KIT | KIT-Bibliothek | Impressum | Datenschutz

Machine-learning-based Bayesian state estimation in electrical energy systems

Jongh, Steven de; Mueller, Felicitas; Li, Hongfei; Georgieva, Boyana; Suriyah, Michael; Leibfried, Thomas

Abstract:

In many algorithmic applications in electrical power grids, state estimation (SE) represents the first step of a
process chain. In SE, sensor measurements are processed to infer the most probable grid state. Classical methods
such as weighted least squares (WLSs) based approaches use statistical methods that can be based on sensor noise
and erroneous measurements. With these methods, only point estimates are made, which results in a lack of

knowledge about prediction uncertainties. In this study, machine-learning-based methods for determining the actual
state of the grid are proposed. Bayesian optimisation is applied to find the optimal hyperparameter configurations for

neural networks (NNs) for SE tasks. The application of Bayesian inference using Bayesian NNs is proposed, which
allows the prediction of point estimates as well as uncertainty intervals for the system states. The advantages of using
Bayesian approaches in comparison to classical SE methods like WLS are shown.


Verlagsausgabe §
DOI: 10.5445/IR/1000138913
Veröffentlicht am 13.10.2021
Originalveröffentlichung
DOI: 10.1049/oap-cired.2021.0053
Dimensions
Zitationen: 1
Cover der Publikation
Zugehörige Institution(en) am KIT Institut für Elektroenergiesysteme und Hochspannungstechnik (IEH)
Publikationstyp Proceedingsbeitrag
Publikationsdatum 23.04.2021
Sprache Englisch
Identifikator ISSN: 2515-0855
KITopen-ID: 1000138913
Erschienen in CIRED 2020 Berlin Workshop (CIRED 2020), 22-23 September 2020
Veranstaltung International Conference and Exhibition on Electricity Distribution (CIRED 2020), Online, 22.09.2020 – 23.09.2020
Verlag Institution of Engineering and Technology (IET)
Seiten 341-344
Serie CIRED - Open Access Proceedings Journal ; 2020
Nachgewiesen in Scopus
Dimensions
KIT – Die Forschungsuniversität in der Helmholtz-Gemeinschaft
KITopen Landing Page