Pith. sign in

REVIEW 4 major objections 4 minor 50 references

A Principled Framework to Evaluate Quality of AC-OPF Datasets for Machine Learning: Benchmarking a Novel, Scalable Generation Method

T0 review · 4 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash

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

Pith's one-line read The paper claims that sampling power-grid load setpoints by their system-wide total, rather than independently, produces AC-OPF datasets with more diverse optima and constraint activation, and that this heuristic scales to grids with over 4

desk verdict Valuable, reproducible AC-OPF dataset generator; the load-slack validity check is the key missing evidence for the 'best balance' claim. read the letter →

arxiv 2508.19083 v1 pith:UUMEU3R2 submitted 2025-08-26 eess.SY cs.SY

Matteo Baù (1) , Luca Perbellini (2) , Samuele Grillo (2) ((1) Ricerca sul Sistema Energetico , (2) Politecnico di Milano) This is my paper
classification eess.SYcs.SY
keywords ACoptimalpowerflowmachinelearningdatasetssyntheticdatagenerationconstraintactivationloadsamplingdatasetqualitymetricsgridscalability
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

Machine learning models that directly approximate AC-OPF solutions are only as good as the datasets they train on, but the standard way of building those datasets—sampling each load uniformly around its nominal value—quietly destroys diversity on large systems. The paper shows why: the total active load across independently sampled setpoints concentrates in a narrow band as the number of loads grows, so most AC-OPF instances look alike and many generators sit pinned at their limits. Its proposed generator samples the load polytope in thin slices at controlled total-load values and solves an AC-OPF with load slack variables to keep convergence high. To make comparisons fair, it proposes three interpretable metrics—marginal entropy, activation-pattern Hamming distance, and bound activation frequency—and uses them to benchmark its method against uniform sampling, a polytope-coverage method, and a bi-level exploration method. Across nine test systems up to 4,661 buses, the heuristic ranked best among the scalable methods and consistently beat uniform sampling, which is what the paper is trying to establish.

What carries the argument

The load-slicing sampler: draw total active-load values uniformly, add a constraint slicing the load polytope into a thin band around each value, and sample each slice with a coordinate-directions hit-and-run walk. The AC-OPF is augmented with load slack variables, penalized by a value-of-lost-load cost, so convergence no longer requires preemptive polytope reduction. The three quality metrics—Q1, mean normalized Shannon entropy of primal-variable marginals; Q2, mean normalized Hamming distance between ternary constraint-activation patterns; and Q3, mean normalized activation frequency of non-redundant bounds—carry the comparative evaluation.

What would settle it

Train the same neural-network architecture on a HEDGeOPF dataset and on a uniform-sampling dataset that scores lower on the three metrics, evaluate both on a fixed held-out set of feasible AC-OPF instances, and compare out-of-sample constraint violation and optimality gap; if the lower-scored dataset is not worse, the metrics do not carry the paper's claim. Independently, record the fraction of converged HEDGeOPF instances in which any load slack variable is nonzero; a large fraction would show the dataset labels correspond to modified loads, not the sampled setpoints.

Watch

Extended reading notes

Core claim

The central claim is that a simple change in how load setpoints are sampled—forcing the system-wide total active load to vary uniformly instead of letting it concentrate—fixes the diversity deficit of AC-OPF datasets. The paper argues that under independent or polytope-uniform sampling, the total active load concentrates near its mean by the central limit theorem and the concentration of measure effect, so large systems produce nearly identical optima. Its solver samples a thin slice of the load polytope for each drawn total-load value, then solves an AC-OPF with load slack variables and a large penalty so that even infeasible setpoints become feasible with minimal load adjustments. The pape

Load-bearing premise

The ranking of datasets depends on the premise that the three proposed metrics predict how well a machine-learning model generalizes and respects constraints on AC-OPF, and that the slack-augmented solutions used as training labels still represent the sampled load setpoints; the paper provides no neural-network validation and no slack-activation statistics.

