Fault Diagnosis Technology Of Motor Bearing Based On Electrical Method

The electrical method can achieve non intrusive diagnosis of motor bearing failures and reduce the economic cost of additional installation sensors,which is of great significance.Different diagnostic results can be obtained for different signal carriers.The signal carrier used mainly is current,torque and speed signals.Considering the high accuracy of the encoder in the permanent magnet motor,the speed signal can get higher fault diagnosis results.In this paper,a double inertia model for the integration of motor and motor bearings is established.The theoretical derivation of electrical method based on the double inertia model in MATLAB/Simulink is verified by simulation,and the validity of the theoretical derivation is proved.After the failure of the FFT experiment,an adaptive resonance demodulation method,fast spectral kurtosis,is used to diagnose the fault.In the case of different experimental conditions and rotational speed,the diagnosis effect of three different signal carriers is compared and analyzed.And the better adaptability of the speed signal to the embedded system is verified.It can get relatively better result of fault diagnosis.In order to improve the effect of fault diagnosis,a method of signal reconstruction with ensemble empirical node decomposition combined with intercorrelation coefficient is adopted to improve the signal to noise ratio of the signal to a certain extent,and the range of fault diagnosis is expanded.In the process of signal reconstruction,the influence of a large number of noise is filtered,and the correlation with the fault signal is strengthened,and the reconfigurable signal which is easier to diagnose the fault can be obtained.After that,the fast spectral kurtosis map will be improved by the shortcoming of high peak pulse interference,and an improved method for the average of the seed band spectrum kurtosis is obtained,and the combination of multiple crest frequency and bandwidth with high kurtosis is used to diagnose.First of all,this method is used to verify the improvement of the fault diagnosis.Then the method of signal reconstruction with EEMD and intercorrelation coefficient is combined to continue to broaden the range of fault diagnosis.At last,the fault diagnosis failure under the condition of drag loading is analyzed.It is inferred that the fault information is drowning due to the instability or eccentricity of the loading.And with the increase of the drag load,the unstable signal will get a greater increase of the fault than the fault impact and lead to further reduction of the diagnosis range.The load contrast test of magnetic particle brake is carried out,which can basically verify the previous analysis results.It is difficult todiagnose electrical method of motor bearing in the case of drag and load under the method of digital signal processing.It is necessary to use the method of combining with motor drive technology to collect signal for diagnosis under the condition of restraining these fluctuations.