Physics-Informed Neural Network for Modeling the Dynamic Behavior of Grid-Forming Converters

arXiv (Cornell University)

Abstract

This paper investigates physics-informed neural networks for modeling the full dynamic behavior of droop-controlled grid-forming converters. The approach is trained on synthetic data generated via numerical solvers and benchmarked against both traditional integration methods and a vanilla neural network. Results show higher predictive accuracy than the vanilla network using the same training data and substantially reduced runtime compared with numerical solvers.

Authors 5

1. Hussein Jaffal Aachen Institute of High Voltage Equipment and Grids

   RWTH Aachen University

   Affiliation as printed

   Institute of High Voltage Equipment and Grids, Digitalization and Energy Economics , RWTH Aachen University , Aachen , Germany

2. Arianna Fois Aachen Institute of High Voltage Equipment and Grids

   RWTH Aachen University

   Affiliation as printed

   Institute of High Voltage Equipment and Grids, Digitalization and Energy Economics , RWTH Aachen University , Aachen , Germany

3. Sarra Bouchkati Aachen Institute of High Voltage Equipment and Grids

   RWTH Aachen University

   Affiliation as printed

   Institute of High Voltage Equipment and Grids, Digitalization and Energy Economics , RWTH Aachen University , Aachen , Germany

4. Amirali Mahjoob Aachen Institute of High Voltage Equipment and Grids

   RWTH Aachen University

   Affiliation as printed

   Institute of High Voltage Equipment and Grids, Digitalization and Energy Economics , RWTH Aachen University , Aachen , Germany

5. Andreas Ulbig Aachen Institute of High Voltage Equipment and Grids

   RWTH Aachen University

   Affiliation as printed

   Institute of High Voltage Equipment and Grids, Digitalization and Energy Economics , RWTH Aachen University , Aachen , Germany