Model Identification and Parameter Tuning of Dynamic Loads in Power Distribution Grid: Digital Twin Approach
Nils Huxoll; Mohannad Aldebs; Payam Teimourzadeh Baboli; Sebastian Lehnhoff; Davood Babazadeh
2021 International Conference on Smart Energy Systems and Technologies (SEST)
With the ongoing changes in power systems, not only on the generation side but also on the load side, new approaches are necessary to monitor and control power systems. Therefore, this paper investigates the Digital Twin technology for power system loads with a novel parameter identification method based on Bayesian Inference. A framework for load model Digital Twins is proposed based on an existing model, and a novel approach to load model identification is investigated and compared to existing methods. Even though Bayesian Inference relies on prior knowledge of the model, compared to other approaches, it returns a Probability Density Function for the whole model and each model parameter and fares very well with sparse data as well as an increased level of measurement noise. The results promise to use Bayesian Inference as the primary identification method for a Digital Twin as proposed in this paper. This Digital Twin framework can be utilized to overcome new challenges arising for power system control and monitoring.
September / 2021
inproceedings
International Conference on Smart Energy Systems and Technologies, SEST 2021
1-6
ENSURE Phase 2
Neue EnergieNetzStruktURen für die Energiewende (Phase 2)
Neue EnergieNetzStruktURen für die Energiewende (Phase 2)
