Method and system for performing fault diagnosis by bearing noise detection
The present disclosure provides a method and system for performing fault diagnosis by bearing noise detection. The method provides: collecting noise in bearing detection, the noise comprising bearing noise and operating condition noise; subjecting the collected noise to pre-processing, to obtain a first time domain signal and a second frequency domain signal; and inputting the first time domain signal and second frequency domain signal to a bearing fault diagnosis model. The bearing fault diagnosis model includes a characteristic extraction model and a fault discrimination model, the characteristic extraction model subjects the first time domain signal and second frequency domain signal to characteristic extraction separately to obtain a first characteristic associated with a time domain shock peak and a second characteristic associated with a fault frequency peak. Moreover, the fault discrimination model combines the first characteristic and second characteristic, and performs fault discrimination on the basis of a combined characteristic.
1. A method for performing fault diagnosis by bearing noise detection, comprising:
collecting noise in bearing detection, the noise comprising bearing noise and operating condition noise;
subjecting the collected noise to pre-processing, to obtain a first time domain signal and a second frequency domain signal; and
inputting the first time domain signal and second frequency domain signal to a bearing fault diagnosis model;
wherein the bearing fault diagnosis model provides a characteristic extraction model and a fault discrimination model, the characteristic extraction model subjects the first time domain signal and second frequency domain signal to characteristic extraction separately to obtain a first characteristic associated with a time domain shock peak and a second characteristic associated with a fault frequency peak; and the fault discrimination model combines the first characteristic and second characteristic, and performs fault discrimination on the basis of a combined characteristic.
2. The method according to claim 1 , wherein the characteristic extraction model provides a first sub-model for processing the first time domain signal to obtain the first characteristic, and a second sub-model for processing the second frequency domain signal to obtain the second characteristic.
3. The method according to claim 2 , wherein
based on the first time domain signal, first peak value data and first mean value data are extracted separately by means of the first sub-model, wherein the first peak value data represents a noise shock at a specific time in the time domain, and the first mean value data represents average operating condition noise in the time domain; and
based on the second frequency domain signal, second peak value data and second mean value data are extracted separately by means of the second sub-model, wherein the second peak value data represents a noise shock on a specific spectrum in the frequency domain, and the second mean value data represents average operating condition noise in the frequency domain.
4. The method according to claim 3 , further comprising:
combining the first peak value data and first mean value data, to obtain the first characteristic; and
combining the second peak value data and second mean value data, to obtain the second characteristic.
5. The method according to claim 1 , wherein the characteristic extraction model is a model based on a convolutional neural network (CNN) and the fault discrimination model is a model based on a fully connected network.
6. The method according to claim 1 , wherein the first time domain signal is a signal time domain envelope, and the second frequency domain signal is a signal spectrum envelope.
7. The method according to claim 1 , further comprising establishing a fault mode data set on the basis of a historical fault mode, and storing a discriminated fault mode in order to update the fault mode data set.
8. The method according to claim 1 , wherein the pre-processing comprises:
subjecting the collected noise to bandpass filtering;
acquiring time domain waveform data of a filtered signal to serve as the first time domain signal;
subjecting the filtered signal to a Fourier transform, and acquiring envelope spectrum data of the Fourier transformed signal to serve as the second frequency domain signal;
normalizing the first time domain signal and second frequency domain signal separately; and
resampling the normalized first time domain signal and second frequency domain signal.
9. A system for performing fault diagnosis by bearing noise detection, comprising:
a data collector, configured to collect noise in bearing detection, the noise comprising bearing noise and operating condition noise;
a processor connected to the data collector; the processor being configured to:
subject the collected noise to pre-processing, to obtain a first time domain signal and a second frequency domain signal; and
input the first time domain signal and second frequency domain signal to a bearing fault diagnosis model;
wherein the bearing fault diagnosis model comprises a characteristic extraction model and a fault discrimination model, the characteristic extraction model subjects the first time domain signal and second frequency domain signal to characteristic extraction separately to obtain a first characteristic associated with a time domain shock peak and a second characteristic associated with a fault frequency peak; and the fault discrimination model combines the first characteristic and second characteristic, and performs fault discrimination on the basis of a combined characteristic; and
a memory, configured to be connected to the processor and store a discriminated fault mode in order to update a fault mode data set.
10. A non-transitory computer readable storage medium comprising:
instructions executed by a computer, and
a system for performing fault diagnosis by bearing noise detection by the instructions executed by the computer, the system providing:
a data collector, configured to collect noise in bearing detection, the noise comprising bearing noise and operating condition noise;
a processor connected to the data collector; the processor being configured to:
subject the collected noise to pre-processing, to obtain a first time domain signal and a second frequency domain signal; and
input the first time domain signal and second frequency domain signal to a bearing fault diagnosis model;
wherein the bearing fault diagnosis model comprises a characteristic extraction model and a fault discrimination model, the characteristic extraction model subjects the first time domain signal and second frequency domain signal to characteristic extraction separately to obtain a first characteristic associated with a time domain shock peak and a second characteristic associated with a fault frequency peak; and the fault discrimination model combines the first characteristic and second characteristic, and performs fault discrimination on the basis of a combined characteristic; and
a memory, configured to be connected to the processor and store a discriminated fault mode in order to update a fault mode data set.