Pith. sign in

REVIEW 4 major objections 5 minor 13 references

Design and Implementation of Scalable Communication Interfaces for Reliable and Stable Real-time Co-Simulation of Power Systems

T0 review · 4 major / 5 minor · reviewed 2026-08-08 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper shows that a real-time data extrapolation filter, applied to coarse incoming measurements, removes the frequency spikes that destabilize remote transmission–distribution co-simulations with inverter-based resources.

desk verdict Useful engineering write-up of co-simulation interfaces with an honest but under-validated smoothing fix; deserves rework, not rejection. read the letter →

arxiv 2502.07866 v1 pith:PS2PRLBP submitted 2025-02-11 eess.SY cs.SY

Qi Xiao , Jongha Woo , Lidong Song , Ning Lu , Victor Paduani This is my paper · ORCID
classification eess.SYcs.SY
keywords co-simulationreal-timesimulationcommunicationinterfaceinverter-basedresourcesdataextrapolationphase-lockedlooptransmission-distributionresolutionmismatch
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper reports a practical architecture for real-time co-simulation of power transmission and distribution systems with high penetration of inverter-based resources, and its main claim is a method for keeping such simulations stable when the coupled simulators exchange data at very different rates. The authors build and test two communication interfaces, one for subsystems on a local network and one for geographically separated subsystems using either cloud file sharing or a VPN connection. In the VPN case, the coarse 10-millisecond voltage and frequency data arriving from a transmission simulator into a 100-microsecond electromagnetic-transient distribution model caused the phase-locked loops of inverter-based resources to oscillate and spike. The proposed real-time data extrapolation method smooths the received data before it feeds the phase-locked loop, and in the demonstrated fault case it removes those oscillations with less delay and error than a low-pass filter. If the claim holds, co-simulation becomes a more reliable tool for studying grids with large shares of inverter-based resources across institutions and time scales.

What carries the argument

The load-bearing object is the real-time data extrapolation method defined in Equations (1)–(3). It is a lightweight recursive predictor: the current extrapolated value is the previous extrapolated value plus the average variation of the last received data points plus a scaled error correction that pulls the output toward the latest actual measurement. The tunable parameters, $N$ (how many received data points define the average variation) and $K$ (how strongly the error increment corrects the curve), trade smoothness against accuracy. The machinery sits between the remote transmission simulator and the EMT distribution model, converting coarse, irregularly arriving phasor-domain data into a smooth waveform at the 100 μs simulation rate before the phase-locked loop (the control circuit that locks an inverter to grid frequency) synchronizes on it.

What would settle it

Re-run the VPN-based transmission–distribution fault case with the number of averaging steps set to 2 or 3 and the correction strength set to 0.01 or 0.1, then compare the extrapolated PCC voltage and phase-locked-loop-measured frequency against the transmission simulator's own recorded values. If the frequency spikes reappear, the extrapolated curve diverges, or the error grows for latencies inside the observed 17–35 ms range, the method is not robustly stable beyond its hand-picked setting.

Watch

Extended reading notes

Core claim

The central discovery, stated on the paper's own terms, is that resolution mismatch, rather than raw communication latency, is the main destabilizing factor in time-sensitive VPN-based transmission–distribution co-simulation. When 10 ms transmission-side voltage, frequency, and phase data arrive at an EMT distribution model running at 100 μs, the phase-locked loop (the control circuit that locks an inverter to grid frequency) synchronizes onto a staircase-like frequency signal and produces large oscillations and spikes during and after a fault. The proposed remedy is a real-time data extrapolation method: at each EMT timestep the interface computes an extrapolated value $X_t = X_{t-1}^{ext} + \Delta X + \Delta e$, where $\Delta X$ is the average variation of previously received data and $\Delta e = K (X_T - X_{t-1}^{ext})$ is an error-increment correction scaled by $K$. With $N=1$ and $K=0.001$, the extrapolated PCC voltage and frequency curves are smoother than both the raw received data and a low-pass-filtered version, and the phase-locked-loop-measured frequency is free of the spikes and oscillations that appear without the method. The paper concludes that the method significantly enhances the stability and reliability of distributed inverter-based-resource co-simulations.

