Research On Motor Bearing Fault Detection Based On Electrical Characteristics Analysis

Induction motors have become indispensable electromechanical power equipment in industrial manufacturing due to their powerful driving,conduction,and traction capabilities.When implementing the corresponding power transmission process,they often cooperate with many parts and components to jointly assist the work of the motor.Rolling bearings,as a key mechanical component in almost all types of motor drive systems,bear the connection of the rotating part,and at the same time,they also serve 'as a supporting part,which needs to have sufficient capacity to bear the load.This also causes bearing failure to become one of the common causes of failure of electromechanical equipment.Accurately grasp the running status of the bearing and timely detect the fault and damage,which will help the safe operation of the motor and avoid accidents in production and life.Since most bearing applications are often subject to environmental disturbances such as noise and vibration,when the motor bearing fails,the weak fault feature is submerged in a strong noise background,and the electrical characteristic signal is an available signal that is less affected by noise.The stator current is an effective way to carry the fault characteristic signal in the electrical characteristics.Due to the vibration and eccentricity of the bearing during the fault,the fault characteristic in the current signal can be regarded as a non-periodic quasi-periodic cyclic stationary process.Through the analysis based on the cyclic autocorrelation function and the analysis based on the cyclic bispectrum,the cyclically stable characteristics of the fault frequency characteristics in the current signal are verified,and the manifestation of the stator current fault characteristics in the case of bearing faults is obtained,which provides a theoretical basis for the detection of fault characteristics.In view of the fact that the bearing fault features in the actual stator current signal are weak,the feature extraction and detection are difficult,and the efficiency of the cyclic bispectrum analysis is limited.By analyzing the influence of delay parameters on cyclic bispectrum analysis,a parameter optimized cyclic bispectrum analysis method is proposed to detect bearing fault characteristics.According to the physical meaning of the delay parameter,the accurate value of the delay parameter is calculated under the condition that the influence effect of the delay parameter is satisfied.Through simulation and experiment,the results of cyclic bispectrum analysis of different delay parameter values are compared to verify the accuracy of optimized parameter values and the effectiveness of cyclic bispectrum method based on parameter optimization to detect bearing fault characteristics.Based on the above theoretical analysis and experimental verification,the project built a motor drive system composed of FPGA and IPM.Then install the healthy and damaged bearings in the motor,collect the stator currents of different bearings,verify the cyclic stability of the fault characteristics,and verify the effectiveness of the parameter optimized cyclic bispectrum method to detect bearing failures.