Research On The Method Of Local Mean Decomposition In The Fault Diagnosis Of Gearbox

Based on the fault diagnosis of gearbox as background, the local mean decomposition (LMD) researched in the paper is a newly self-adaptive time-frequency analysis method in recent years. It can apply to non-stationary signal and can get significative Instantaneous frequency, which provides a new method for the diagnosis of mechanical equipment.This paper researched the basic theory and realizing process of LMD. According to the Actual application problems of LMD, the paper proposed the improved method of LMD based on three cubic interpolations to solve the problems of long calculating time and non-convergence brought by improper smooth span during sliding smooth. Then the paper analyzed the end effect of LMD, and tried the method of extension of extreme point to solve endpoint effect.The paper applied the method of improved LMD to the field of fault diagnosis of gearbox, and the type of JZQ250 gearbox was made as test object. Based on the fault mechanism and vibration trait of the gearbox, the vibration signals of the gearbox were tested and the method of improved LMD was applied to analyze the signal and to diagnose the fault of snaggletooth, bearing cup and bearing cone, and got a good diagnosis effect. The test results proved the availability of LMD in gearbox fault diagnosis field.