pith. sign in

arxiv: 2604.18732 · v1 · submitted 2026-04-20 · 📡 eess.SY · cs.SY

Stiffness-Aware Decentralized Dynamic State Estimation for Inverter-Dominated Power Systems

Xingyu Zhao , Marcos Netto , Junbo Zhao This is my paper

Pith reviewed 2026-05-10 03:35 UTC · model grok-4.3

classification 📡 eess.SY cs.SY
keywords dynamic state estimationinverter-based resourcesstiffnessstatistical linearizationmatrix exponentialdecentralized estimationpower system monitoring
0
0 comments X p. Extension

The pith

Statistical linearization lets decentralized estimators track stiff inverter dynamics stably at coarse sampling intervals.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper seeks to overcome numerical instability in dynamic state estimation for modern power grids dominated by inverter-based resources. These resources introduce multi-timescale stiff dynamics that force conventional explicit discretization methods to use impractically small sampling steps, raising communication and computation costs. The proposed approach builds local linear surrogate models through statistical linearization, then applies matrix-exponential discretization to propagate uncertainty analytically in discrete time. This combination keeps estimation accurate and stable even when sampling intervals are coarse. A sympathetic reader would care because it makes real-time monitoring feasible as renewable inverters replace traditional synchronous machines.

Core claim

The central claim is that statistical linearization constructs sufficiently accurate local linear surrogate models of the nonlinear multi-timescale inverter dynamics, which in turn permits matrix-exponential discretization to achieve analytical uncertainty propagation in discrete time and thereby enables stable, accurate, stiffness-aware decentralized dynamic state estimation at sampling rates that would destabilize conventional explicit schemes.

What carries the argument

Local linear surrogate models obtained via statistical linearization, followed by matrix-exponential discretization that performs analytical uncertainty propagation without explicit integration.

If this is right

  • Conventional explicit discretization schemes require impractically small sampling intervals to remain stable under stiff inverter dynamics, while the proposed method does not.
  • Decentralized dynamic state estimation can operate with reduced communication and computational burden while retaining accuracy.
  • The method explicitly reveals the mechanism by which stiff multi-timescale dynamics destabilize standard DSE approaches.
  • Numerical tests confirm efficient and accurate estimation under coarse sampling conditions for inverter-dominated systems.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same linear-surrogate-plus-matrix-exponential pattern could be tested on other stiff nonlinear systems such as fast chemical reactors or multi-scale mechanical control loops.
  • Integration with phasor measurement unit data streams would provide a direct field test of whether the coarse-sampling accuracy holds under realistic measurement noise.
  • Because uncertainty is propagated analytically, the method may naturally support extensions that quantify how inverter parameter uncertainty affects overall grid observability.

Load-bearing premise

Statistical linearization must yield a local linear model accurate enough for the matrix-exponential discretization to remain stable and produce estimates that track the true nonlinear behavior at large sampling intervals.

What would settle it

Run the estimator on a high-fidelity nonlinear simulation of an inverter-dominated grid at sampling intervals where explicit methods diverge; if state-estimate errors grow unbounded or the filter becomes unstable, the central claim is falsified.

Figures

Figures reproduced from arXiv: 2604.18732 by Junbo Zhao, Marcos Netto, Xingyu Zhao.