Editorial extensions

If this is right

  • A dataset generated by uniform total-load slicing has higher Q1, Q2, and Q3 scores than independent uniform sampling on every tested grid, so machine-learning models will see more varied generator setpoints and constraint behavior.
  • Because the degeneracy argument scales with load count, any uniform-sampling-based dataset for large grids is likely to be degenerate regardless of the variation range, making explicit total-load control essential.
  • Removing polytope reduction makes AC-OPF dataset generation practical for grids above 4,000 buses, with per-instance solve time remaining modest.
  • Across all compared methods, bus-voltage and branch-flow bounds activate rarely, so networks trained on these datasets may not learn those constraints; feasibility claims on such datasets should be treated cautiously.

Reading between the lines

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

  • The paper never trains a neural network; the next test would be whether datasets ranked higher by Q1–Q3 actually improve the generalization of a fixed architecture on an unseen feasible test set.
  • The slack-augmented formulation means some dataset instances are solutions for slightly adjusted loads rather than the nominal setpoints; the paper does not report how often slacks activate, so the effective input distribution may deviate from the intended one.
  • Total-load slicing could apply to security-constrained or stochastic OPF datasets, where aggregate loading similarly drives which constraints bind and which generators are marginal.
  • A hybrid strategy is suggested by the comparison: use the scalable heuristic for large grids, and reserve the more expensive bi-level exploration for small systems where its higher activation diversity can pay off.
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 / 4 minor

Summary. This paper proposes HEDGeOPF, a heuristic for generating AC-OPF datasets by sampling load setpoints from slices of a convex polytope with explicit control of total active power, then solving an AC-OPF with load slack variables to improve convergence. It introduces three quality metrics (Q1: mean normalized Shannon entropy of primal variables; Q2: average normalized Hamming distance of ternary constraint-activation patterns; Q3: normalized bound-activation frequency) and compares HEDGeOPF (MX) with OPFData (M0/20), independent uniform sampling (M0), OPFLearn (M1), and RAMBO (M2) on nine PGLib test cases up to 4661 buses. The central claim is that MX offers the best balance between dataset quality and scalability, consistently beating uniform random sampling while approaching RAMBO's quality at much lower computational cost.

Significance. The contribution is timely and practically oriented. If the framework is accepted, it provides a standardizable way to compare AC-OPF dataset generators and a scalable baseline for large power systems. Strengths include the open-source Julia implementation, the use of large PGLib cases, the explicit discussion of the concentration-of-measure effect, and the interpretability of the three proposed metrics. The paper also makes a useful empirical observation that the widely used OPFData-style sampling leads to very low constraint-activation diversity. However, the validity of the generated training pairs is not established, the metrics are introduced and operationalized by the same paper that claims its method scores best, and the comparative claims are stronger than the single-run tables support.

