Pith. sign in

REVIEW 6 major objections 5 minor 35 references

Scalable Fairness Shaping with LLM-Guided Multi-Agent Reinforcement Learning for Peer-to-Peer Electricity Markets

T0 review · 6 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read LLM-scored fairness bonuses shift peer-to-peer energy markets toward local trading, cutting consumer bills by up to 30% without destroying grid profit.

desk verdict Clever architecture, but the central claim is untested: no control isolates the LLM fairness critic, and the LLM scoring is underspecified, so the reported fairness and savings are not yet attributable to the method. read the letter →

arxiv 2508.18610 v1 pith:UYWHWYML submitted 2025-08-26 eess.SY cs.LGcs.SY

Shrenik Jadhav , Birva Sevak , Srijita Das , Akhtar Hussain , Wencong Su , Van-Hai Bui This is my paper · ORCID
classification eess.SYcs.LGcs.SY
keywords peer-to-peerenergytradingcontinuousdoubleauctionmulti-agentreinforcementlearningfairnessshapinglargelanguagemodelcriticrewardprosumermarketspartialobservability
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 tries to establish that fairness can be injected into peer-to-peer electricity trading without sacrificing profit. A large language model reads each hour's trade ledger and returns three fairness scores—grid reliance, seller balance, and price spread—which are added as reward bonuses to multi-agent reinforcement learning agents bidding in a continuous double auction. Across a 30-day pilot, a 90-day scaled community, and a 90-day real-world dataset, the trained policies shift roughly 54% of energy to local peer-to-peer trades, cut consumer costs by 27–30% versus grid-only procurement, keep seller participation balanced, and leave the grid profitable. The claim a sympathetic reader would take away is that LLM-guided fairness shaping is a modular, scalable way to align private bidding incentives with community-level equity, and that it can be layered onto standard PPO training without destabilizing learning.

What carries the argument

The load-bearing mechanism is the fairness-shaped reward in Eq. (17): a prosumer's slot reward is raw profit plus three LLM-derived fairness bonuses—FTG (grid-reliance penalty), FPP (price-spread tightness), and FBS scaled by the seller's volume share—with coefficients $\lambda$ ramped linearly over episodes so that fairness guidance fades in only after basic trading is learned. This sits inside a continuous double auction in which the lowest asks match the highest bids while $p_{\text{buy}} \geq p_{\text{sell}}$, with agents trained by Proximal Policy Optimization under partial observability: each agent sees only its own forecasts, weather, time, and battery state, not the order book. The LLM critic is the novel component, converting the post-slot ledger into $[0,1]$ fairness scores that become part of the learning signal.

What would settle it

Fix one hour's post-slot ledger from Case 1 and run the LLM critic ten or more times, also under several different prompts or models; if the returned FTG/FBS/FPP triplets vary by more than a small tolerance (say 0.1 on [0,1]), Eq. (17) is not a well-defined function of the ledger and the reported fairness outcomes cannot be attributed to a stable shaping signal.

Watch

Extended reading notes

Core claim

The central discovery is that post-slot fairness feedback can act as a shaping signal rather than a post-hoc evaluation. After each auction slot, the LLM critic converts the ledger into normalized scores FTG (1 if all demand is met via P2P), FBS (1 if seller quantities are evenly shared), and FPP (1 if clearing prices cluster tightly around the median), and these enter each prosumer's reward as scaled bonuses with coefficients that ramp from zero to full strength over training (Eqs. 17–20). The fairness bonuses complement rather than overwhelm economic incentives; the peer-sharing term is additionally weighted by the seller's share of the slot's traded volume. The paper argues that this aligns private incentives with community equity under partial observability, and reports that across all three cases the market stabilizes at roughly 54% peer-to-peer energy, consumer bills fall by 27–30%, seller-sharing entropy stays high, and grid net profit remains positive.

Load-bearing premise

The whole argument rests on the unstated premise that the LLM's fairness scores are stable, repeatable, and accurate reflections of the trade ledger, since the paper never specifies the scoring functions, the LLM model, the prompt, or any calibration, and Eq. (17) feeds those scores directly into training.

Editorial extensions

If this is right

  • If the central claim holds, fairness shaping transfers from a small pilot to a larger community and to a real-world mixed-asset dataset without retraining, since Cases 2 and 3 reuse policies from earlier checkpoints.
  • The ramp schedule lets fairness incentives phase in during training, and the paper reports stable convergence in roughly 1,000–1,500 episodes across cases, suggesting the shaping does not destabilize PPO.
  • The concrete economic corollaries are the roughly 54/46 peer-to-grid split and 27–30% consumer bill reductions: local trading becomes the default without the grid losing money.
  • Sensitivity results around ±20% PV and ±10% load imply that the shaped policies adapt proportionally, preserving seller-sharing entropy and grid viability under supply and demand shocks.
  • Because the framework is modular, swapping the LLM scorer or the fairness metrics would change which community-level objective is being shaped, making the design adaptable to other equity targets.

