Research On Motor Bearing Fault Diagnosis Based On Time-Frequency Analysis

Whether the motor equipment can be operated efficiently and with low power consumption is closely related to the working condition of the bearing or gear transmission system.Early detection of abnormal bearings can replace the faulty parts in advance and end the abnormal state of long-term working conditions.This paper attempts to identify the early weak feature fault state of the bearing,and conducts research in the aspects of stator current power supply noise suppression,environmental noise suppression,fault feature extraction,and time-frequency feature enhancement of fault features.This paper first analyzes the fault characteristics of single-point damaged bearings,demonstrates the principle of fault characteristics from the aspects of motor air gap change and torque fluctuation,and further modifies the existing torque single pulse model according to the calculation results of the dynamic model to The torque double-pulse model describes the fault form more accurately,and further analyzes the effect of bearing faults on the stator current through mechanical-magnetic-electrical coupling conduction.In this paper,the time-frequency analysis method—Adaptive Noise Complete Empirical Mode Decomposition(CEEMDAN)is used to process the non-linear and non-stationary stator currents,and the effective information is extracted from the noise reduction filtering and recombination of the current signal according to the kurtosis value and the required ’signal frequency band;Adopt time-shifting method to suppress power fundamental frequency and odd multiple frequency energy,reduce the frequency leakage of power frequency component and mask the fault characteristic phenomenon;explain the improved principle of multi-resolution energy operator relative to energy operator,and improve its resolution according to the fault signal The self-adaptive parameter,the adaptive multi-resolution energy operator demodulate the fault signal is more targeted than the basic energy operator demodulation effect,adaptively enhance the fault signal spectrum characteristics.The effectiveness of the proposed method is verified by simulation data.The laboratory builds a bearing fault experiment platform to reduce the motor speed and enhance the fault modulation characteristics and recognizability.The test method for bearing fault stator current diagnosis verified by the test data is more effective than other methods.The abnormal state of the bearing can be clearly found in the map and finally used Labview and Matlab software have developed a bearing fault detection system.The system realizes online/remote detection of bearing fault functions,which can clearly detect the existence of faults and achieve good diagnostic results.