major comments (4)
  1. [§III-B, Eq. (7d) and Alg. 1] The slack-augmented formulation keeps the AC-OPF feasible by allowing the actual load sd in KCL (7b) to differ from the sampled setpoint ŝd via sd = ŝd + sdw_d − sup_d (7d). Algorithm 1 stores every converged solve as a dataset instance and never reports how often sup_d/sdw_d are nonzero or their magnitudes. If the slacks activate materially, the stored (input, solution) pairs are either not AC-OPF solutions for the sampled inputs, or the intended distribution over load setpoints is silently distorted. The manuscript itself motivates the slacks precisely because the unclassified load polytope (6) contains infeasible setpoints, so this is not a remote edge case. Please report slack-activation statistics (share of instances with nonzero slacks, maximum magnitudes) and either post-filter instances with nonzero slacks or record the effective load sd as the input feature.
  2. [§V-B, Tables III–V; Conclusion] The headline claim that MX 'consistently outperforms' random sampling is not supported by the tables as stated. For Q1 on generator reactive power (Table III-b), M0 has higher values than MX on grids 39 (43.7 vs 41.8), 500 (40.1 vs 37.2), and 1354 (37.1 vs 34.1), while M1 beats MX on grids 30, 39, and 57. Yet the Conclusion states 'MX consistently outperforms M0 and M1 on all test cases.' The claim should be restricted to the variable classes and metrics where it holds, or supported by a pre-specified aggregate ranking with confidence intervals. As written, the abstract's central claim is overstated.
  3. [§IV-B, §V] The paper defines Q1–Q3 as 'dataset quality for ML applications' but never validates these metrics against any downstream ML outcome. No NN is trained; no evidence is given that datasets scoring higher on Q1–Q3 improve generalization, constraint satisfaction, or sample efficiency. Since the same paper proposes the metrics and then uses them to evaluate its own method, the benchmark is at risk of being circular. The authors justify avoiding NN benchmarks on reproducibility grounds, but then the abstract's 'dataset quality' claim should either be softened to 'quality as measured by Q1–Q3' or supported by at least one NN experiment on a small and a large test case.
  4. [§V-C, Table VI] The computational comparison rests on a single run per method and grid, with no confidence intervals or repeated-seed variability. Table VI covers only four grids for M2 versus MX, and the sentence 'MX's solve time only exceeds that of M2 on the 118-bus system when scaling up to the 4661-bus grid' is difficult to parse. Since scalability is a central contribution, please report multiple independent repetitions or at least give per-instance timing distributions, and clarify the comparison on large grids where M2 could not be run.
minor comments (4)
  1. [Throughout] Typos: 'seperately' (Sec. V-A), 'al least once' (Sec. III-A), 'fis feasible' (Alg. 1).
  2. [§V-C] The sentence about MX's solve time on the 4661-bus system is ambiguous and should be rewritten for clarity.
  3. [§IV-B, Eq. (8)] The bin count B is fixed to 100 without sensitivity analysis. Since Q1 values depend on the discretization, please report robustness to B or give a principled criterion for choosing it.
  4. [§V-A, Table II] The comparison is not on identical input spaces: M0/M0/20 have no qd/pd ratio constraint while M1 and MX do, and M1 has a wider reactive variation. This should be stated more prominently in the fairness discussion, as it weakens the direct comparability of M1 against the others.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the central comparison is run against external open-source baselines and the proposed metrics do not by construction favor HEDGeOPF (RAMBO wins on Q2/Q3).

full rationale

The paper's central claim is that HEDGeOPF offers the best quality/scalability balance. The supporting evidence is an empirical benchmark against external open-source methods (M0, M1, M2) using metrics Q1-Q3. These metrics are not defined in terms of HEDGeOPF's outputs, nor is HEDGeOPF defined in terms of the metrics; they are general descriptive statistics (marginal entropy, activation-pattern Hamming distance, activation frequency) proposed before the comparison. The results do not automatically crown the authors' method: RAMBO systematically outperforms HEDGeOPF on Q2 and Q3, so the metrics are not rigged by construction. The self-citations to the authors' prior work [1] are background for the sampling heuristic and are not load-bearing for the new empirical findings. The main weakness is not circularity but an unvalidated validity assumption: in formulation (7), load slack variables can alter the sampled load setpoint, and the paper does not report how often slacks activate or whether stored instances correspond to the nominal sampled inputs. This is a correctness/validity concern about the benchmark, not a circular derivation. The paper also explicitly declines NN-based validation of its quality metrics, which limits external grounding but does not make the comparison definitionally circular. Overall, no step reduces to its inputs by construction, so the circularity score is low.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central claim rests on a hand-built sampling polytope (delta_p, delta_q, power-factor parameters, slice width epsilon) and on unvalidated quality criteria; no new physical entities are introduced.

