| Fatigue failure of bearings is a common cause of failure of marine rotating machinery.Bearing failure can be found in the early detection of bearing fatigue failure to avoid mechanical equipment accidents.The fatigue failure detection of marine bearings is affected by the vibration,shaking and environment factors,the collected vibration signals are more complicated.In this paper,the bearings used in the ship hoisting equipment are chosen as the experimental objects,the actual working condition of the bearing is simulated by the fatigue test machine.The Labview virtual machine is used to analyze the data collected before and after fatigue failure.The operating state of the bearing is determined by the temperature rise and the vibration acceleration signal during the operation of the bearing.When the temperature of the bearing rises obviously and reaches a certain range,we can judge the running state of the bearing may be unusual,At the same time,the peak factor and kurtosis value of vibration acceleration signal are analyzed to judge whether abnormal vibration impact occurs during the running of bearing.Labview software is used to carry out envelope demodulation of the vibration acceleration signal,through Comparison of the effects of Hilbert envelope method and extreme envelope method on vibration signal,finally found that carry on the Hilbert envelope to the data first,then carry on the extreme value envelope to get better effect of envelope.And the empirical mode decomposition(EMD)of vibration acceleration signal can observe waveform components in each frequency band,it is easier to extract the useful components of the signal,then according to the fault frequency of each part of the bearing,the specific parts of the fatigue damage of the bearing can be determined.In the empirical mode decomposition,a method of endpoint extreme value data extension is proposed.This method can effectively suppress the endpoint effect,the decomposed components are more regular.Finally,Hilbert Huang(HHT)transform is used on vibration acceleration signal,it can be seen from the transformation of the Hilbert spectrum whether there is a shock,as well as the frequency of impact and the impact time. |