Load-bearing premise

The load-bearing premise is that the hand-chosen smoothing settings from the single demonstrated setup, one averaging step and a tiny 0.001 correction factor, will keep the extrapolation stable in other co-simulations; the paper provides no sensitivity analysis or stability proof for changing either setting.

Editorial extensions

If this is right

  • If the method works as reported, low-pass filtering is no longer the only option for resolution mismatches, because the extrapolation method yields smoother phase-locked-loop frequency with smaller delay and magnitude error than the low-pass filter.
  • Time-sensitive VPN-based remote co-simulation can tolerate 17–35 ms data delivery intervals and roughly 20 ms communication latency without destabilizing inverter-based resources, provided the incoming data are extrapolated.
  • The file-sharing interface supports non-time-critical load-following coordination with typical 1.5–4 s synchronization delays, fixing the operating envelope of that approach.
  • The local interface carries out minute-level energy management over a local network with negligible communication delay, so device-level controllers can stay on the real-time simulator.
  • Making the implementation code open allows other groups to reproduce both interfaces and the extrapolation filter, a direct corollary of the paper's claimed transferability.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • Beyond the demonstrated power-system case, the extrapolation filter is a generic receiver-side smoother for any co-simulation where a fast subsystem receives slow, time-stamped data; the same equations could serve couplings between gas, water, or transportation networks.
  • The paper's hand-picked parameter values are a single operating point, so a natural extension is an online auto-tuning rule for the averaging length and the correction strength, or a stability analysis of the recursion; the paper itself defers parameter selection to future work.
  • A sensitivity sweep over the observed 17–35 ms delivery jitter would test whether the smoothing benefit is robust, an experiment implied but not reported.
  • Sending-side upsampling of the transmission data before transfer might remove the need for receiver-side extrapolation, but that would rely on predicting the remote model's schedule rather than reacting to received data.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper reports the design and implementation of communication interfaces for real-time co-simulation of power systems: a local interface over LAN using Modbus and TCP/IP sockets, and two remote interfaces using cloud file sharing and VPN-based TCP/IP sockets. The interfaces are evaluated on OPAL-RT with a microgrid testbed and a transmission-distribution co-simulation. To address instability caused by data resolution mismatches in the remote VPN case, the authors propose a real-time data extrapolation method described by Eqs. (1)-(3), and show, in a five-cycle fault scenario at Bus 94, that the method smooths received frequency and voltage data and reduces PLL-measured frequency oscillations compared with raw data and with a low-pass filter. The central claim is that this extrapolation method significantly enhances the stability and reliability of time-sensitive co-simulations.

Significance. If established with stronger evidence, the contribution would be practically useful: the paper provides an open-source implementation on GitHub, uses commercial real-time simulators, and addresses a genuine problem of resolution mismatch in distributed co-simulation. The comparison with LPF is a useful practical result. However, the current evidence for the central stability claim is a single qualitative case study with hand-selected parameters, and the authors themselves state that parameter selection is not explored. The paper is therefore better viewed as a preliminary engineering demonstration than as a validated method for reliable and stable co-simulation. The lack of error metrics, sensitivity analysis, and repeated trials limits the significance to the specific demonstrated configuration.

