Intelligent fault diagnosis and localization of hydropower station equipment using multi-source data fusion and deep residual networks
The results indicate the potential value of integrating physical residuals, simulation-assisted augmentation, and modality-aware representation learning for rotating-equipment diagnosis, while plant-specific digital-twin calibration, multi-sensor hydropower validation, quantitative domain-discrepancy analysis, and edge...