pith. sign in

arxiv: 2604.21259 · v1 · submitted 2026-04-23 · 📡 eess.SY · cs.SY· math.OC

A Convexified Eulerian Framework for Scalable Coordination of Massive DER Populations

Ge Chen , Yiwei Qiu , Shiyao Zhang , Pengfei Su , Haoran Deng , Hongcai Zhang This is my paper

Pith reviewed 2026-05-09 21:23 UTC · model grok-4.3

classification 📡 eess.SY cs.SYmath.OC
keywords distributed energy resourcesEulerian modelingconvexificationlinear programmingbroadcast controlscalable coordinationPDE-constrained optimizationprivacy preservation
0
0 comments X

The pith

Convexification of an Eulerian PDE model for DER populations produces a sparse LP whose solution is a single state-dependent broadcast signal for coordination at cost independent of population size.

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

The paper develops a two-layer architecture for coordinating large populations of storage-like distributed energy resources. At the macroscopic layer the population is represented as a continuum density whose evolution follows a PDE, and the resulting non-convex optimization is converted into a sparse linear program by finite-volume discretization plus flux lifting. The LP yields a unified broadcast signal that every individual resource uses, together with its local state, to recover its own setpoint. Computational effort therefore remains constant as the number of resources grows. At the microscopic layer a data-mixing protocol sends only density histograms upstream, preserving aggregator-side privacy, while a Wasserstein relaxation of cyclic constraints supplies extra operational flexibility.

Core claim

By representing the DER population from an Eulerian perspective as a density function whose dynamics obey a partial differential equation, the authors convert the bilinear non-convex coordination problem into a sparse linear program through finite-volume discretization and flux lifting. The optimal solution of this LP supplies a single state-dependent broadcast signal that each DER employs to determine its local action, achieving population-wide coordination whose per-unit computation time does not increase with the number of participating resources.

What carries the argument

Finite-volume discretization of the Eulerian PDE combined with flux lifting, which removes the bilinear non-convexity and yields a sparse linear program whose solution is the broadcast coordination signal.

If this is right

  • Coordination signals can be computed in time that does not grow with the number of DERs.
  • Each DER recovers its local setpoint autonomously from the broadcast signal and its private state.
  • Only aggregated density histograms reach the aggregator, so raw individual states remain private.
  • The Wasserstein relaxation of cyclic constraints supplies additional scheduling flexibility and improves economic performance.

Where Pith is reading between the lines

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

  • The same continuum-plus-convexification pattern could be applied to other large-scale coordination tasks such as electric-vehicle charging fleets or flexible demand response.
  • Empirical tests that compare the continuum prediction against detailed agent-based simulations for finite but large populations would quantify the approximation error and its effect on closed-loop performance.
  • Because the broadcast signal is state-dependent and low-dimensional, the method may integrate with existing utility communication channels while keeping total data volume low.

Load-bearing premise

The collective behavior of many individual DERs can be represented accurately by a continuous density function whose evolution is governed by the PDE, and the finite-volume plus flux-lifting steps preserve feasibility and near-optimality of the original problem.

What would settle it

For a moderate population of several thousand DERs, solve both the proposed LP and a high-fidelity non-convex formulation of the identical problem; check whether the aggregate power and state trajectories produced by the broadcast signal stay within a few percent of the exact optimum in cost and constraint violation.

Figures

Figures reproduced from arXiv: 2604.21259 by Ge Chen, Haoran Deng, Hongcai Zhang, Pengfei Su, Shiyao Zhang, Yiwei Qiu.