Figure 1
Figure 1. Figure 1: DSE trajectories of SA-UKF and RK4-UKF in the SMIB case under [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: NIS consistency assessment (10 different random-seed selections) of [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: RK4 stability regions in the complex plane under three sampling rates. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 2
Figure 2. Figure 2: The first row indicates that the proposed method is [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: The modified inverter-dominated 39-bus power system [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 7
Figure 7. Figure 7: NIS consistency assessment (10 different random-seed selections) for [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: NIS consistency assessment (10 different random-seed selections) for [PITH_FULL_IMAGE:figures/full_fig_p007_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: RK4 stability regions in the complex plane for different sampling [PITH_FULL_IMAGE:figures/full_fig_p007_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: DSE results of SA-UKF and RK4-UKF for G8 (a GFM) at 1800 fps sampling rate. The left, middle, and right columns show the virtual frequency ω oc and the converter-side dq-axis currents i cv d and i cv q , respectively. The filtered state covariances are shown as 95% confidence intervals (CIs) using shaded regions. Time (s) 0 0.5 1 1.5 ! pll (p.u.) 0.97 0.98 0.99 1 1.01 1.02 1.03 True SA-UKF 95% CI SA-UKF R… view at source ↗
Figure 11
Figure 11. Figure 11: DSE results of SA-UKF and RK4-UKF for G10 (a GFL) at 1800 fps sampling rate. The left, middle, and right columns show the PLL frequency ω pll and the converter-side dq-axis currents i cv d and i cv q , respectively. The filtered state covariances are shown as 95% confidence intervals (CIs) using shaded regions. responding statistical consistency assessments are given in [PITH_FULL_IMAGE:figures/full_fig_… view at source ↗
read the original abstract

Dynamic state estimation (DSE) is becoming increasingly important for monitoring inverter-dominated power systems. Due to their cascading control structures, inverter-based resources (IBRs) exhibit multi-timescale dynamics, leading to stiff system models that pose significant challenges for conventional DSE methods. In particular, explicit discretization schemes often require impractically small sampling intervals to maintain numerical stability, increasing computational and communication burdens. To address this issue, this paper proposes a stiffness-aware decentralized DSE method for inverter-dominated power systems. The statistical linearization is used to construct a local linear surrogate model for the nonlinear dynamics, which allows matrix-exponential discretization to enable analytical uncertainty propagation in discrete time, rather than relying on explicit integration schemes. This enables stable DSE at lower sampling rates. Numerical results reveal the mechanism by which stiff dynamics destabilize conventional DSE and demonstrate that the proposed method achieves efficient and accurate estimation under coarse sampling conditions.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 1 minor

Summary. The paper claims to introduce a stiffness-aware decentralized DSE method for inverter-dominated power systems. By applying statistical linearization to build local linear surrogate models of the nonlinear IBR dynamics, it enables matrix-exponential based discretization that supports analytical uncertainty propagation. This allows the method to perform stable and accurate state estimation at sampling rates that would cause conventional explicit integrators to fail due to stiffness from multi-timescale dynamics. The numerical results are said to demonstrate both the destabilization mechanism in standard DSE and the advantages of the proposed approach under coarse sampling.

Significance. This research addresses a timely and important problem in power system monitoring as the grid transitions to inverter-based resources. If the proposed method proves robust, it could enable more efficient decentralized estimation with reduced communication and computation requirements. The approach leverages established techniques (statistical linearization and matrix-exponential discretization) in a novel way for this application. Credit is due for focusing on the stiffness issue explicitly and proposing an analytical alternative to numerical integration.

major comments (1)
  1. [Numerical Results] The central claim depends on statistical linearization yielding a sufficiently accurate surrogate for the stiff nonlinear dynamics over coarse sampling intervals. Since statistical linearization only matches first and second moments at the operating point, higher-order nonlinearities and time-varying linearization points in multi-timescale IBR systems may cause the surrogate to diverge within one large step. The numerical results section must include direct trajectory comparisons between the surrogate and the full nonlinear model, along with quantitative error metrics over the sampling interval, to confirm that the discretization remains stable and accurate.
minor comments (1)
  1. [Abstract] The abstract states that numerical results demonstrate the mechanism and the method's performance, yet provides no quantitative metrics, baseline comparisons, error bars, or details on how stiffness was quantified. Adding these would make the summary more informative.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their insightful and constructive review of our manuscript. The feedback highlights an important aspect of validating the proposed method's core approximation, and we address it directly below.

read point-by-point responses

  1. Referee: [Numerical Results] The central claim depends on statistical linearization yielding a sufficiently accurate surrogate for the stiff nonlinear dynamics over coarse sampling intervals. Since statistical linearization only matches first and second moments at the operating point, higher-order nonlinearities and time-varying linearization points in multi-timescale IBR systems may cause the surrogate to diverge within one large step. The numerical results section must include direct trajectory comparisons between the surrogate and the full nonlinear model, along with quantitative error metrics over the sampling interval, to confirm that the discretization remains stable and accurate.

