Uncertainty quantification for data-driven motor drive temperature estimation models / by Shubham Gupta ; first examiner: Prof. Dr.-Ing. Jakub Kucka, second examiner: Prof. Dr.-Ing. Erdal Kayacan. Paderborn, 2026
Inhalt
- Introduction
- Fundamentals
- Thermal modeling of PMSMs
- Uncertainty Quantification
- Scoring Rules
- Gaussian Processes
- Bayesian regression with Gaussian prior
- Kernel Trick
- Variational Inference
- SVGP Approximation using VI
- Bayesian Neural Networks
- Deep Ensembles
- Evidential Deep Learning
- Uncertainty Calibration
- Uncertainty Evaluation Metrics
- Implementation
- Dataset
- Baseline TNN architecture
- Dropout
- Deep Ensembles
- Evidential Deep Learning
- Gaussian Processes
- HPO
- Uncertainty calibration
- Results
- Conclusion
- Appendix
- HPO plots for dropout
- HPO plots for deep ensembles
- HPO plots for EDL (NLL loss)
- HPO plots for EDL (CRPS loss)
- Artificial Intelligence use declaration
- Lists
- References