free parameters (4)
  • Total active power slice half-width epsilon = 0.1% of support range (4)
    Set to 0.1% of the uniform support to approximate uniform total load; no sensitivity analysis is reported, and the claim of uniformity depends on this value.
  • Power factor distortion parameters delta_pf, alpha_min, alpha_max = 0.05, 0.01, 0.99 (Section V-A)
    Hand-chosen to shape the reactive-power range; affects all MX datasets and comparisons.
  • Active/reactive load variation ranges delta_p, delta_q = 100% for M0, M1, M2, MX; 20% for M0/20
    Boundary assumptions of the load polytope; chosen per method in Table II and inherited from prior work, not fitted.
  • KDE smoothing offset eta and batch sizes ns, nb = not reported in manuscript
    Algorithm 1 line 6 defines a sampling distribution proportional to 1/(KDE+eta) but eta and the batch sizes are not given; they affect sampling focus and truncation of extreme loading regions.
assumptions (5)
  • standard math Central limit theorem and concentration of measure in high-dimensional convex bodies predict that total active power concentrates near its mean under independent or polytope-uniform sampling (Section II, citing [36]).
    This motivates the total-active-power sampling design.
  • domain assumption AC-OPF optimal solutions are strongly influenced by total active power load through economic dispatch (Section II, citing [4]).
    If false, the claimed diversity mechanism would not transfer to OPF variables.
  • domain assumption Higher marginal entropy, higher activation-pattern Hamming distance, and higher bound activation frequency make datasets better for training ML models (Section IV-A).
    None of the three criteria are calibrated against ML generalization in the paper.
  • ad hoc to paper Load slack variables in (7d)-(7j) keep the AC-OPF feasible for infeasible sampled setpoints, and the resulting solution is treated as a valid dataset instance.
    The paper does not filter nonzero slack or report its magnitude; many instances may not correspond to the sampled load.
  • standard math Coordinate-direction hit-and-run samples the sliced H-polytope uniformly enough to approximate the slice (Section III-A, Volesti [37]).
    The uniformity of the final dataset depends on this mixing assumption.

how reviews work

0 comments
Cite this review

Pith. "Pith review of A Principled Framework to Evaluate Quality of AC-OPF Datasets for Machine Learning: Benchmarking a Novel, Scalable Generation Method." pith.science (2026). https://pith.science/paper/UUMEU3R2

@misc{pith2026250819083,
  author       = {Pith},
  title        = {Pith review of: A Principled Framework to Evaluate Quality of AC-OPF Datasets for Machine Learning: Benchmarking a Novel, Scalable Generation Method},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UUMEU3R2}},
  note         = {Machine review of arXiv:2508.19083}
}
read the original abstract

Several methods have been proposed in the literature to improve the quality of AC optimal power flow (AC-OPF) datasets used in machine learning (ML) models. Yet, scalability to large power systems remains unaddressed and comparing generation approaches is still hindered by the absence of widely accepted metrics quantifying AC-OPF dataset quality. In this work, we tackle both these limitations. We provide a simple heuristic that samples load setpoints uniformly in total load active power, rather than maximizing volume coverage, and solves an AC-OPF formulation with load slack variables to improve convergence. For quality assessment, we formulate a multi-criteria framework based on three metrics, measuring variability in the marginal distributions of AC-OPF primal variables, diversity in constraint activation patterns among AC-OPF instances and activation frequency of variable bounds. By comparing four open-source methods based on these metrics, we show that our heuristic consistently outperforms uniform random sampling, whether independent or constrained to a convex polytope, scoring as best in terms of balance between dataset quality and scalability.

Figures

Figures reproduced from arXiv: 2508.19083 by the authors.

Figure 1
Figure 1. KDE distribution of min-max scaled active [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. Share of generators with normalized Shannon entropy in active power values below 10%, with partition into three entropy sub-intervals. Results are [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Comparison among every method of Table II in terms cumulative mean normalized Hamming distance (i.e., [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

50 extracted references · 44 canonical work pages

  1. [1]

    An Efficient and Scalable Algo- rithm for the Creation of Representative Synthetic AC-OPF Datasets,

    L. Perbellini, M. Ba `u, and S. Grillo, “An Efficient and Scalable Algo- rithm for the Creation of Representative Synthetic AC-OPF Datasets,” in 2024 IEEE 8th Forum on Research and Technologies for Society and Industry Innovation (RTSI) , 2024, pp. 653–658

    work page 2024

  2. [8]