Reading between the lines

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

  • The LLM's specific role may be replaceable: FTG, FBS, and FPP are aggregate statistics of the ledger, so a hand-coded formula could plausibly produce the same scores; the paper's real contribution would then be the reward-shaping scheme rather than the language model itself. A direct ablation replacing the LLM with a deterministic calculator would settle this.
  • If the shaping scheme generalizes, the same template could steer markets toward other normative goals—carbon intensity, local resilience, or participation of vulnerable households—by swapping the scorer that reads the ledger.
  • The fairness metrics are defined at community level, so the promise depends on the community being the right unit; in larger systems with heterogeneous neighborhoods, aggregating one set of scores per slot may hide intra-community inequities.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

6 major / 5 minor

Summary. The paper proposes FairMarket-RL, a multi-agent reinforcement learning framework for peer-to-peer (P2P) electricity markets in which a large language model (LLM) computes three slot-level fairness metrics—FTG, FBS, and FPP—from the post-trade ledger and adds them as reward-shaping bonuses to a PPO-trained bidding policy. The market is modeled as a continuous double auction with discrete price–quantity actions, partial observability, storage dynamics, and grid settlement. Three case studies are presented: a 30-day, three-prosumer pilot; a 90-day, ten-prosumer synthetic community; and a 90-day, twelve-household real-world dataset with battery-equipped and PV-only prosumers. The paper reports that the framework shifts trades toward local P2P exchange (about 54% P2P share), reduces consumer costs by 27–30% relative to grid-only procurement, maintains high fairness scores, preserves grid profitability, and behaves reasonably under PV and load perturbations.

Significance. If the central claim were established, the paper would make a useful contribution: it would show that LLM-computed fairness signals can be embedded into MARL reward shaping for electricity markets without sacrificing economic performance. The manuscript has several strengths: the environment is detailed and physically grounded, the authors provide source code, the case-study progression from a small pilot to a mixed-asset real-world dataset is sensible, and the economic outcomes (consumer bills, grid profit) are independent of the fairness metrics and provide some external grounding. However, the load-bearing causal claim—that the LLM-guided fairness shaping, rather than the auction mechanism or the unshaped PPO policy, produces the reported outcomes—is not supported by the experiments as presented. The treatment condition is incompletely specified, and no control condition is run.

