As the sustaining element of rotor in motor

bearing fault owns great proportion in motor common faults.In allusion to the character of multi-component and AF/FM in motor rolling bearing vibrating signal

a process of motor bearing fault feature extraction based on local mean decomposition（LMD） and smoothed Teager energy operator is proposed.Multi-component AF-FM signal is decomposed into some single component AF-FM signal at first

then the component including main fault character is demodulated by means of smooth Teager energy operator

the fault feature of motor is analyzed accurately thereby.The effectiveness of the means is proved through simulation and example.