    Initial estimate of ac optimal power flow with graph neural networks,

    A. Deihim, D. Apostolopoulou, and E. Alonso, “Initial estimate of ac optimal power flow with graph neural networks,”Electric Power Systems Research, vol. 234, p. 110782, 2024

    2024

  3. [2]

    Contribution ´a l’ ´etude du dispatching ´economique,

    J. Carpentier, “Contribution ´a l’ ´etude du dispatching ´economique,” Bulletin de la Soci ´et´e Francaise des ´electriciens, vol. 3, pp. 431–447, 1962. 10

    work page 1962

  4. [3]

    Zero duality gap in optimal power flow problem,

    J. Lavaei and S. H. Low, “Zero duality gap in optimal power flow problem,” IEEE Transactions on Power Systems , vol. 27, no. 1, pp. 92–107, 2012

    work page 2012

  5. [4]

    History of Optimal Power Flow and Formulations,

    M. Cain, R. O’Neill, and A. Castillo, “History of Optimal Power Flow and Formulations,” Federal Energy Regulatory Commission Washington DC, pp. 1–36, 2012

    work page 2012

  6. [5]

    Critical review of recent advances and further devel- opments needed in ac optimal power flow,

    F. Capitanescu, “Critical review of recent advances and further devel- opments needed in ac optimal power flow,” Electric Power Systems Research, vol. 136, pp. 57–68, 2016

    work page 2016

  7. [6]

    Learning warm-start points for ac optimal power flow,

    K. Baker, “Learning warm-start points for ac optimal power flow,” in 2019 IEEE 29th International Workshop on Machine Learning for Signal Processing (MLSP), 2019, pp. 1–6

    work page 2019

  8. [7]

    Smart-PGSim: Using Neural Network to Accelerate AC-OPF Power Grid Simulation,

    W. Dong, Z. Xie, G. Kestor, and D. Li, “Smart-PGSim: Using Neural Network to Accelerate AC-OPF Power Grid Simulation,” in SC20: International Conference for High Performance Computing, Networking, Storage and Analysis , 2020, pp. 1–15

    work page 2020

Show all 50 references
  1. [9]

    Leveraging Power Grid Topology in Machine Learning Assisted Optimal Power Flow,

    T. Falconer and L. Mones, “Leveraging Power Grid Topology in Machine Learning Assisted Optimal Power Flow,” IEEE Transactions on Power Systems, vol. 38, no. 3, pp. 2234–2246, 2023

    2023

  2. [10]

    Hybrid learning aided inactive constraints filtering algorithm to enhance ac opf solution time,

    F. Hasan, A. Kargarian, and J. Mohammadi, “Hybrid learning aided inactive constraints filtering algorithm to enhance ac opf solution time,” IEEE Transactions on Industry Applications , vol. 57, no. 2, pp. 1325– 1334, 2021

    2021

  3. [11]

    Learning optimal solutions for extremely fast ac optimal power flow,

    A. S. Zamzam and K. Baker, “Learning optimal solutions for extremely fast ac optimal power flow,” in 2020 IEEE International Conference on Communications, Control, and Computing Technologies for Smart Grids (SmartGridComm), 2020, pp. 1–6

    2020

  4. [12]

    Predicting ac optimal power flows: Combining deep learning and lagrangian dual methods,

    F. Fioretto, T. W. Mak, and P. Van Hentenryck, “Predicting ac optimal power flows: Combining deep learning and lagrangian dual methods,” in Proceedings of the AAAI conference on artificial intelligence , vol. 34, no. 01, 2020, pp. 630–637

    2020

  5. [13]

    Optimal power flow using graph neural networks,

    D. Owerko, F. Gama, and A. Ribeiro, “Optimal power flow using graph neural networks,” in 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , 2020, pp. 5930–5934

    2020

  6. [14]

    DC3: A learning method for optimization with hard constraints,

    P. L. Donti, D. Rolnick, and J. Z. Kolter, “DC3: A learning method for optimization with hard constraints,” in International Conference on Learning Representations, 2021