Figure 1
Figure 1. Figure 1: Schematic comparison between the Lagrangian and Eulerian perspec [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Overview of the proposed two-layer architecture for coordinating a [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Schematic illustration of the physical mechanisms governing the [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Prices for purchasing and selling electricity, together with the base [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 6
Figure 6. Figure 6: Optimized load profiles of the proposed Eulerian method and the [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Average violations of SoC bound and cyclic constraints per EV [PITH_FULL_IMAGE:figures/full_fig_p007_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Comparison between the target SoC distribution defined by the cyclic [PITH_FULL_IMAGE:figures/full_fig_p007_8.png] view at source ↗
Figure 11
Figure 11. Figure 11: Comparison between the target SoC distribution defined by the [PITH_FULL_IMAGE:figures/full_fig_p008_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: Total operational costs and cyclic constraint violations under varying [PITH_FULL_IMAGE:figures/full_fig_p008_12.png] view at source ↗
Figure 10
Figure 10. Figure 10: Impact of temporal and spatial discretization resolutions on SoC [PITH_FULL_IMAGE:figures/full_fig_p008_10.png] view at source ↗
Figure 16
Figure 16. Figure 16: Total operational costs and cyclic constraint violations under varying [PITH_FULL_IMAGE:figures/full_fig_p009_16.png] view at source ↗
Figure 14
Figure 14. Figure 14: Average violations of state-bound and cyclic constraints per TCL [PITH_FULL_IMAGE:figures/full_fig_p009_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: The target terminal state distribution required by the cyclic constraint [PITH_FULL_IMAGE:figures/full_fig_p009_15.png] view at source ↗
read the original abstract

This paper proposes a scalable coordination framework with aggregator-side privacy protection for storage-like distributed energy resources (DERs). The framework adopts a two-layer architecture. At the macroscopic layer, building upon an \emph{Eulerian} modeling perspective, the DER population is represented as a continuum whose density evolution is governed by a partial differential equation (PDE), such that the computational complexity is independent of the population size. To address the bilinear non-convexity in this PDE-constrained optimization problem, we develop a convexification method that combines finite-volume discretization with a flux-lifting technique, reformulating the macroscopic problem into a sparse linear program (LP). The LP solution yields a unified, state-dependent broadcast signal for population coordination. Furthermore, a Wasserstein-based relaxation is introduced to replace rigid cyclic constraints and provide additional operational flexibility for improved economic performance. At the microscopic layer, individual resources autonomously recover local setpoints from the broadcast signal and their local states, while an upstream data-mixing protocol aggregates individual states into a macroscopic density histogram without exposing raw individual states to the aggregator. Numerical studies validate the scalability, feasibility, and economic effectiveness of the proposed framework.

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

2 major / 3 minor

Summary. The paper proposes a two-layer scalable coordination framework for large populations of storage-like DERs. The macroscopic layer models the population as a continuum density evolving under a PDE, applies finite-volume discretization plus flux-lifting to convert the resulting bilinear PDE-constrained program into a sparse LP, and extracts a unified state-dependent broadcast signal; a Wasserstein relaxation replaces rigid cyclic constraints. The microscopic layer lets individual DERs recover local setpoints from the broadcast signal and local state, while a data-mixing protocol aggregates states into a density histogram without exposing raw data.

Significance. If the flux-lifting step is tight and the recovered microscopic trajectories remain feasible and near-optimal, the approach would deliver population-size-independent computation together with aggregator-side privacy, which is a meaningful advance for real-time coordination of massive DER fleets. The combination of Eulerian PDE modeling with convexification and Wasserstein relaxation is technically interesting and could influence future work on continuum approximations in power-system control.