    Authors: We agree that explicit validation of the statistical linearization surrogate's accuracy over coarse sampling intervals is necessary to fully support the central claim. Our existing numerical results demonstrate the destabilization of conventional DSE and the overall estimation performance of the proposed approach, but they do not include direct side-by-side trajectory comparisons or quantitative error metrics (e.g., RMSE or maximum deviation) between the surrogate linear model and the full nonlinear IBR dynamics within each large step. In the revised manuscript, we will add these comparisons and error metrics in the Numerical Results section, using the same test cases and sampling rates already presented, to confirm that the surrogate remains sufficiently accurate for the discretization and uncertainty propagation steps. revision: yes

Circularity Check

0 steps flagged

No circularity: method applies standard techniques to new application domain

full rationale

The derivation chain consists of applying statistical linearization to obtain a local linear surrogate for the nonlinear IBR dynamics, followed by matrix-exponential discretization to enable stable analytical uncertainty propagation at coarse sampling intervals. These steps are standard numerical techniques whose validity does not depend on the paper's own fitted parameters or outputs. Numerical results are presented to illustrate performance differences versus conventional explicit schemes, but the core claims are not obtained by renaming or fitting quantities defined from the method itself. No self-citation chain, self-definitional construction, or fitted-input-as-prediction pattern appears in the load-bearing steps.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Based solely on the abstract, the central claim rests on standard mathematical techniques (statistical linearization and matrix exponential) whose validity is assumed from prior literature; no free parameters, ad-hoc axioms, or newly invented entities are introduced in the summary.

pith-pipeline@v0.9.0 · 5457 in / 1243 out tokens · 24877 ms · 2026-05-10T03:35:35.126944+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

29 extracted references · 4 canonical work pages

  1. [1]

    1,200 MW fault induced solar photovoltaic resource interruption dis- turbance report: Southern California 8/16/2016 event,

    “1,200 MW fault induced solar photovoltaic resource interruption dis- turbance report: Southern California 8/16/2016 event,” North American Electric Reliability Corporation (NERC), Tech. Rep., Jun. 2017

    2016

  2. [2]

    A review of current- limiting control of grid-forming inverters under symmetrical distur- bances,

    B. Fan, T. Liu, F. Zhao, H. Wu, and X. Wang, “A review of current- limiting control of grid-forming inverters under symmetrical distur- bances,”IEEE Open J. Power Electron., vol. 3, pp. 955–969, 2022

    2022

  3. [3]

    Power system recovery from momentary cessation with transient stability improvement,

    M. Savastianov, K. Smedley and J. Cao, “Power system recovery from momentary cessation with transient stability improvement,”IEEE Trans. Power Syst., 2023

    2023

  4. [4]

    doi:10.1016/j.asoc.2022.108744

    P. Marchi, P. Gill Estevez, and C. Galarza, “DSE-assisted DEF strat- egy for locating forced oscillations in synchronous generators,”Int. J. Electr. Power Energy Syst., vol. 146, p. 108744, Mar. 2023, doi: 10.1016/j.ijepes.2022.108744

    work page doi:10.1016/j.ijepes.2022.108744 2023

  5. [5]

    Cyberattack on phase-locked loops in inverter-based energy resources,

    A. Bamigbade, Y . Dvorkin and R. Karri, “Cyberattack on phase-locked loops in inverter-based energy resources,”IEEE Trans. Smart Grid, vol. 15, no. 1, pp. 821-833, Jan. 2024

    2024

  6. [6]

    Power system robust decentralized dynamic state estimation based on multiple hypothesis testing,

    J. B. Zhao and L. Mili, “Power system robust decentralized dynamic state estimation based on multiple hypothesis testing,”IEEE Trans. Power Syst., vol. 33, no. 4, pp. 4553–4562, Jul. 2018, doi: 10.1109/TP- WRS.2017.2785348

    work page doi:10.1109/tp- 2018

  7. [7]

    Data-driven adaptive unscented Kalman filter for time-varying inertia and damping estimation of utility-scale IBRs considering current limiter,

