1361. Fault diagnosis of gears based on local mean decomposition combing with kurtosis

Local Mean Decomposition (LMD) is a new self-adaptive time frequency analysis method. In present paper, the effectiveness of LMD method to extract fault features of gears, which are multi-component amplitude modulation (AM) and frequency modulation (FM), is demonstrated. A series of tests on tooth wearing, breaking and spalling gears are conducted and analyzed by LMD. And the fault features extracted by LMD are compared with those obtained from conventional Hilbert transform (HT). Moreover, the gear faults are identified by kurtosis based on LMD decomposed signals. The results demonstrate that the scheme combining LMD method with kurtosis analysis is effective to extract the characteristics of fault gears and improve the accuracy of fault diagnosis of gears.