major comments (6)
  1. [Section 2.3 and Eq. (17)] The LLM fairness critic is not specified in any reproducible way. Section 2.3 states only that "we send the post-slot ledger to LLM, which returns three fairness metrics scaled to the interval [0,1]", but the paper does not identify the LLM, the prompt, the temperature, the parsing procedure, or any calibration. Moreover, FTG, FBS, and FPP are defined only verbally (e.g., "1 when quantities are evenly shared"), with no mathematical formulas. Because these scores enter the agent rewards through Eq. (17) and are then plotted in Figures 4, 8, and 12 as evidence of success, the reported fairness results cannot be independently verified and the treatment condition is unreproducible. A deterministic, formula-based definition of FTG, FBS, and FPP, together with an ablation that replaces the LLM with those formulas, is essential to support the paper's claims.
  2. [Sections 3.1–3.3, Eq. (17), and Section 4] No control condition isolates the effect of the fairness-shaping terms. All experiments train (or transfer) policies with the full reward in Eq. (17), including the LLM-derived bonuses; there is no baseline with beta_grid = beta_price = beta_peer = 0, no PPO-only baseline without fairness shaping, and no ablation that replaces the LLM outputs with deterministic fairness scores. Consequently, the approximately 54% P2P share, the 27–30% consumer savings, and the reported FTG/FBS/FPP values are all compatible with the possibility that the CDA clearing and unshaped PPO alone produce these outcomes. Since the title and the conclusion in Section 4 attribute the results to LLM-guided fairness shaping, this missing control is a load-bearing gap.
  3. [Sections 3.2.1–3.2.2 and 3.3.1–3.3.2, Figures 8 and 12] There is a direct internal inconsistency in the experimental protocol. Section 3.2.1 states that "Fairness-shaped PPO policies trained in Case 1 are transferred without retraining," and Section 3.3.1 states that "Policies are fixed from previous cases checkpoints to evaluate generalization without retraining or hyperparameter changes." Yet Sections 3.2.2 and 3.3.2 and Figures 8 and 12 present "training diagnostics" with Episode returns and fairness metrics over 10,000 training episodes. If the policies are not retrained, these curves cannot be training curves; if they are retrained, the transfer/generalization claim is contradicted. The manuscript must clarify which setting produced Figures 8 and 12 and what those curves actually represent.
  4. [Sections 3.1.5, 3.2.6, and 3.3.6] The economic results depend on tariff assumptions that are never reported. The claimed consumer cost reductions of 27–30% are computed relative to a grid-only baseline, and the grid's net profit (e.g., $57.7 in Case 1, about $590 in Case 2, about $379 in Case 3) depends directly on the retail and feed-in tariffs. The paper does not state these tariff values or the baseline procurement cost formula, so the economic claims cannot be reproduced or assessed. The authors should report the tariffs used, the grid-only baseline calculation, and the resulting sensitivity of the savings numbers to tariff choices.
  5. [Sections 3.1.2, 3.2.2, and 3.3.2] The reported fairness values are internally inconsistent across cases and do not support the conclusion of stable high fairness. Case 1 states that "FTG stabilizes around 0.80–0.85" (Section 3.1.2), while Case 2 reports that "FTG improves to about 0.72 by late training" (Section 3.2.2), and Case 3 says that FTG converges "slightly below" an 0.8 threshold (Section 3.3.2), although no threshold is defined anywhere. The paper should explain whether these differences are expected and should not describe fairness as uniformly "high" or "stable" when the headline metric moves from about 0.8–0.85 to about 0.72.
  6. [Figures 4, 8, and 12; Section 3.3.3] All reported results are from single training runs with no error bars, seeds, or statistical measures. The sensitivity analysis in Section 3.3.3 also appears to be a single fixed-policy counterfactual per perturbation. Given that the central claims concern robustness and generalization, the absence of multiple seeds or confidence intervals makes it difficult to distinguish systematic effects from noise in the PPO training and in the stochastic environment.
minor comments (5)
  1. [Section 2.3 and Introduction] The fairness metric names are inconsistent: the Introduction uses "Fairness-by-Share (FBS)" and "Fairness-by-Price (FPP)", while Section 2.3 uses "Fairness-Between-Sellers (FBS)" and "Fairness-of-Pricing (FPP)". The names should be standardized throughout.
  2. [Section 3.2.3 and 3.2.4] Two different figures are both labeled "Figure 9": one for hourly energy volumes and one for market fairness metrics. The latter should be renumbered.
  3. [Section 2.1.3] The text contains an incomplete word: "At the start of each slotuseholds receive..." should be "At the start of each slot, households receive...".
  4. [Equation (17)] The notation for the seller-share weighting is ambiguous: the denominator "∑_{k=1}^{m} q_{k,t}^{sold}" is not typeset cleanly, and the sum should be over all prosumers in that slot. Please clarify the indexing and define the case when the denominator is zero.
  5. [Section 3.2.3] The claim that the largest single-hour peer trade in Case 2 was about 2.4 kWh, while the 90-day P2P total is 2,260 kWh, is surprising; please verify whether this value is in kWh or reflects a different aggregation unit, and clarify the units consistently with Case 1, where the largest hourly trade is reported as about 24 kWh.

Circularity Check

1 steps flagged · score 6.0 of 10

The fairness metrics reported as evidence of success are the same LLM scores inserted into the reward in Eq. (17), making the headline fairness claim partly circular; economic outcomes and Case-2 JFI/entropy provide some independent grounding.

  1. self definitional [Section 2.3, Eq. (17); Section 3.1.2 / Figure 4; also Figures 8 and 12]
    "For each prosumer i the slot-level reward is: R_i,t = π_i,t + λ_grid(e)·β_grid·FTG_t + λ_price(e)·β_price·FPP_t + λ_peer(e)·β_peer·FBS_t·(q_sold_i / Σ q_sold_k) ... Fairness metrics remain high throughout training: FTG stabilizes around 0.80–0.85, FBS near 0.90, and FPP consistently exceeds 0.95. These trajectories show that the learned policies balance efficiency (high rewards) with equity (stable, high fairness)."

