Physics-Informed Learning of Proprietary Inverter Models for Grid Dynamic Studies

TL;DR

This paper introduces a physics-informed neural ODE framework to accurately emulate proprietary inverter dynamics for grid studies, overcoming industry confidentiality barriers and enhancing simulation fidelity.

Contribution

It presents a novel physics-informed latent neural ODE model that captures unmodeled inverter behaviors, improving dynamic simulation accuracy in power systems.

Findings

Enhanced simulation accuracy over purely data-driven models

Successful validation with a grid-forming inverter case study

Effective modeling of proprietary inverter dynamics without revealing internal controls

Abstract

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.