    B. Tan and J. Zhao, “Data-driven adaptive unscented Kalman filter for time-varying inertia and damping estimation of utility-scale IBRs considering current limiter,”IEEE Trans. Power Syst., vol. 39, no. 6, pp. 7331–7345, 2024, doi: 10.1109/TPWRS.2024.3379956

    work page doi:10.1109/tpwrs.2024.3379956 2024

  8. [8]

    Zhao et al., ”Roles of dynamic state estimation in power system modeling, monitoring and operation,”IEEE Trans

    J. Zhao et al., ”Roles of dynamic state estimation in power system modeling, monitoring and operation,”IEEE Trans. Power Syst., vol. 36, no. 3, pp. 2462-2472, May 2021,

    2021

  9. [9]

    Liu et al., ”Dynamic state estimation for power system control and protection,”IEEE Trans

    Y . Liu et al., ”Dynamic state estimation for power system control and protection,”IEEE Trans. Power Syst., vol. 36, no. 6, pp. 5909-5921, Nov. 2021

    2021

  10. [10]

    Dynamic state estimation for multi-machine power system by unscented Kalman filter with enhanced numerical stability,

    J. Qi, K. Sun, J. Wang and H. Liu, “Dynamic state estimation for multi-machine power system by unscented Kalman filter with enhanced numerical stability,”IEEE Trans. Smart Grid

  11. [11]

    A general decentralized dynamic state estimation with synchronous generator magnetic saturation,

    B. Tan, J. Zhao and M. Netto, “A general decentralized dynamic state estimation with synchronous generator magnetic saturation,”IEEE Trans. Power Syst., vol. 38, no. 1, pp. 960-963, Jan. 2023

    2023

  12. [12]

    Revisiting power systems time-domain simulation methods and models,

    J. D. Lara, R. Henriquez-Auba, D. Ramasubramanian, S. Dhople, D. S. Callaway and S. Sanders, “Revisiting power systems time-domain simulation methods and models,”IEEE Trans. Power Syst., 2023

    2023

  13. [13]

    Recursive dynamic state estimation for power systems with an incomplete nonlinear DAE model,

    M. Katanic, J. Lygeros, G. Hug. “Recursive dynamic state estimation for power systems with an incomplete nonlinear DAE model,”IET Gener Transm Distrib, vol. 18, no. 22, pp. 3657-3668, 2024

    2024

  14. [14]

    Protection of multi- terminal hybrid transmission lines based on dynamic states estimation,

    J. Qiu, Y . Liu, B. Wang, Y . Xie and W. Huang, “Protection of multi- terminal hybrid transmission lines based on dynamic states estimation,” 2023 IEEE PESGM, Orlando, FL, USA, 2023, pp. 1-5

    2023

  15. [15]

    A novel discrete-time state-space model for decentralized dynamic state estimation of grid-forming inverters,

    X. Zhao, M. Netto, and J. Zhao, “A novel discrete-time state-space model for decentralized dynamic state estimation of grid-forming inverters,” IEEE Trans. Power Syst., 2025

    2025

  16. [16]

    Zhao and J

    X. Zhao and J. Zhao, ”A robust and reduced-order power system dy- namic state estimator for grid-forming inverters,”IEEE Trans. Instrum. Meas., 2025

    2025

  17. [17]

    Decentralized dynamic state estimation in power systems using unscented transformation,

    A. K. Singh and B. C. Pal, “Decentralized dynamic state estimation in power systems using unscented transformation,”IEEE Trans. Power Syst., vol. 29, no. 2, pp. 794-804, March 2014

    2014

  18. [18]

    Dynamic state estimation of a grid- connected converter of a renewable generation system using adaptive cubature Kalman filtering,

    J. Zhang, T. Bi, and H. Liu, “Dynamic state estimation of a grid- connected converter of a renewable generation system using adaptive cubature Kalman filtering,”Int. J. Electr. Power Energy Syst., vol. 143, p. 108470, 2022

    2022

  19. [19]

    Dynamic state estimation based control strategy for DFIG wind turbine connected to complex power systems,