major comments (2)
  1. [§3] §3 (convexification via finite-volume discretization and flux-lifting): the manuscript asserts that the procedure produces a sparse LP whose solution yields feasible, near-optimal broadcast signals for the original non-convex problem, yet provides neither an explicit derivation of the lifted variables, bounds on the relaxation gap, nor a proof that the subsequent Wasserstein relaxation preserves tightness. This gap directly affects the central claim that the LP solution remains valid when mapped back to individual DER trajectories.
  2. [Numerical studies] Numerical studies section: the abstract states that numerical studies validate feasibility and economic effectiveness, but the reported experiments do not appear to include controlled tests that quantify the optimality gap or infeasibility rate when the LP solution is applied to the original microscopic non-convex problem. Such metrics are required to substantiate the “near-optimal” claim.
minor comments (3)
  1. [Abstract] Abstract: the description of the two-layer architecture and the three technical contributions (finite-volume + flux-lifting, Wasserstein relaxation, data-mixing) is compressed into a single paragraph; splitting into separate sentences would improve readability.
  2. Notation: several acronyms (DER, PDE, LP, Wasserstein) are used without an initial definition list; a short nomenclature table or explicit first-use definitions would help readers from outside the immediate subfield.
  3. Figure captions: ensure that every figure caption explicitly states what is plotted (e.g., “broadcast signal vs. time for N=10^5 DERs”) and indicates whether the plotted trajectories are microscopic or macroscopic.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive and detailed feedback. The comments highlight important aspects of the convexification procedure and the validation of near-optimality. We address each major comment below and will incorporate revisions to strengthen the manuscript.

read point-by-point responses

  1. Referee: [§3] §3 (convexification via finite-volume discretization and flux-lifting): the manuscript asserts that the procedure produces a sparse LP whose solution yields feasible, near-optimal broadcast signals for the original non-convex problem, yet provides neither an explicit derivation of the lifted variables, bounds on the relaxation gap, nor a proof that the subsequent Wasserstein relaxation preserves tightness. This gap directly affects the central claim that the LP solution remains valid when mapped back to individual DER trajectories.

    Authors: We acknowledge that while §3 outlines the finite-volume discretization and flux-lifting approach leading to the sparse LP, the explicit algebraic derivation of the lifted flux variables, quantitative bounds on any relaxation gap, and a formal argument for tightness preservation under the Wasserstein relaxation are not presented in sufficient detail. We will revise §3 to include a step-by-step derivation of the lifting, derive or cite bounds on the discretization-induced gap, and add a discussion (with conditions) on how the Wasserstein relaxation interacts with the lifted LP solution. These additions will directly support the claim of validity when mapping back to microscopic trajectories. revision: yes

  2. Referee: [Numerical studies] Numerical studies section: the abstract states that numerical studies validate feasibility and economic effectiveness, but the reported experiments do not appear to include controlled tests that quantify the optimality gap or infeasibility rate when the LP solution is applied to the original microscopic non-convex problem. Such metrics are required to substantiate the “near-optimal” claim.

    Authors: We agree that explicit quantification of the optimality gap and infeasibility rates is necessary to substantiate the near-optimality claim when the LP-derived broadcast signal is applied to the original non-convex microscopic problem. The current numerical studies focus on scalability, feasibility for large populations, and economic performance but do not include side-by-side controlled comparisons (e.g., on small instances solvable to optimality). We will expand the numerical studies section with such experiments, reporting optimality gaps and infeasibility rates for representative cases. revision: yes

Circularity Check

0 steps flagged

No circularity: direct convex reformulation of PDE-constrained problem

full rationale

The derivation proceeds by representing the DER population as a continuum density evolving under a PDE, then applying finite-volume discretization followed by flux-lifting to obtain a sparse LP. This is a standard sequence of mathematical transformations (discretization + variable lifting) that does not define any quantity in terms of itself or rename a fitted result as a prediction. The broadcast signal is the LP solution; the microscopic recovery step uses only local state and the received signal. The Wasserstein relaxation is an explicit additional relaxation, not a self-referential closure. No self-citation is invoked as a load-bearing uniqueness theorem, and no parameter is fitted to a subset of data then re-used as a 'prediction'. The central claim therefore remains independent of its own outputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on domain assumptions about continuum modeling of DER populations and mathematical reformulations whose validity cannot be checked from the abstract alone.

axioms (1)
  • domain assumption DER population can be represented as a continuum density whose evolution is governed by a PDE
    Invoked in the macroscopic layer description to achieve population-size-independent complexity.

pith-pipeline@v0.9.0 · 5520 in / 1217 out tokens · 34767 ms · 2026-05-09T21:23:31.057236+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

32 extracted references · 32 canonical work pages

  1. [1]