    2021

  7. [15]

    Physics-Informed Neural Net- works for AC Optimal Power Flow,

    R. Nellikkath and S. Chatzivasileiadis, “Physics-Informed Neural Net- works for AC Optimal Power Flow,” Electric Power Systems Research, vol. 212, p. 108412, 2022

    2022

  8. [16]

    Learning to Solve the AC-OPF Using Sensitivity-Informed Deep Neural Networks,

    M. K. Singh, V . Kekatos, and G. B. Giannakis, “Learning to Solve the AC-OPF Using Sensitivity-Informed Deep Neural Networks,” IEEE Transactions on Power Systems , vol. 37, no. 4, pp. 2833–2846, 2022

    2022

  9. [17]

    Deepopf: A feasibility- optimized deep neural network approach for ac optimal power flow problems,

    X. Pan, M. Chen, T. Zhao, and S. H. Low, “Deepopf: A feasibility- optimized deep neural network approach for ac optimal power flow problems,” IEEE Systems Journal , vol. 17, no. 1, pp. 673–683, 2023

    2023

  10. [18]

    Topology-aware graph neural networks for learning feasible and adaptive ac-opf solutions,

    S. Liu, C. Wu, and H. Zhu, “Topology-aware graph neural networks for learning feasible and adaptive ac-opf solutions,” IEEE Transactions on Power Systems, vol. 38, no. 6, pp. 5660–5670, 2023

    2023

  11. [19]

    Unsupervised optimal power flow using graph neural networks,

    D. Owerko, F. Gama, and A. Ribeiro, “Unsupervised optimal power flow using graph neural networks,” in 2024 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) , 2024, pp. 6885– 6889

    2024

  12. [20]

    Optimal power flow with physics-informed typed graph neural networks,

    T. B. Lopez-Garcia and J. A. Dom ´ınguez-Navarro, “Optimal power flow with physics-informed typed graph neural networks,” IEEE Transactions on Power Systems, vol. 40, no. 1, pp. 381–393, 2025

    2025

  13. [21]

    CANOS: A Fast and Scalable Neural AC-OPF Solver Robust To N-1 Perturbations,

    L. Piloto, S. Liguori, S. Madjiheurem, M. Zgubic, S. Lovett, H. Tomlinson, S. Elster, C. Apps, and S. Witherspoon, “CANOS: A Fast and Scalable Neural AC-OPF Solver Robust To N-1 Perturbations,”

  14. [22]

    Optimization proxies using limited labeled data and training time – a semi-supervised bayesian neural network approach,

    P. Pareek, A. Jayakumar, K. Sundar, D. Deka, and S. Misra, “Optimization proxies using limited labeled data and training time – a semi-supervised bayesian neural network approach,” 2025. [Online]. Available: https://arxiv.org/abs/2410.03085

    2025 arXiv

  15. [23]

    Self-supervised primal-dual learning for constrained optimization,

    S. Park and P. Van Hentenryck, “Self-supervised primal-dual learning for constrained optimization,” Proceedings of the AAAI Conference on Artificial Intelligence, vol. 37, no. 4, pp. 4052–4060, 2023

    2023

  16. [24]

    A learning-augmented approach for AC optimal power flow,

    J. Rahman, C. Feng, and J. Zhang, “A learning-augmented approach for AC optimal power flow,” International Journal of Electrical Power & Energy Systems, vol. 130, p. 106908, 2021

    2021

  17. [25]

    Topology-transferable physics-guided graph neural network for real-time optimal power flow,

    M. Yang, G. Qiu, J. Liu, Y . Liu, T. Liu, Z. Tang, L. Ding, Y . Shui, and K. Liu, “Topology-transferable physics-guided graph neural network for real-time optimal power flow,” IEEE Transactions on Industrial Informatics, vol. 20, no. 9, pp. 10 857–10 872, 2024

    2024

  18. [26]

    OPF-Learn: An Open- Source Framework for Creating Representative AC Optimal Power Flow Datasets,