major comments (4)
  1. [§III-B, Eqs. (1)-(3)] The central stability claim rests on the extrapolation recurrence, but no stability, bounded-error, or convergence analysis is provided for Eqs. (1)-(3). The recurrence is a heuristic predictor-corrector with gain K and slope-window N, and the paper's Conclusions explicitly state that 'the impact and selection of parameters for the proposed extrapolation approach are not explored in detail.' Since N=1 and K=0.001 are selected on the demonstrated case and the same case is used for evaluation, the reader cannot assess whether the method remains stable or avoids over-smoothing under other resolution mismatches, latencies, or fault locations. Please add a formal analysis or a systematic sensitivity study over N and K, including a case with a genuine transient introduced between received samples.
  2. [§III-B, Fig. 9] The performance comparison of raw data, LPF, and extrapolation is purely qualitative. No error metrics (e.g., RMSE, maximum delay, settling time) and no statistical validation are reported, and only one fault scenario is shown. The claim that the method 'significantly enhances stability and reliability' requires quantitative results. I recommend reporting numerical errors and delay measures for both the extrapolation and LPF methods, and repeating the experiment over multiple fault locations, fault durations, and communication-latency conditions.
  3. [§III-A, local communication interface] The claim that the local interface remains stable over extended durations is not demonstrated in this paper; it is deferred to reference [10], an arXiv preprint, with the text 'Results, as discussed in [10], demonstrate that the interface remains stable over extended durations.' The accuracy comparison in Fig. 6 is also qualitative. Since the title and abstract claim reliable and stable interfaces generally, the paper should either include representative long-duration results or explicitly limit the stability claim to the tested scenarios.
  4. [§III-B, VPN-based T&D co-simulation] The paper states that received data are updated every 17-35 ms and that the EMT model runs at 100 μs, but it does not specify how the extrapolated data are inserted between received samples, nor does it quantify the actual resolution mismatch in the reported experiments. Without this information, the proposed method cannot be reproduced or applied to other configurations. Please report the exact data-exchange timeline, the number of extrapolated points per received interval, and how the extrapolated values are synchronized with the EMT timestep.
minor comments (5)
  1. [§III-B, Eqs. (1)-(3)] The notation is inconsistent: the text defines N as the number of extrapolated data points, but Eq. (2) uses n. Please use one symbol consistently throughout.
  2. [Fig. 9] Add axis labels, units, legends, and a vertical line marking the fault instant. The current figures do not show the 10 ms and 100 μs time grids, which makes it difficult to interpret the claimed improvement.
  3. [Fig. 7(b)] The delay distribution plot has no sample size or confidence interval. State how many file-sharing cycles were measured and over what time period.
  4. [References] The GitHub repository is cited as [9], but a versioned release or DOI would improve reproducibility, since the code may change after publication.
  5. [§II-A] The expression for the propagation delay in the time-sequence description appears garbled by formatting; please check that the equation renders correctly.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the extrapolation method is an empirical heuristic evaluated against raw data and LPF, not a derived prediction that reduces to its own inputs.

full rationale

The paper's central contribution is an engineering communication interface with a proposed real-time data extrapolation heuristic (Eqs. 1-3). The extrapolation is not derived from first principles, and the paper does not claim that it is; it is presented as a smoothing/prediction heuristic whose parameters are hand-chosen (n=1, K=0.001) for the demonstrated VPN-based T&D co-simulation. The evaluation compares the extrapolated output against raw received data and LPF-filtered data, so the claimed improvement is an empirical demonstration against external benchmarks, not a quantity forced by construction. The parameter values are admittedly not analyzed or generalized ('the impact and selection of parameters for the proposed extrapolation approach are not explored in detail'), which is a robustness/limitation concern, not a circularity concern. The self-citations in the paper support background architecture and prior testbed results rather than substituting for the present demonstration. No equation is defined in terms of the quantity it is used to predict, and no fitted parameter is renamed as a prediction. Therefore the derivation chain is not circular; the appropriate finding is no significant circularity with a score of 0.

Assumptions & free parameters 2 free parameters · 3 assumptions · 0 invented entities

The central claim rests on two hand-tuned extrapolation parameters and several domain assumptions inherited from prior work; no new physical entities are introduced.

free parameters (2)
  • N = 1
    Number of previous data points used in Equation (2) for average variation; chosen by hand to balance smoothness and accuracy, with no sensitivity analysis.
  • K = 0.001
    Gain for the error increment term in Equation (3); hand-tuned for the single test case in Figure 9.
assumptions (3)
  • domain assumption The EMT-phasor co-simulation update using a first-order low-pass filter with delay compensation is stable and accurate.
    Inherited from reference [3]; this paper relies on it for the local co-simulation architecture without re-validation.
  • domain assumption The observed PLL frequency oscillations are caused by data resolution mismatch rather than by other modeling or communication artifacts.
    The instability attribution is asserted in Section III.B but not demonstrated with a controlled experiment.
  • domain assumption Time-stamping exchanged data ensures synchronization and data alignment.
    Relied on throughout Sections II and III; no mechanism for clock drift or out-of-order delivery is analyzed.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Design and Implementation of Scalable Communication Interfaces for Reliable and Stable Real-time Co-Simulation of Power Systems." pith.science (2026). https://pith.science/paper/PS2PRLBP

