Skip to content
Open access

Fast Harmonic Resonance Assessment Under Power-System Uncertainty Using Adaptive Shift-Invert Arnoldi Extraction

Sep 2026 · International Workshop on Applied Measurements for Power Systems · pp. 1-6 · 0 citations · 16 references

Abstract

This paper presents a computationally efficient framework for probabilistic harmonic resonance assessment based on resonance mode analysis (RMA). In Monte Carlo (MC) studies, conventional RMA requires repeated eigenvalue decomposition of frequency-dependent network admittance matrices, resulting in a substantial computational burden when large numbers of uncertainty realizations are considered. To address this limitation, an adaptive shift-invert Arnoldi (ASIA) method is proposed for the direct extraction of the critical modal impedance (CMI), i.e., the modal impedance associated with the admittance eigenvalue closest to zero. The method is integrated with a coarse–fine frequency-scanning strategy that concentrates modal computations around candidate resonance regions while avoiding full eigenspectrum evaluation. The resulting framework enables efficient probabilistic resonance assessment under uncertain operating conditions and network parameters. The approach is evaluated on a modified IEEE 14-bus harmonic benchmark system considering uncertainties in loads, line parameters, and capacitor/filter components. The retained resonance peaks are used to identify resonance bands and to quantify both resonance occurrence probability and modal-impedance severity distributions. The results show that different uncertainty classes affect resonance behavior in distinct ways, with line-parameter uncertainty mainly influencing resonance frequency locations and load uncertainty primarily affecting resonance severity. Compared with full eigenvalue decomposition (FEVD) and a modified power-iteration (MPI) approach, the proposed method preserves CMI accuracy while reducing the average computational time by factors of 4.31 and 1.46, respectively. These results demonstrate the suitability of the proposed framework for large-scale probabilistic harmonic-resonance studies as a model-based complement to harmonic monitoring and impedance-estimation techniques.

Read PDF

We use cookies to run the site and, with your consent, for analytics and to show ads. See our Cookie Policy.