    T. Joswig-Jones, K. Baker, and A. S. Zamzam, “OPF-Learn: An Open- Source Framework for Creating Representative AC Optimal Power Flow Datasets,” in 2022 IEEE Power and Energy Society Innovative Smart Grid Technologies Conference (ISGT) , 2022, pp. 1–5

    2022

  19. [27]

    Scalable Bilevel Optimization for Gen- erating Maximally Representative OPF Datasets,

    I. V . Nadal and S. Chevalier, “Scalable Bilevel Optimization for Gen- erating Maximally Representative OPF Datasets,” in 2023 IEEE PES Innovative Smart Grid Technologies Europe (ISGT EUROPE) , 2023, pp. 1–6

    2023

  20. [28]

    A large synthetic dataset for machine learning applications in power transmission grids,

    M. Gillioz, G. Dubuis, and P. Jacquod, “A large synthetic dataset for machine learning applications in power transmission grids,” Scientific Data, vol. 12, no. 1, 2025

    2025

  21. [29]

    Generating quality datasets for real-time security assessment: Balancing historically rele- vant and rare feasible operating conditions,

    A.-A. B. Bugaje, J. L. Cremer, and G. Strbac, “Generating quality datasets for real-time security assessment: Balancing historically rele- vant and rare feasible operating conditions,” International Journal of Electrical Power & Energy Systems , vol. 154, p. 109427, 2023

    2023

  22. [30]

    Efficient creation of datasets for data-driven power system applications,

    A. Venzke, D. K. Molzahn, and S. Chatzivasileiadis, “Efficient creation of datasets for data-driven power system applications,” Electric Power Systems Research, vol. 190, p. 106614, 2021

    2021

  23. [31]

    Split-based sequential sampling for realtime security assessment,

    A.-A. B. Bugaje, J. L. Cremer, and G. Strbac, “Split-based sequential sampling for realtime security assessment,” International Journal of Electrical Power & Energy Systems , vol. 146, p. 108790, 2023

    2023

  24. [32]

    Enriching Neural Network Training Dataset to Improve Worst-Case Performance Guarantees,

    R. Nellikkath and S. Chatzivasileiadis, “Enriching Neural Network Training Dataset to Improve Worst-Case Performance Guarantees,” in 2023 IEEE Belgrade PowerTech, 2023, pp. 1–6

    2023

  25. [33]

    Optimal power flow based on physical-model-integrated neural network with worth-learning data generation,

    Z. Hu, H. Zhang, R. Yang, Y . Chen, and H. Wu, “Optimal power flow based on physical-model-integrated neural network with worth-learning data generation,” CSEE Journal of Power and Energy Systems, pp. 1–10, 2025

    2025

  26. [34]

    Generating high- quality datasets with critical state samples for data-driven power system applications,

    Z. Zhang, X. Bai, Z. Weng, J. Xiao, and J. Zhang, “Generating high- quality datasets with critical state samples for data-driven power system applications,” IEEE Access, vol. 13, pp. 27 507–27 515, 2025

    2025

  27. [35]

    HEDGeOPF: High-quality, Efficient Dataset Genererator for the AC Optimal Power Flow,

    M. Ba `u and L. Perbellini, “HEDGeOPF: High-quality, Efficient Dataset Genererator for the AC Optimal Power Flow,” 2025. [Online]. Available: https://github.com/mttb91/HEDGeOPF.jl

    2025

  28. [36]

    Vershynin, High-dimensional probability: An introduction with ap- plications in data science

    R. Vershynin, High-dimensional probability: An introduction with ap- plications in data science . Cambridge University Press, 2018, vol. 47

    2018

  29. [37]

    volesti: V olume Approximation and Sampling for Convex Polytopes in R,

    A. Chalkis and V . Fisikopoulos, “volesti: V olume Approximation and Sampling for Convex Polytopes in R,” The R Journal, vol. 13, no. 2, p. 561, 2021

    2021

  30. [38]

