Skip to main content
Advertisement
Browse Subject Areas
?

Click through the PLOS taxonomy to find articles in your field.

For more information about PLOS Subject Areas, click here.

< Back to Article

Fig 1.

Architecture and training workflow of the Physics-Informed Self-Supervised Diagnosis (PI-SSD) framework.

More »

Fig 1 Expand

Table 1.

PI-SSD architecture, training configuration, and hyperparameter summary.

More »

Table 1 Expand

Table 2.

Summary of datasets used for training and evaluation.

More »

Table 2 Expand

Table 3.

Performance comparison on NASA Gearbox and Aalto Rotor Datasets (mean ± 95% CI).

More »

Table 3 Expand

Table 4.

Effect of hyperparameter variations on PI-SSD performance (mean ± 95% CI).

More »

Table 4 Expand

Table 5.

Cross-domain generalization results (Train on NASA, Test on Aalto Rotor, No Fine-Tuning).

More »

Table 5 Expand

Table 6.

Ablation study: Full Model vs. Component Variants (mean ± 95% CI).

More »

Table 6 Expand

Table 7.

Physics consistency breakdown by operating speed.

More »

Table 7 Expand

Table 8.

Calibration metrics on NASA Gearbox and Aalto Rotor datasets.

More »

Table 8 Expand

Fig 2.

Violin plots of physics residuals (PR-MSE) for healthy vs. faulty windows comparing Full PI-SSD and –Physics variant.

More »

Fig 2 Expand

Fig 3.

Training curves (loss, PR-MSE, AUROC) for benchmark models over 50 epochs.

PI-SSD consistently achieves the best balance between constraint satisfaction and classification.

More »

Fig 3 Expand