Research Article
KPCA and AE Based Local-Global Feature Extraction Method for Vibration Signals of Rotating Machinery
Table 6
Classification comparison of the rotor dataset.
| | Methods | Description | Dimension of feature vector | Testing accuracy (mean accuracy standard deviation) | Cost time (s) |
| | 1 (proposed) | SFS-KPCA-AE + KNN | 3 | 95.17% 2.87% | 3.63 | | 2 | ST-KPCA-AE + KNN | 3 | 47.00% 2.53% | 3.48 | | 3 | SFS-KPCA + KNN | 3 | 86.67% | 1.04 | | 4 | SFS-AE + KNN | 3 | 84.55% + 1.34% | 19.99 | | 5 | F20-KPCA + KNN | 3 | 87.78% | 1.33 | | 6 | F20-AE + KNN | 3 | 83.67% 4.40% | 3.87 | | 7 | EEMD-CC + KNN | 8 | 85.49% 4.38% | 28.31 |
|
|