@misc{pith2026250207866,
  author       = {Pith},
  title        = {Pith review of: Design and Implementation of Scalable Communication Interfaces for Reliable and Stable Real-time Co-Simulation of Power Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PS2PRLBP}},
  note         = {Machine review of arXiv:2502.07866}
}
read the original abstract

Co-simulation offers an integrated approach for modeling the large-scale integration of inverter-based resources (IBRs) into transmission and distribution grids. This paper presents a scalable communication interface design and implementation to enable reliable and stable real-time co-simulation of power systems with high IBR penetration. The communication interface is categorized into two types: local and remote. In local scenarios, where subsystems are connected within a single local area network (LAN), low-latency communication facilitates the seamless integration of electromagnetic transient (EMT) and phasor-domain models, enabling efficient interactions with power and energy management algorithms. For remote scenarios, data exchange is achieved via internet-based file sharing or VPN-enabled communication. The performance of both methods is evaluated using OPAL-RT as a real-time simulator, demonstrating scalability, effectiveness, and challenges specific to real-time co-simulation applications. To mitigate instability arising from data resolution mismatches in time-sensitive co-simulations, a real-time data extrapolation method is proposed. This approach significantly enhances stability and reliability, ensuring more accurate simulation outcomes. The implementation code is available on GitHub, providing researchers the tools to replicate and expand upon this work.

Figures

Figures reproduced from arXiv: 2502.07866 by the authors.

Figure 1
Figure 1. Layout of the local and remote real-time co-simulation testbed [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Workflow of the local communication interfa [PITH_FULL_IMAGE:figures/full_fig_p002_2.png] view at source ↗
Figure 3
Figure 3. Time sequence of the local co-simulation te [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: Workflow of the remote communication interfa [PITH_FULL_IMAGE:figures/full_fig_p003_4.png]
Figure 5
Figure 5. Figure 5: Time sequence of the remote communication i [PITH_FULL_IMAGE:figures/full_fig_p003_5.png]
Figure 6
Figure 6. Figure 6: Comparison of PCC voltage, frequency, and p [PITH_FULL_IMAGE:figures/full_fig_p004_6.png]
Figure 7
Figure 7. Figure 7: Simulation results of file-sharing approach [PITH_FULL_IMAGE:figures/full_fig_p004_7.png]
Figure 9
Figure 9. Figure 9: Comparison of the received, extrapolated an [PITH_FULL_IMAGE:figures/full_fig_p005_9.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

13 extracted references · 9 canonical work pages

  1. [10]

    Optimal Control Design for Operating a Hybrid PV Plant with Robust Power Reserves for Fast Frequency Regulation Services

    V. Paduani, Q. Xiao, B. Xu, D. Lubkeman, and N. Lu, “Optimal Control Design for Operating a Hybrid PV Plant with Robust Power Reserves for Fast Frequency Regulation Services”, arXiv preprint arXi v:2212.03803

    work page Pith review arXiv

  2. [1]

    Each subsystem writes its local data to a shared file and retrieves remote data for simulation

    File-Sharing Approach The file-sharing method uses platforms such as clou d-based storage services to exchange data between subsystem s. Each subsystem writes its local data to a shared file and retrieves remote data for simulation. As shown in Fig. 4(a), the wor kflow involves two steps: 1) write local data to the shared file, synchronized via the cloud ...

    work page

  3. [2]

    VPN-based Approach The VPN-based approach creates a secure, encrypted virtual private network to emulate a local network over the internet. This enables synchronized, low-latency data exchange, ma king it ideal for time-sensitive simulations while offering stronger security than WebSocket, MQTT, or public IP setups with port forw arding. Each subsystem's ...

    work page

  4. [3]

    Analysis of transforming dq impedances of different converters to a common reference frame in complex converter networks,