    The FTG, FBS, and FPP entering Eq. (17) are the same LLM-returned scores described in Section 2.3, and PPO is trained to maximize the shaped reward containing them. The paper then plots 'fairness metrics (FPP, FBS, FTG)' over training and treats their high values as evidence that the framework 'balance[s] efficiency with equity.' That evaluation reduces to the training objective: the agent is rewarded for increasing these exact scores, so high values are partly constructed by the reward definition rather than independently measured. Because the scoring functions and LLM outputs are not specified, the fairness outcome and the reward input cannot be separated.

full rationale

The central fairness evaluation is partially circular: the LLM-derived scores FTG/FBS/FPP are added directly to every prosumer's reward in Eq. (17), and the same scores are then plotted and cited as evidence of equitable outcomes in Cases 1–3. This is a genuine 'prediction equals input' reduction for those metrics. The paper also reports independent quantities—consumer bill reductions, grid profitability, seller-share entropy, Jain's fairness index, and the ~54% P2P trade share—that are not literally identical to the reward bonuses, giving the empirical claims some external grounding. The self-citation to the authors' earlier letter [28] is acknowledged and not load-bearing, and no uniqueness theorem is imported from prior work. Because the main fairness claim partially reduces to its own reward inputs, the score is 6 rather than 0–2; the independent economics and market-quality metrics keep it from being 8–10.

Assumptions & free parameters 5 free parameters · 6 assumptions · 1 invented entities

This ledger highlights that the simulation rests on hand-tuned reward coefficients, ramp endpoints, and unreported tariffs, plus a black-box LLM component. The fairness scores are simultaneously the objective and the evaluation metric. No physical entity is invented, but the LLM fairness critic is an unspecified engineered component with no independent validation.

free parameters (5)
  • Reward scaling factors beta_grid, beta_price, beta_peer = not reported
    Introduced in Eq. (17) as tunable scaling of the LLM fairness bonuses; their values are never given, yet they control the strength of fairness shaping and thus the reported trade-off.
  • Fairness ramp endpoints e_start, e_full = grid/price: 0.02E to 0.30E; peer: 0.30E to 0.80E
    Hand-chosen training schedule in Section 2.3 determines when fairness incentives activate and affects convergence and final fairness.
  • Weather intensity alpha = not reported; alpha in (0,1)
    Eq. (1) scales PV and load via a sunny/cloudy flag; the value is not given, so scenario generation is under-specified.
  • Grid retail and feed-in tariffs = not reported
    Consumer-cost reductions and grid profitability are reported in dollars, but tariff levels are never stated; the P2P price menu of 10-30 cents/kWh alone cannot verify the economic claims.
  • PPO hyperparameters = not reported
    Learning rate, clip range, entropy coefficient, and seeds are not listed, so the training dynamics cannot be reproduced or checked for variance.
assumptions (6)
  • domain assumption Canonical 24-hour load and PV templates, scaled by household peaks and multiplicative log-normal noise, represent realistic residential behavior (Eqs. 4-5).
    The synthetic profiles are not validated against the profiles used in the real-world Case 3, so simulation-to-reality transfer is assumed.
  • domain assumption Greedy continuous double auction clearing (Eq. 2) yields a market outcome agents can learn to exploit.
    The CDA is standard [41], but its price and quantity outcomes depend on the assumed bid/ask menus and the grid tariff.
  • domain assumption Each agent observes only local forecasts, public time/weather signals, and its own SOC; it never sees the order book (Eq. 10).
    Partial observability is a modeling choice that shapes what PPO can learn.
  • ad hoc to paper The LLM returns stable, normalized fairness scores from the market ledger (Section 2.3).
    This is the paper's core mechanism and it is entirely unspecified: no model, prompt, temperature, or scoring formulas.
  • domain assumption Consumers submit bids at the grid tariff simultaneously, while the grid acts as a price-taking residual supplier.
    This design likely guarantees P2P prices at or below the grid tariff, making consumer bill reductions partially structural rather than learned.
  • domain assumption PPO with 10,000 episodes and fairness-ramped rewards converges to stationary policies.
    Training curves are shown, but no convergence criterion, hyperparameters, seeds, or policy-checkpoint rules are reported.
invented entities (1)
  • LLM fairness critic module
    purpose: Converts the post-slot market ledger into normalized fairness scores FTG, FBS, and FPP that are added to agent rewards.
    No specification of model, prompt, or scoring formulas; no external validation that these scores are accurate measures of fairness. This is the paper's central novel component and currently a black box.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Scalable Fairness Shaping with LLM-Guided Multi-Agent Reinforcement Learning for Peer-to-Peer Electricity Markets." pith.science (2026). https://pith.science/paper/UYWHWYML

@misc{pith2026250818610,
  author       = {Pith},
  title        = {Pith review of: Scalable Fairness Shaping with LLM-Guided Multi-Agent Reinforcement Learning for Peer-to-Peer Electricity Markets},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UYWHWYML}},
  note         = {Machine review of arXiv:2508.18610}
}
read the original abstract