    MATPOWER User’s Manual, Version 7.1

    R. D. Zimmerman and C. E. Murillo-Sanchez, “MATPOWER User’s Manual, Version 7.1.” 2020. [Online]. Available: https: //matpower.org/docs/MATPOWER-manual-7.1

    2020

  31. [39]

    Powermodels.jl: An open-source framework for exploring power flow formulations,

    C. Coffrin, R. Bent, K. Sundar, Y . Ng, and M. Lubin, “Powermodels.jl: An open-source framework for exploring power flow formulations,” in Power Systems Computation Conference (PSCC) , 2018, pp. 1–8

    2018

  32. [40]

    JuMP 1.0: Recent improvements to a modeling language for mathematical optimization,

    M. Lubin, O. Dowson, J. Dias Garcia, J. Huchette, B. Legat, and J. P. Vielma, “JuMP 1.0: Recent improvements to a modeling language for mathematical optimization,” Mathematical Programming Computation , vol. 15, p. 581–589, 2023

    2023

  33. [41]

    Delving into Deep Imbalanced Regression,

    Y . Yang, K. Zha, Y . Chen, H. Wang, and D. Katabi, “Delving into Deep Imbalanced Regression,” in Proceedings of the 38th International Conference on Machine Learning , vol. 139, 2021, pp. 11 842–11 851

    2021

  34. [42]

    Generalized Simpson-diversity,

    H.-R. Gregorius and E. M. Gillet, “Generalized Simpson-diversity,” Ecological Modelling, vol. 211, no. 1, pp. 90–96, 2008

    2008

  35. [43]

    Identifying Redundant Constraints for AC OPF: The Challenges of Local Solutions, Relaxation Tightness, and Approximation Inaccuracy,

    A. D. O. Aquino, L. A. Roald, and D. K. Molzahn, “Identifying Redundant Constraints for AC OPF: The Challenges of Local Solutions, Relaxation Tightness, and Approximation Inaccuracy,” in 2021 North American Power Symposium (NAPS) , 2021, pp. 1–6

    2021

  36. [44]

    OPFData: Large-scale datasets for AC optimal power flow with topological perturbations,

    S. Lovett, M. Zgubic, S. Liguori, S. Madjiheurem, H. Tomlinson, S. Elster, C. Apps, S. Witherspoon, and L. Piloto, “OPFData: Large-scale datasets for AC optimal power flow with topological perturbations,”

  37. [45]

    The Power Grid Library for Benchmarking AC Optimal Power Flow Algorithms,

    S. Babaeinejadsarookolaee, A. Birchfield, R. D. Christie, C. Coffrin, C. DeMarco, R. Diao, M. Ferris, S. Fliscounakis, S. Greene, R. Huang, C. Josz, R. Korab, B. Lesieutre, J. Maeght, T. W. K. Mak, D. K. Molzahn, T. J. Overbye, P. Panciatici, B. Park, J. Snodgrass, A. Tbaileh,...

    2021 arXiv

  38. [46]

    Available: https://arxiv.org/abs/2406.07234

    [Online]. Available: https://arxiv.org/abs/2406.07234

    arXiv

  39. [47]

    I. V . Nadal and S. Chevalier, “RAMBO,” 2023. [Online]. Available: https://github.com/SamChevalier/RAMBO

    2023

  40. [48]

    OPFLearn.jl,

    T. Joswig-Jones and A. S. Zamzam, “OPFLearn.jl,” 2023. [Online]. Available: https://github.com/NREL/OPFLearn.jl/releases/tag/v0.1.2

    2023

  41. [50]

    On the implementation of an interior- point filter line-search algorithm for large-scale nonlinear programming,

    A. W ¨achter and L. T. Biegler, “On the implementation of an interior- point filter line-search algorithm for large-scale nonlinear programming,” Mathematical Programming, vol. 106, no. 1, pp. 25–57, 2006

    2006

  42. [2024]

    Available: https://arxiv.org/abs/2403.17660

    [Online]. Available: https://arxiv.org/abs/2403.17660

    arXiv

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.