Research Article
Simultaneous-Fault Diagnosis of Automotive Engine Ignition Systems Using Prior Domain Knowledge and Relevance Vector Machine
Table 10
Evaluation results of the binarization approach and the proposed framework for fault diagnosis.
| | Binarization | Proposed framework |
| Feature Extraction | None | WPT + PCA + DK | WPT + PCA + DK | Classification | SVM | SVM | PCRVM | threshold | None | None | PSO-optimized threshold = 0.7147 | -measure over | | | | Overall cases | 0.4518 | 0.6792 | 0.8891 | Single fault cases | 0.4500 | 0.7404 | 0.9567 | Simultaneous fault cases | 0.4528 | 0.6455 | 0.7714 |
|
|