Peer-to-peer (P2P) energy trading is becoming central to modern distribution systems as rooftop PV and home energy management systems become pervasive, yet most existing market and reinforcement learning designs emphasize efficiency or private profit and offer little real-time guidance to ensure equitable outcomes under uncertainty. To address this gap, a fairness-aware multiagent reinforcement learning framework, FairMarket-RL, is proposed in which a large language model (LLM) critic shapes bidding policies within a continuous double auction under partial observability and discrete price-quantity actions. After each trading slot, the LLM returns normalized fairness scores Fairness-to-Grid (FTG), Fairness-Between-Sellers (FBS), and Fairness-of-Pricing (FPP) that are integrated into the reward via ramped coefficients and tunable scaling, so that fairness guidance complements, rather than overwhelms, economic incentives. The environment models realistic residential load and PV profiles and enforce hard constraints on prices, physical feasibility, and policy-update stability. Across a progression of experiments from a small pilot to a larger simulated community and a mixed-asset real-world dataset, the framework shifts exchanges toward local P2P trades, lowers consumer costs relative to grid-only procurement, sustains strong fairness across participants, and preserves utility viability. Sensitivity analyses over solar availability and aggregate demand further indicate robust performance, suggesting a scalable, LLM-guided pathway to decentralized electricity markets that are economically efficient, socially equitable, and technically sound.

Figures

Figures reproduced from arXiv: 2508.18610 by the authors.

Figure 5
Figure 5. Case 1 f Hourly energy volumes (Days 1–10). Peer-to-peer exchanges, grid imports, and PV exports to the grid. 3.1.4 24-hour role-switching snapshot A representative day ( [PITH_FULL_IMAGE:figures/full_fig_p012_5.png] view at source ↗
Figure 7
Figure 7. Case 1 — Economics (30 days). (Left) Prosumer net profit; (Center) Consumer total cost; (Right) Grid revenue, cost, and net profit. 3.2 Case 2 — 90-Day Scaled Community 3.2.1 Experimental setup and inputs We extend FairMarket-RL from the Case 1 pilot to a larger mixed-asset community comprising ten rooftop-PV prosumers (P₁–P₁₀), three consumers (C₁–C₃), and a single price-taking grid node. The market operates as 2,1… view at source ↗
Figure 9
Figure 9. Case 2 — Hourly energy volumes (Days 1–10). Peer-to-peer exchanges, grid imports, and PV exports to the grid. 3.2.4 Market quality and inclusivity Market health is assessed using three indicators ( [PITH_FULL_IMAGE:figures/full_fig_p015_9.png] view at source ↗
Figures from the paper (4 more)
Figure 11
Figure 11. Figure 11: Case 2 — Economics (90 days). (Left) Prosumer net profit; (Center) Consumer average monthly bill and source composition; (Right) Grid revenue, cost, and net profit. 3.3 Case 3- 90-day real world community 3.3.1 Experimental setup and inputs This final case evaluates F…
Figure 12
Figure 12. Figure 12: Case 3 — Training diagnostics. (Left) Total reward per episode for battery-equipped prosumers; (Right) fairness metrics (FPP, FBS, FTG) over 10,000 training episodes. 3.3.3 Sensitivity analysis (robustness to PV and load shocks) We evaluate resilience via four counter…
Figure 15
Figure 15. Figure 15: Scenario comparison − Community energy flows (15 days). (Top) No P2P; (Middle) P2P without batteries; (Bottom) P2P with batteries. 3.3.6 Community Economics Battery-equipped prosumers achieve the largest net profits via arbitrage and self-consumption; PV￾only prosumer…
Figure 16
Figure 16. Figure 16: Case 3 — Economics (90 days). (Left) Prosumer net profit; (Center) Consumer average monthly bill and source composition; (Right) Grid revenue, cost, and net profit. Across 90 days, peers exchanged approximately 4,900 kWh while grid transactions totaled about 4,250 kWh…

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

35 extracted references · 20 canonical work pages

  1. [5]

    Peer -to-peer energy systems for connected communities: A review of recent advances and emerging challenges,

    W. Tushar, C. Yuen, T. K. Saha, T. Morstyn, A. C. Chapman, M. J. E. Alam, S. Hanif, and H. V. Poor, “Peer -to-peer energy systems for connected communities: A review of recent advances and emerging challenges,” Applied Energy , vol. 282, Art. no. 116131, Jan. 2021, doi: 10.1016/j.apenergy.2020.116131

    work page arXiv 2021

  2. [7]

    Hierarchical Hybrid Multi -Agent Deep Reinforcement Learning for Peer -to-Peer Energy Trading Among Multiple Heterogeneous Microgrids,

    Y. Wu, T. Zhao, H. Yan, M. Liu, and N. Liu, “Hierarchical Hybrid Multi -Agent Deep Reinforcement Learning for Peer -to-Peer Energy Trading Among Multiple Heterogeneous Microgrids,” IEEE Transactions on Smart Grid , vol. 14, no. 6, pp. 4649 –4665, 2023, doi: 10.1109/TSG.2023.3250321

    arXiv 2023

  3. [8]