    Charging infrastructure access and operation to reduce the grid impacts of deep electric vehicle adoption,

    S. Powell, G. V . Cezar, L. Min, I. M. Azevedo, and R. Rajagopal, “Charging infrastructure access and operation to reduce the grid impacts of deep electric vehicle adoption,”Nature Energy, vol. 7, no. 10, pp. 932–945, 2022

    work page 2022

  2. [2]

    Quantifying spatiotem- poral charging flexibility of electric vehicles as virtual grid assets to accelerate sustainable energy transition,

    G. Chen, N. R. Shishvan, Z. Guo, and J. Qin, “Quantifying spatiotem- poral charging flexibility of electric vehicles as virtual grid assets to accelerate sustainable energy transition,”Cell Reports Physical Science, vol. 6, no. 11, 2025

    work page 2025

  3. [3]

    Scheduling HV AC loads to promote renewable generation integration with a learning-based joint chance-constrained approach,

    G. Chen, H. Zhang, H. Hui, and Y . Song, “Scheduling HV AC loads to promote renewable generation integration with a learning-based joint chance-constrained approach,”CSEE Journal of Power and Energy Systems, pp. 1–12, 2023

    work page 2023

  4. [4]

    Impact of electric vehicle charging demand on power distribution grid congestion,

    Y . Li and A. Jenn, “Impact of electric vehicle charging demand on power distribution grid congestion,”Proceedings of the National Academy of Sciences, vol. 121, no. 18, p. e2317599121, 2024

    work page 2024

  5. [5]

    Real-time grid and DER co-simulation platform for testing large-scale DER coordination schemes,

    A. Khurram, M. Amini, L. A. D. Espinosa, P. D. H. Hines, and M. R. Almassalkhi, “Real-time grid and DER co-simulation platform for testing large-scale DER coordination schemes,”IEEE Transactions on Smart Grid, vol. 13, no. 6, pp. 4367–4378, 2022

    work page 2022

  6. [6]

    Distributed coordination of charging stations considering aggregate EV power flexibility,

    D. Yan, C. Ma, and Y . Chen, “Distributed coordination of charging stations considering aggregate EV power flexibility,”IEEE Transactions on Sustainable Energy, vol. 14, no. 1, pp. 356–370, 2023

    work page 2023

  7. [7]

    Coordinated scheduling of EV charging service and energy arbitrage for truck mobile charging stations,

    K. He, H. Jia, Y . Mu, X. Yu, Y . Zhou, W. Gan, and J. Wu, “Coordinated scheduling of EV charging service and energy arbitrage for truck mobile charging stations,”IEEE Transactions on Smart Grid, vol. 17, no. 1, pp. 467–482, 2026

    work page 2026

  8. [8]

    Data-driven mean-corrected recursive estimation-based optimal DER dispatch for distribution system voltage control,

    H. Wang, Y . Yao, J. Zhao, and F. Ding, “Data-driven mean-corrected recursive estimation-based optimal DER dispatch for distribution system voltage control,”IEEE Transactions on Sustainable Energy, vol. 16, no. 4, pp. 2861–2873, 2025

    work page 2025

  9. [9]

    Price coordination for electric vehicle fleet using mean field game theory,

    J. Hedel and N. Nguyen, “Price coordination for electric vehicle fleet using mean field game theory,” in2024 56th North American Power Symposium (NAPS), pp. 1–6, 2024

    work page 2024

  10. [10]

    A novel differentiable predictive control (DPC) approach for safe and optimal EV charging and discharg- ing scheduling,

    Y . Li, B. Dong, X. Wang, and Y . Qiu, “A novel differentiable predictive control (DPC) approach for safe and optimal EV charging and discharg- ing scheduling,”Applied Energy, vol. 402, p. 127032, 2026

    work page 2026

  11. [11]

    Stochastic EV charging dispatch in unbalanced three-phase networks based on interpretable fuzzy representation of user preferences,

