Analysis Of The Feature Of Motor Bearing Fault Based On Cyclostationary Theory

Three phase induction motor has simple structure,reliable operation and high efficiency,which is widely applied in the production and living and other fields.As an important part of the motor,bearing is easy to occur failure.So the early discovery of fault and taking measures is very important.Aiming at the size of the bearing fault damaged area,bearing dynamic structure and different bearing capacity of the ball in and out of the pit,this thesis put forward a model of double pulse torque ripple model and the characteristic frequency expression in stator current was deduced based on it.Then the current signal was separately carried on the cyclic autocorrelation function and spectral correlation density function analysis,it can be found that comparing with the traditional bearing fault detection method,the cyclostationary theory had the obvious superiority of extraction and recognition of fault characteristic frequency?reduction of fundamental frequency and harmonic noise.This thesis studied the relationship between the amplitude of fault characteristic frequency with the width of fault damage.Using cyclic autocorrelation function to analyze the amplitude variation of the first to third side frequency of fault characteristic frequency under different width of fault damage.Then,the spectral correlation density function analysis was carried out on the current signal,it can be seen that it had obvious spectral correlation in cyclic frequency domain and frequency domain,using it can identify the fault feature from different angles,not only reflected information of cyclic autocorrelation,also showed more fault characteristic information in spectral frequency domain,It provided more judgment in the diagnosis of bearing fault.Using the motor bearing fault experiment platform to collect the stator current signal under different width of fault damage in lab environment,and the current signal was carried on the cyclic autocorrelation function and spectral correlation density function analysis.Comparing the results of actual analysis with simulation,it verified the correctness of theory and method in this thesis,also showed the superiority of cyclostationary theory in recognition and extraction of bearing fault characteristic.