Diagnostic of erosive cavitation in hydraulic turbines from indirect measurements using SCADA data with uncertainty quantification

Keyphrases

• Acquisition Data 33%
• Bayesian Neural Network 33%
• Calibration Metrics 33%
• Cavitation 100%
• Cavitation Monitoring 33%
• Cavitation Rate 33%
• Condition Monitoring System 66%
• Data-driven Framework 33%
• Deep Ensemble 66%
• Deep Learning 33%
• Degradation Trajectory 33%
• Distribution Shift 33%
• Francis Turbine 33%
• Heteroscedastic Regression 33%
• Hydraulic Turbine 100%
• Indirect Measurement 100%
• Industrial Conditions 33%
• Instantaneous Prediction 33%
• Maintenance Decision 33%
• Measurement Approach 33%
• Model Transferability 33%
• Network Split 33%
• Operational Data 33%
• Operational Variables 33%
• Point Prediction 33%
• Real Industrial Data 33%
• Reliable Decision Making 33%
• SCADA Data 100%
• Split Conformal Prediction 33%
• Supervisory Control 33%
• Uncertainty Estimation 33%
• Uncertainty Quantification 100%
• Uncertainty-aware 33%

Engineering

• Accident Prevention 50%
• Bounded Uncertainty 50%
• Condition Monitoring System 100%
• Deep Learning Method 50%
• Francis Turbines 50%
• Hydraulic Turbines 100%
• Industrial Condition 50%
• Industrial Data 50%
• Metrics 50%
• Operational Data 50%
• Progression 50%
• Supervisory Control 50%
• Uncertainty Quantification 100%

Chemical Engineering

• Condition Monitoring 100%
• Deep Learning Method 50%
• Neural Network 50%
• Operational Variables 50%