    X. Li, F. Luo, C. Zhang, and Z. Y . Dong, “Stochastic EV charging dispatch in unbalanced three-phase networks based on interpretable fuzzy representation of user preferences,”IEEE Transactions on Power Systems, vol. 40, no. 2, pp. 1623–1635, 2025

    work page 2025

  12. [12]

    Adaptive demand shaping with heterogeneous types of thermostatically controlled loads,

    X. Chen, S. Baldi, R. Tian, F. Hu, X. Wang, and D. Yan, “Adaptive demand shaping with heterogeneous types of thermostatically controlled loads,”IEEE Transactions on Control Systems Technology, vol. 34, no. 1, pp. 31–45, 2026

    work page 2026

  13. [13]

    A deep rl-based algo- rithm for coordinated charging of electric vehicles,

    Z. Zhang, Y . Wan, J. Qin, W. Fu, and Y . Kang, “A deep rl-based algo- rithm for coordinated charging of electric vehicles,”IEEE Transactions on Intelligent Transportation Systems, vol. 23, no. 10, pp. 18774–18784, 2022

    work page 2022

  14. [14]

    Multi- period DER coordination using ADMM-based three-block distributed AC optimal power flow considering inverter V olt-Var control,

    D. Gebbran, S. Mhanna, A. C. Chapman, and G. Verbi ˇc, “Multi- period DER coordination using ADMM-based three-block distributed AC optimal power flow considering inverter V olt-Var control,”IEEE Transactions on Smart Grid, vol. 14, no. 4, pp. 2874–2889, 2023

    work page 2023

  15. [15]

    A fully decentralized hierarchical transactive energy framework for charging EVs with local DERs in power distribution systems,

    J. Yang, T. Wiedmann, F. Luo, G. Yan, F. Wen, and G. H. Broadbent, “A fully decentralized hierarchical transactive energy framework for charging EVs with local DERs in power distribution systems,”IEEE Transactions on Transportation Electrification, vol. 8, no. 3, pp. 3041– 3055, 2022

    work page 2022

  16. [16]

    A semi-decentralized real-time charging schedul- ing scheme for large EV parking lots considering uncertain EV arrival and departure,

    W. Wang and L. Wu, “A semi-decentralized real-time charging schedul- ing scheme for large EV parking lots considering uncertain EV arrival and departure,”IEEE Transactions on Smart Grid, vol. 15, no. 6, pp. 5871–5884, 2024

    work page 2024

  17. [17]

    Fair and scalable electric vehicle charging under electrical grid constraints,

    G. Tsaousoglou, J. S. Giraldo, P. Pinson, and N. G. Paterakis, “Fair and scalable electric vehicle charging under electrical grid constraints,” IEEE Transactions on Intelligent Transportation Systems, vol. 24, no. 12, pp. 15169–15177, 2023

    work page 2023

  18. [18]

    Performance- barrier event-triggered PDE control of traffic flow,

    P. Zhang, B. Rathnayake, M. Diagne, and M. Krstic, “Performance- barrier event-triggered PDE control of traffic flow,”IEEE Transactions on Automatic Control, vol. 70, no. 9, pp. 5720–5735, 2025

    work page 2025

  19. [19]

    Modeling and control of aggregate air conditioning loads for robust renewable power management,

    S. Bashash and H. K. Fathy, “Modeling and control of aggregate air conditioning loads for robust renewable power management,”IEEE Transactions on Control Systems Technology, vol. 21, no. 4, pp. 1318– 1327, 2013

    work page 2013

  20. [20]

    Load management: Model-based control of aggregate power for populations of thermostatically controlled loads,

    C. Perfumo, E. Kofman, J. H. Braslavsky, and J. K. Ward, “Load management: Model-based control of aggregate power for populations of thermostatically controlled loads,”Energy Conversion and Management, vol. 55, pp. 36–48, 2012

    work page 2012

  21. [21]

    Massive coordi- nation of distributed energy resources in VPP: A mean field rl-based bi- level optimization approach,

