Summary
This paper presents a comparative study of model order reduction (MOR) criteria for interaction analysis in converter-dominated power systems. Multiple reduction metrics are evaluated on a common benchmark system to assess their accuracy and suitability for eigenvalue-based interaction studies. We first review representative MOR criteria in both timeand frequency-domain frameworks, including eigenvalue-sensitivity–based methods (e.g., participation factor analysis) and Gramian-based techniques.
Read more Read lessIn practical AC/DC systems, parts of the network are often represented as black-box subsystems, which complicates interaction analysis and MOR-based model simplification, particularly for modular multilevel converter (MMC) high-voltage direct current (HVDC) applications. Furthermore, as power systems increasingly transition toward cable-dominated networks, high-order dynamics introduced by frequency-response fitting (e.g., vector fitting) can significantly increase the state-space order, making conventional interaction analysis more computationally demanding and less transparent.
Based on the comparison results, this work demonstrates that different MOR criteria yield reduced-order models with varying degrees of accuracy in preserving interaction-relevant dynamics. In contrast to prior studies, we explicitly evaluate the impact of MOR on parametric sweeps by comparing root-locus trajectories, enabling stability-oriented design insights and supporting efficient operational planning in converter-dominated AC/DC grids.
Additional informations
| Publication type | Session Materials |
|---|---|
| Reference | C2_12629_2026 |
| Publication year | |
| Publisher | CIGRE |
| Country | Serbia |
| Study committees | |
| File size | 1 MB |
| Price for non member | 30 € |
| Price for member | 30 € |
Authors
LEE Dongyeong - KU Leuven, ESAT-Electa/Etch-EnergyVille Belgium; BEERTEN Jef - KU Leuven, ESAT-Electa/Etch-EnergyVille Belgium
Keywords
Interaction analysis, Sensitivity analysis, Converter-dominated system, Model order reduction, Reduced order modeling, Gramian analysis, Black-Box system
