Research on bearing fault diagnosis of reciprocating compressors based on GMDE and MFO-MKELM algorithms
ObjectiveAiming at the problem that the bearing fault feature extraction is difficult and the recognition accuracy is not high due to the characteristics of local strong non-stationarity and nonlinearity of the vibration signal of reciprocating compressor bearing clearance, a new method based on GMD...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | zho |
| Published: |
Editorial Office of Journal of Mechanical Transmission
2025-02-01
|
| Series: | Jixie chuandong |
| Subjects: | |
| Online Access: | http://www.jxcd.net.cn/thesisDetails#10.16578/j.issn.1004.2539.2025.02.022 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | ObjectiveAiming at the problem that the bearing fault feature extraction is difficult and the recognition accuracy is not high due to the characteristics of local strong non-stationarity and nonlinearity of the vibration signal of reciprocating compressor bearing clearance, a new method based on GMDE and MFO-MKELM algorithm was proposed.MethodsFirst, spread on multiscale entropy in the process of coarse graining, the average coarse graining way to a certain extent, “neutralize” the dynamics of the original signal mutation behavior, and reduce the accuracy of the entropy analysis. A generalized multiscale entropy algorithm spread, application of the reciprocating compressor vibration signals of bearing clearance for fault feature extraction was proposed; then, the polynomial kernel function and the improved Gaussian kernel function were linearly combined to construct the multiple kernel extreme learning machine intelligent recognition algorithm, and the fault diagnosis research was carried out on the extracted feature vector set.ResultsSimulation results show that the recognition accuracy of the fault diagnosis method is as high as 98.6%, which effectively realizes the intelligent diagnosis of different types of bearing faults. |
|---|---|
| ISSN: | 1004-2539 |