Physics-Informed Learning of Proprietary Inverter Models for Grid Dynamic Studies
This addresses a challenge for grid operators and researchers in performing accurate dynamic simulations and stability analyses when manufacturers do not disclose inverter details, representing an incremental improvement by integrating physics with neural learning.
The paper tackles the problem of accurately simulating grid dynamics when inverter models are proprietary by developing a physics-informed neural ODE framework, resulting in improved dynamic simulation accuracy over purely data-driven methods in a grid-forming inverter case study.
This letter develops a novel physics-informed neural ordinary differential equations-based framework to emulate the proprietary dynamics of the inverters -- essential for improved accuracy in grid dynamic simulations. In current industry practice, the original equipment manufacturers (OEMs) often do not disclose the exact internal controls and parameters of the inverters, posing significant challenges in performing accurate dynamic simulations and other relevant studies, such as gain tunings for stability analysis and controls. To address this, we propose a Physics-Informed Latent Neural ODE Model (PI-LNM) that integrates system physics with neural learning layers to capture the unmodeled behaviors of proprietary units. The proposed method is validated using a grid-forming inverter (GFM) case study, demonstrating improved dynamic simulation accuracy over approaches that rely solely on data-driven learning without physics-based guidance.