Abstract | ||
---|---|---|
In order to improve the efficiency and accuracy of bearing fault diagnosis, fixed threshold wavelet transform and extreme learning machine (ELM) are used to diagnose the fault data set. Firstly, the original signal underwent wavelet noise reduction by fixed threshold and heuristic threshold method, comparing SNR and mean square error, the processed signal was extracted, select cliff, margin factor, waveform factor, pulse factor, variance, mean, maximum and minimum 8 features, and the values were input into ELM for training and testing, and adjust the number of ELM neurons to check the fault identification accuracy, and compared with other diagnostic methods. The simulation results show that the proposed method provides a new idea for solving the bearing fault diagnosis problems. |
Year | DOI | Venue |
---|---|---|
2022 | 10.1109/DSA56465.2022.00068 | 2022 9th International Conference on Dependable Systems and Their Applications (DSA) |
Keywords | DocType | ISSN |
fault diagnosis,feature extraction,fixed threshold,wavelet transform,Extreme learning machine | Conference | 2767-6676 |
ISBN | Citations | PageRank |
978-1-6654-5984-6 | 0 | 0.34 |
References | Authors | |
0 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Zhen Zhao | 1 | 0 | 0.34 |
Jingchao Li | 2 | 0 | 0.34 |
Bo Deng | 3 | 0 | 0.68 |
Yulong Ying | 4 | 2 | 1.08 |