PSFCL: A Probabilistic Slow Feature Contrastive Learning approach for incipient fault diagnosis in industrial processes

Puntos clave

• The method identifies incipient faults with high precision, advancing diagnostic capabilities for industrial applications.
• Key evidence includes a fault detection rate exceeding 90% across diverse process datasets, indicating robustness.
• Assessment using probabilistic slow feature contrastive learning demonstrates improved model performance over traditional techniques.
• Highlights the need for further validation in real-world settings to optimize fault detection strategies.