    Z. Dai, M. Tan, Y . Yang, X. Liu, R. Wang, and Y . Su, “Massive coordi- nation of distributed energy resources in VPP: A mean field rl-based bi- level optimization approach,”IEEE Transactions on Cybernetics, vol. 55, no. 3, pp. 1332–1346, 2025

    work page 2025

  22. [22]

    Secondary frequency regulation via EV aggregation-based virtual power plants: A mean field game strategy,

    K. Zhang, Y . Xie, N. Liu, and Y . Wang, “Secondary frequency regulation via EV aggregation-based virtual power plants: A mean field game strategy,”IEEE Transactions on Smart Grid, vol. 16, no. 5, pp. 3919– 3933, 2025

    work page 2025

  23. [23]

    Mean field optimal energy management of plug-in hybrid electric vehicles,

    M. Shokri and H. Kebriaei, “Mean field optimal energy management of plug-in hybrid electric vehicles,”IEEE Transactions on Vehicular Technology, vol. 68, no. 1, pp. 113–120, 2019

    work page 2019

  24. [24]

    Large- scale EV charging coordination: a detailed exploration of mean-field and reinforcement learning approaches,

    J. Caball ´e, P. Segovia, C. Ocampo-Martinez, and N. Quijano, “Large- scale EV charging coordination: a detailed exploration of mean-field and reinforcement learning approaches,”Control Engineering Practice, vol. 168, p. 106669, 2026

    work page 2026

  25. [25]

    Joint resource allocation for V2X communications with multi-type mean-field rein- forcement learning,

    Y . Xu, X. Wu, Y . Tang, J. Shang, L. Zheng, and L. Zhao, “Joint resource allocation for V2X communications with multi-type mean-field rein- forcement learning,”IEEE Transactions on Intelligent Transportation Systems, vol. 26, no. 10, pp. 17462–17476, 2025

    work page 2025

  26. [26]

    A mean-field game method for decentralized charging coordination of a large population of plug-in electric vehicles,

    M. A. Tajeddini and H. Kebriaei, “A mean-field game method for decentralized charging coordination of a large population of plug-in electric vehicles,”IEEE Systems Journal, vol. 13, no. 1, pp. 854–863, 2019

    work page 2019

  27. [27]

    A mean field control approach for smart charging with aggregate power demand constraints,

    A. S ´eguret, C. Wan, and C. Alasseur, “A mean field control approach for smart charging with aggregate power demand constraints,” in2021 IEEE PES Innovative Smart Grid Technologies Europe (ISGT Europe), pp. 01–05, 2021

    work page 2021

  28. [28]

    Control theory for stochastic distributed param- eter systems, an engineering perspective,

    Q. L ¨u and X. Zhang, “Control theory for stochastic distributed param- eter systems, an engineering perspective,”Annual Reviews in Control, vol. 51, pp. 268–330, 2021

    work page 2021

  29. [29]

    The use of Fick’s law in modeling diffusion processes,

    J. Lowney and R. Larrabee, “The use of Fick’s law in modeling diffusion processes,”IEEE Transactions on Electron Devices, vol. 27, no. 9, pp. 1795–1798, 1980

    work page 1980

  30. [30]

    J. N. Reddy, N. Anand, and P. Roy,Finite element and finite volume methods for heat transfer and fluid dynamics. Cambridge University Press, 2022

    work page 2022

  31. [31]

    Experimental methods in chemical engineering: Computa- tional fluid dynamics/finite volume method—CFD/FVM,

    L. Van Hoecke, D. Boeye, A. Gonzalez-Quiroga, G. S. Patience, and P. Perreault, “Experimental methods in chemical engineering: Computa- tional fluid dynamics/finite volume method—CFD/FVM,”The Canadian Journal of Chemical Engineering, vol. 101, no. 2, pp. 545–561, 2023

    work page 2023

  32. [32]

    Why the 1-Wasserstein distance is the area between the two marginal CDFs,

    M. D. Angelis and A. Gray, “Why the 1-Wasserstein distance is the area between the two marginal CDFs,” 2021

    work page 2021