network is used to classify, and the result is good.By simulation analysis, the results show that the fault diagnosis of N205EM rollingbearing achieves the good results and the expected goals. Not only can it solve theproblem of huge amount of data which can reduce the efficiency of the operation, but alsoit can solve the problem of the real-time.