    S. Yu, T. Fernando, K. Emami and H. H. -C. Iu, “Dynamic state estimation based control strategy for DFIG wind turbine connected to complex power systems,”IEEE Trans. Power Syst., vol. 32, no. 2, pp. 1272-1281, March 2017

    2017

  20. [20]

    Dynamic state estimation for DFIG with unknown inputs based on cubature Kalman filter and adaptive interpolation,

    M. Zhu, H. Liu, J. Zhao, B. Tan, T. Bi and S. S. Yu, “Dynamic state estimation for DFIG with unknown inputs based on cubature Kalman filter and adaptive interpolation,”J. Mod. Power Syst. Clean Energy, vol. 11, no. 4, pp. 1086-1099, July 2023

    2023

  21. [21]

    Robust dynamic state estimation for DFIG via the generalized maximum correntropy criterion ensemble Kalman Filter,

    W. Ma, C. Wang, L. Dang, X. Zhang and B. Chen, “Robust dynamic state estimation for DFIG via the generalized maximum correntropy criterion ensemble Kalman Filter,”IEEE Trans. Instrum. Meas., vol. 72, pp. 1-13, 2023

    2023

  22. [22]

    Dynamic state estimation for inverter-based resources: a control-physics dual estimation framework,

    H. Huang, Y . Lin, X. Lu, Y . Zhao and A. Kumar, “Dynamic state estimation for inverter-based resources: a control-physics dual estimation framework,”IEEE Trans. Power Syst., vol. 39, no. 5, pp. 6456-6468, Sept. 2024

    2024

  23. [23]

    Power system dynamic state estimation of grid-forming inverters with current limiter,

    X. Zhao, B. Tan and J. Zhao, “Power system dynamic state estimation of grid-forming inverters with current limiter,”IEEE Trans. Power Syst., 2024

    2024

  24. [24]

    Switching dynamic state estimation and event detection for inverter-based resources with multiple control modes,

    H. Huang and Y . Lin, “Switching dynamic state estimation and event detection for inverter-based resources with multiple control modes,” IEEE Trans. Power Syst., 2024

    2024

  25. [25]

    Dynamic state estimation of hybrid systems: Inverters that switch between grid-following and grid-forming control schemes,

    B. G. Odunlami and M. Netto, “Dynamic state estimation of hybrid systems: Inverters that switch between grid-following and grid-forming control schemes,”arXiv preprintarXiv:2511.13872, 2025

    work page arXiv 2025

  26. [26]

    A new method for the nonlinear transformation of means and covariances in filters and estimators,

    S. Julier, J. Uhlmann, and H. F. Durrant-Whyte, “A new method for the nonlinear transformation of means and covariances in filters and estimators,”IEEE Trans. Autom. Control, vol. 45, no. 3, pp. 477–482, Mar. 2000

    2000

  27. [27]

    S ¨arkk¨a,Bayesian Filtering and Smoothing, Institute of Mathematical Statistics Textbooks, no

    S. S ¨arkk¨a,Bayesian Filtering and Smoothing, Institute of Mathematical Statistics Textbooks, no. 3. Cambridge, U.K.: Cambridge Univ. Press, 2013

    2013

  28. [28]

    Input-Output State-Space Representation of Grid-Following Inverters,

    X. Zhao, M. Netto, and J. Zhao, “Input-Output State-Space Representation of Grid-Following Inverters,” [Online]. Available: https:// drive.google.com/file/d/1SLX5UreMmAeWdhNQysbgBvrl9GdrQVuC/ view?usp=sharing

  29. [29]

    IEEE/IEC International Standard - Measuring relays and protection equipment - Part 118-1: Synchrophasor for power systems - Measure- ments,

    “IEEE/IEC International Standard - Measuring relays and protection equipment - Part 118-1: Synchrophasor for power systems - Measure- ments,”IEC/IEEE 60255-118-1:2018, pp. 1-78, 2018. APPENDIXA THEFOURTH-ORDERRUNGE–KUTTASCHEME A. Formulation Given the continuous-time process model in (23), the fourth- order Runge–Kutta (RK4) discretization is given by k1 ...

    2018