    A dynamic internal trading price strategy for networked microgrids: A deep reinforcement learning-based game-theoretic approach,

    V. H. Bui, A. Hussain, and W. Su, “A dynamic internal trading price strategy for networked microgrids: A deep reinforcement learning-based game-theoretic approach,” IEEE Transactions on Smart Grid, vol. 13, no. 5, pp. 3408–3421, Sept. 2022, doi: 10.1109/TSG.2022.3168856

    arXiv 2022

  4. [9]

    A deep reinforcement learning -based bidding strategy for participants in a peer-to-peer energy trading scenario,

    F. Zhang, Q. Yang, and D. Li, “A deep reinforcement learning -based bidding strategy for participants in a peer-to-peer energy trading scenario,” Frontiers in Energy Research, vol. 10, Art. no. 1017438, Jan. 2023, doi: 10.3389/fenrg.2022.1017438

    arXiv 2023

  5. [10]

    Federated reinforcement learning for smart building joint peer-to-peer energy and carbon allowance trading,

    D. Qiu, J. Xue, T. Zhang, J. Wang, and M. Sun, “Federated reinforcement learning for smart building joint peer-to-peer energy and carbon allowance trading,” Applied Energy, vol. 333, Art. no. 120526, Mar. 2023, doi: 10.1016/j.apenergy.2022.120526

    arXiv 2023

  6. [11]

    A critical review of safe reinforcement learning strategies in power and energy systems,

    V.-H. Bui, S. Mohammadi, S. Das, A. Hussain, G. V. Hollweg, and W. Su, “A critical review of safe reinforcement learning strategies in power and energy systems,” Engineering Applications of Artificial Intelligence , vol. 143, Art. no. 110091, Mar. 2025, doi: 10.1016/j.engappai.2025.110091

    arXiv 2025

  7. [12]

    Market Mechanisms for Local Electricity Markets: A review of models, solution concepts and algorithmic techniques,

    G. Tsaousoglou, J. S. Giraldo, and N. G. Paterakis, “Market Mechanisms for Local Electricity Markets: A review of models, solution concepts and algorithmic techniques,” Renewable and Sustainable Energy Reviews, vol. 156, art. no. 111890, 2022, doi: 10.1016/j.rser.2021.111890

    work page arXiv 2022

  8. [13]

    Designing microgrid energy markets: A case study: The Brooklyn Microgrid,

    E. Mengelkamp, J. Gärttner, K. Rock, S. Kessler, L. Orsini, and C. Weinhardt, “Designing microgrid energy markets: A case study: The Brooklyn Microgrid,” Applied Energy, vol. 210, pp. 870–880, Jan. 2018, doi: 10.1016/j.apenergy.2017.06.054

    work page doi:10.1016/j.apenergy.2017.06.054 2018

Show all 35 references
  1. [14]

    Peer -to-Peer Energy Trading of Solar and Energy Storage: A Networked Multiagent Reinforcement Learning Approach,

    C. Feng and A. L. Liu, “Peer -to-Peer Energy Trading of Solar and Energy Storage: A Networked Multiagent Reinforcement Learning Approach,” arXiv preprint arXiv:2401.13947, 2024

    2024 arXiv

  2. [15]

    Comparative analysis of auction mechanisms and bidding strategies for peer-to-peer solar transactive energy markets,

    J. Lin, M. Pipattanasomporn, and S. Rahman, “Comparative analysis of auction mechanisms and bidding strategies for peer-to-peer solar transactive energy markets,” Applied Energy, vol. 255, Art. no. 113687, 2019, doi: 10.1016/j.apenergy.2019.113687

    2019

  3. [16]

    Framework design and optimal bidding strategy for ancillary service provision from a peer -to-peer energy trading community,

    Y. Zhou, J. Wu, G. Song, and C. Long, “Framework design and optimal bidding strategy for ancillary service provision from a peer -to-peer energy trading community,” Applied Energy, vol. 278, Art. no. 115671, Nov. 2020, doi: 10.1016/j.apenergy.2020.115671

    2020

  4. [17]

    Energy Peer -to-Peer Trading in Virtual Microgrids in Smart Grids: A Game-Theoretic Approach,

    K. Anoh, S. Maharjan, A. Ikpehai, Y. Zhang, and B. Adebisi, “Energy Peer -to-Peer Trading in Virtual Microgrids in Smart Grids: A Game-Theoretic Approach,” IEEE Transactions on Smart Grid, vol. 11, no. 2, pp. 1264–1275, Mar. 2020, doi: 10.1109/TSG.2019.2934830

