TY  - CONF
AU  - Buechel, Maximilian
AU  - Zimmer, Marcel
AU  - Carta, Daniele
AU  - Benigni, Andrea
TI  - Distribution Grid Voltage Estimation by Correlated Gaussian Processes
PB  - IEEE
M1  - FZJ-2025-05278
SP  - 1-6
PY  - 2025
AB  - Modern power distribution grids are facing significant monitoring challenges from renewable energy volatility and persistent measurement scarcity. In this paper, we propose a data-driven approach for voltage magnitude estimation in distribution systems using correlated Gaussian processes. We demonstrate that the proposed approach can be applied to a variety of monitoring configurations. Comparable accuracies are achieved even when the number of monitoring units is significantly reduced. Moreover, the results are obtained with the help of small training sets, allowing for short observation periods to generate training data. The proposed approach offers flexibility in terms of monitoring configurations and provides reliable voltage estimation in poorly monitored distribution networks.
T2  - 2025 IEEE 15th International Workshop on Applied Measurements for Power Systems (AMPS)
CY  - 24 Sep 2025 - 26 Sep 2025, Bucharest (Romania)
Y2  - 24 Sep 2025 - 26 Sep 2025
M2  - Bucharest, Romania
LB  - PUB:(DE-HGF)8
DO  - DOI:10.1109/AMPS66841.2025.11219969
UR  - https://juser.fz-juelich.de/record/1049196
ER  -