    Q. Xiao, P. Mattavelli, A. Khodamoradi and F. Tang, "Analysis of transforming dq impedances of different converters to a common reference frame in complex converter networks," in CES Transa ctions on Electrical Machines and Systems, vol. 3, no. 4, pp. 342-350, D ec. 2019, doi: 10.30941/CESTEMS.2019.00046

    arXiv 2019

  5. [4]

    An Asynchronous Real-time Co- simulation Platform for Modeling Interaction between Microgrids and Power Distribution Systems,

    F. Xie, C. McEntee, M. Zhang and N. Lu, "An Asynchronous Real-time Co- simulation Platform for Modeling Interaction between Microgrids and Power Distribution Systems," 2019 IEEE Power & Energy Society General Meeting (PESGM), Atlanta, GA, USA, 2019, pp. 1-5, doi: 10.1109/PESGM40551.2019.8973802

    work page arXiv 2019

  6. [5]

    Novel Re al- Time EMT- TS Modeling Architecture for Feeder Blackstart Simulations,

    V. Paduani, B. Xu, D. Lubkeman and N. Lu, "Novel Re al- Time EMT- TS Modeling Architecture for Feeder Blackstart Simulations," 2022 IEEE Power & Energy Society General Meeting (PESGM), Denver, CO, USA, 2022, pp. 1-5

    work page 2022

  7. [6]

    Networked HIL Simulation System for Modeling Large-scale Power Systems,

    F. Xie et al., "Networked HIL Simulation System for Modeling Large-scale Power Systems," 2020 52nd North American Power Symp osium (NAPS), Tempe, AZ, USA, 2021, pp. 1-6, doi: 10.1109/NAPS50074.2021.9449646

    arXiv 2020

  8. [7]

    Assessment of Transmission-level F ault Impacts on 3-phase and 1-phase Distribution IBR Operation,

    Q. Xiao et al., "Assessment of Transmission-level F ault Impacts on 3-phase and 1-phase Distribution IBR Operation," 2024 IEEE Power & Energy Society General Meeting (PESGM), Seattle, WA, USA, 2024, pp. 1-5, doi: 10.1109/PESGM51994.2024.10688676

    arXiv 2024

Show all 13 references
  1. [8]

    Real-Time Monitoring and Co ntrol of Inverter- based Resources with an Integrated T&D System,

    A. S. Mohammed et al., "Real-Time Monitoring and Co ntrol of Inverter- based Resources with an Integrated T&D System," 2023 8th IEEE Workshop on the Electronic Grid (eGRID), Karlsruhe, Germany, 2023, pp. 1-6, doi: 10.1109/eGrid58358.2023.10380901

    2023

  2. [9]

    Adapti ve cold-load pickup considerations in 2-stage microgrid unit commitment for enhancing microgrid resilience,

    R. Hu, A. Shirsat, V. Muthukaruppan, et al, "Adapti ve cold-load pickup considerations in 2-stage microgrid unit commitment for enhancing microgrid resilience, " in Applied Energy, 2024, 356: 122424

    2024

  3. [11]

    https://github.com/qixiao7/Communication-Interfaces-for-Power-System- Co-Simulation

  4. [12]

    Invisible Manipulation Deep Reinforcement Learning Enhanced S tealthy Attacks on Battery Energy Management Systems,

    Q. Xiao, L. Song L, J. Woo, R. Hu, B. Xu, K. Ye, N. Lu., “Invisible Manipulation Deep Reinforcement Learning Enhanced S tealthy Attacks on Battery Energy Management Systems,” arXiv preprint arXiv:2410.17402. 2024 Oct 22

    2024 arXiv

  5. [13]

    Under-Frequency Load Shedding for Po wer Reserve Management in Islanded Microgrids,

    B. Xu et al., "Under-Frequency Load Shedding for Po wer Reserve Management in Islanded Microgrids," in IEEE Transac tions on Smart Grid, vol. 15, no. 5, pp. 4662-4673, Sept. 2024, doi: 10.1109/TSG.2024.3393426. (a) (b) (c) Fig. 9. Comparison of the received, extrapolated and L...

    2024

Pith tools

Reviewed August 8, 2026 · model on record in the stance chip above.