    2020

  5. [18]

    State-of-the-Art Analysis and Perspectives for Peer- to-Peer Energy Trading,

    Y. Zhou, J. Wu, C. Long, and W. Ming, “State-of-the-Art Analysis and Perspectives for Peer- to-Peer Energy Trading,” Engineering, vol. 6, no. 7, pp. 739 –753, Jul. 2020, doi: 10.1016/j.eng.2020.06.002

    2020 doi

  6. [20]

    Reinforcement Learning Enabled Peer -to-Peer Energy Trading for Dairy Farms,

    M. I. A. Shah, E. Barrett, and K. Mason, “Reinforcement Learning Enabled Peer -to-Peer Energy Trading for Dairy Farms,” in Advances in Practical Applications of Agents, Multi -Agent Systems, and Digital Twins: The PAAMS Collection: Proc. 22nd Int. Conf. Practical Applications ...

    2024 doi

  7. [21]

    Designing fairness in autonomous peer -to-peer energy trading,

    V. N. Behrunani, A. Irvine, G. Belgioioso, P. Heer, J. Lygeros, and F. Dörfler, “Designing fairness in autonomous peer -to-peer energy trading,” IFAC-PapersOnLine, vol. 56, no. 2, pp. 3751–3756, 2023, (Proc. 22nd IFAC World Congress, Yokohama, Japan), doi: 10.1016/j.ifacol.202...

    2023 doi

  8. [22]

    Distributionally Fair Peer -to-Peer Electricity Trading,

    E. Ruiz Irusta and J. M. Morales, “Distributionally Fair Peer -to-Peer Electricity Trading,” arXiv preprint arXiv:2501.09713, 2025

    2025 arXiv

  9. [23]

    A Peer -to-Peer Energy Trading Model for Optimizing Both Efficiency and Fairness,

    E. Kusatake, M. Imahori, and N. Shinomiya, “A Peer -to-Peer Energy Trading Model for Optimizing Both Efficiency and Fairness,” Energies, vol. 16, no. 14, Art. no. 5501, Jul. 2023, doi: 10.3390/en16145501

    2023 doi

  10. [24]

    Federated Reinforcement Learning for Energy Management of Multiple Smart Homes With Distributed Energy Resources,

    Lee, S.; Choi, D.-H. “Federated Reinforcement Learning for Energy Management of Multiple Smart Homes With Distributed Energy Resources,” IEEE Transactions on Industrial Informatics, vol. 18, no. 1, pp. 488–497, Jan. 2022. doi: 10.1109/TII.2020.3035451

    2022

  11. [25]

    Multi-Residential Energy Scheduling Under Time-of-Use and Demand Charge Tariffs With Federated Reinforcement Learning,

    J.-H. Lee, J.-Y. Park, H.-S. Sim, and H.-S. Lee, “Multi-Residential Energy Scheduling Under Time-of-Use and Demand Charge Tariffs With Federated Reinforcement Learning,” IEEE Transactions on Smart Grid , vol. 14, no. 6, pp. 4360 –4372, Nov. 2023, doi: 10.1109/TSG.2023.3251956

    2023

  12. [26]

    Multi-agent reinforcement learning for optimal energy trading and scheduling in integrated electricity and heating networks,

    M. Nazari -Heris, B. Mohammadi-Ivatloo, and M. Shafie -Khah, “Multi-agent reinforcement learning for optimal energy trading and scheduling in integrated electricity and heating networks,” Energy, vol. 238, Art. no. 121934, Jan. 2022, doi: 10.1016/j.energy.2021.121934

    2022

  13. [27]

    Equity -aware peer -to-peer trading with constrained Markov games,

    R. Gupta, P. Mandal, and S. Singh, “Equity -aware peer -to-peer trading with constrained Markov games,” Energy and AI, vol. 12, Art. no. 100248, 2024, doi: 10.1016/j.egyai.2023.100248

    2024

  14. [28]

    Jadhav, B

    S. Jadhav, B. Sevak, S. Das, A. Hussain, W. Su, and V.-H. Bui, FairMarket-RL: LLM-guided fairness shaping for multi-agent reinforcement learning in peer-to-peer markets, 2025. [Preprint]

    2025

  15. [29]

    Electricity market design for the prosumer era,

    Y. Parag and B. K. Sovacool, “Electricity market design for the prosumer era,” Nature Energy, vol. 1, Art. no. 16032, 2016. https://doi.org/10.1038/nenergy.2016.32

    2016 doi

  16. [30]

