Online Long-Term Voltage Stability Margin Estimation for IBR/DER Dominated Power System with Integrated VSM-Aware TSO-DSO Framework

TL;DR

This paper presents a machine learning-based framework for real-time long-term voltage stability margin estimation and enhancement in power systems with high inverter-based resources, integrating TSO-DSO coordination.

Contribution

It introduces a physics-informed ML model that derives an explicit VSM expression, enabling real-time VSM enforcement and optimization in modern power systems.

Findings

The framework accurately estimates VSM under various scenarios.

VSM enhancement improves system stability margins.

Simulation validates effectiveness in IEEE test systems.

Abstract

The rapid growth of inverter-based resources (IBRs) and distributed energy resources (DERs) has fundamentally altered the long-term voltage stability characteristics of modern power systems. This article leverages the advantages of machine learning (ML) for the online estimation of long-term voltage stability margin (VSM) and enhancement of VSM through coordinated transmission system operator-distribution system operator (TSO-DSO) optimization. An explicit analytical VSM expression is derived from offline T&D co-simulation data using a physics-informed ML-trained model under probabilistic loading and generation mix scenarios, while accounting for unbalanced distribution modeling. The resulting closed-form VSM representation is linearized and embedded into the TSO optimization problem, enabling real-time enforcement of minimum VSM constraints. We further enhance operational efficiency by incorporating VSM sensitivities into both transmission and distribution optimization, allowing prioritization of the most influential reactive power resources. Simulation studies conducted on the IEEE 30-bus transmission network integrated with multiple IEEE 37-node distribution feeders validate that the proposed framework successfully achieves the desired VSM enhancement while maintaining high estimation accuracy.