Research On Fault Diagnosis Method Of Roller Bearing Fusion Of Multi-source Features

With the continuous improvement of the degree of automation and precision, the reliability and stability of the equipment is more important for production. Because of the complexity of the printing machine in ink,a single access bearing is very complex,removing too much trouble,frequent bearing fault has led to a lot of printing waste, so the fault diagnosis of the printing machine is very necessary. The main contents of this thesis are as follows:(1) This paper introduces the research status of fault diagnosis of the printing machine, and analyses the existing methods of three kinds of entropy: the advantages and disadvantages of sample entropy, and information entropy and bispectrum entropy, and carries on the comparison,according to the characteristics of fault diagnosis of the existing printing machine, the entropy analysis method combined with Hidden Markov Model. It is applied to the fault diagnosis of roller bearings in printing machine.(2) This paper presents a new method for fault diagnosis of bearing based on multi feature fusion. Most of the existing methods for a single sound fault signal or fault vibration signals for fault diagnosis, but the way that information collection is not complete, the fusion of sound and vibration signal of fault fault signal, and considering the influence factors, so the method of multi feature fusion diagnosis can improve the fault diagnosis rate.(3) The fusion signal of the three fault types and three kinds of bearing fault degree and bearing fault diagnosis are control variables, analysis of roller speed on the printing machine roller bearing diagnosis rate, comparing sound fault signal and vibration signal in fault diagnosis of roller bearing rate.Through the research of this topic, the paper provides a new method for the fault diagnosis of the bearing of the multi feature fusion, and provides a new way for the fault diagnosis of the printing machine.