    Peer -to-peer and community-based markets: A comprehensive review,

    T. Sousa, T. Soares, P. Pinson, F. Moret, T. Baroche, and E. Sorin, “Peer -to-peer and community-based markets: A comprehensive review,” Renewable and Sustainable Energy Reviews, vol. 104, pp. 367–378, 2019. https://doi.org/10.1016/j.rser.2019.01.036

    2019 doi

  17. [31]

    Peer-to-peer energy trading: A review of the literature,

    E. A. Soto, L. Bosman, E. Wollega, and W. D. Leon -Salas, “Peer-to-peer energy trading: A review of the literature,” Applied Energy , vol. 283, Art. no. 116268, 2021. https://doi.org/10.1016/j.apenergy.2020.116268

    2021

  18. [32]

    Using peer-to-peer energy-trading platforms to incentivize prosumers to form federated power plants,

    T. Morstyn, N. Farrell, S. J. Darby, and M. D. McCulloch, “Using peer-to-peer energy-trading platforms to incentivize prosumers to form federated power plants,” Nature Energy, vol. 3, no. 2, pp. 94–101, 2018. https://doi.org/10.1038/s41560-017-0075-y

    2018 doi

  19. [33]

    Federated reinforcement learning for decentralized peer-to-peer energy trading,

    Z. Ye, D. Qiu, S. Li, Z. Fan, and G. Strbac, “Federated reinforcement learning for decentralized peer-to-peer energy trading,” Energy and AI , vol. 20, Art. no. 100500, 2025. https://doi.org/10.1016/j.egyai.2025.100500

    2025

  20. [34]

    Multi -Agent Reinforcement Learning for Automated Peer-to-Peer Energy Trading in Double-Side Auction Market,

    D. Qiu, J. Wang, J. Wang, and G. Strbac, “Multi -Agent Reinforcement Learning for Automated Peer-to-Peer Energy Trading in Double-Side Auction Market,” in Proc. IJCAI-21, pp. 2913–2920, 2021. https://doi.org/10.24963/ijcai.2021/401

    2021 doi

  21. [35]

    Probabilistic forecasting of photovoltaic generation: An efficient statistical approach,

    C. Wan, J. Lin, Y. Song, Z. Xu, and G. Yang, “Probabilistic forecasting of photovoltaic generation: An efficient statistical approach,” IEEE Transactions on Power Systems , vol. 32, no. 3, pp. 2471–2472, May 2017, doi: 10.1109/TPWRS.2016.2608740

    2017

  22. [36]

    Forecasting spot electricity prices: Deep learning approaches and empirical comparison of traditional algorithms,

    J. Lago, F. De Ridder, and B. De Schutter, “Forecasting spot electricity prices: Deep learning approaches and empirical comparison of traditional algorithms,” Applied Energy, vol. 221, pp. 386–405, Jul. 2018, doi: 10.1016/j.apenergy.2018.02.069

    2018 doi

  23. [37]

    A game -theoretic approach to energy trading in the smart grid,

    Y. Wang, W. Saad, Z. Han, H. V. Poor, and T. Başar, “A game -theoretic approach to energy trading in the smart grid,” IEEE Transactions on Smart Grid , vol. 5, no. 3, pp. 1439 –1450, May 2014, doi: 10.1109/TSG.2013.2292991

    2014

  24. [38]

    Overview of current development in electrical energy storage technologies and the application potential in power system operation,

    X. Luo, J. Wang, M. Dooner, and J. Clarke, “Overview of current development in electrical energy storage technologies and the application potential in power system operation,” Applied Energy, vol. 137, pp. 511–536, Jan. 2015, doi: 10.1016/j.apenergy.2014.09.081

    2015 doi

  25. [39]

    Peer -to-peer electricity market analysis: From variational to generalized Nash equilibrium,

    H. Le Cadre, P. Jacquot, C. Wan, and C. Alasseur, “Peer -to-peer electricity market analysis: From variational to generalized Nash equilibrium,” European Journal of Operational Research , vol. 282, no. 2, pp. 753–771, Apr. 2020, doi: 10.1016/j.ejor.2019.09.035

    2020 doi

  26. [40]

    Proximal Policy Optimization Algorithms,

    J. Schulman, F. Wolski, P. Dhariwal, A. Radford, and O. Klimov, “Proximal Policy Optimization Algorithms,” arXiv preprint arXiv:1707.06347, Aug. 2017

    2017 arXiv

  27. [41]

    Trading strategy optimization for a prosumer in continuous double auction-based peer-to-peer market: A prediction-integration model,

    K. Chen, J. Lin, and Y. Song, “Trading strategy optimization for a prosumer in continuous double auction-based peer-to-peer market: A prediction-integration model,” Applied Energy, vol. 242, pp. 1121–1133, May 2019, doi: 10.1016/j.apenergy.2019.03.094

    2019 doi

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.