Pith. sign in

REVIEW 3 major objections 6 minor 34 references

Learning a local trading strategy: deep reinforcement learning for grid-scale renewable energy integration

T0 review · 3 major / 6 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read A model-free DQN battery controller averages 61% of perfect-foresight profit, beating forecast-based control in hard-to-predict markets.

desk verdict Careful empirics with a real market-timing flaw: the RL agent and the rules-based benchmark both see the current interval's real-time LMP, which CAISO only publishes after the interval, so the headline 61% describes a non-causal experiment. read the letter →

arxiv 2411.15422 v1 pith:PZYDMPW6 submitted 2024-11-23 cs.LG cs.AIcs.SYeess.SYmath.OC

Caleb Ju , Constance Crozier This is my paper · ORCID
classification cs.LGcs.AIcs.SYeess.SYmath.OC
keywords deepreinforcementlearningbatteryenergystoragePVgenerationlocationalmarginalpriceDQNrecedinghorizoncontrolarbitragerenewableintegration
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 asks whether a model-free reinforcement learning controller can operate a grid-scale battery co-located with solar generation almost as well as an optimizer that knows the future. Using a deep Q-network (DQN) trained only on current and historical prices and solar output, the authors report that the RL agent earns on average 61% (and up to 96%) of the profit of an approximate perfect-foresight benchmark, and that it outperforms both a tuned rules-based controller and a receding-horizon controller with learned price forecasts on average. The result matters because RL would offer a battery controller that needs no explicit forecast module and can adapt as price and weather patterns shift. The paper also finds that RL shifts solar energy toward high-demand periods and produces more diverse local dispatch decisions than the benchmarks, both concrete advantages for grid integration.

What carries the argument

The central object is a deep Q-network (DQN), a neural network that approximates the Q-function of Q-learning by mapping a state to the expected future reward of each action. The state contains the battery's state-of-charge plus the current and up-to-four-hours-lagged locational marginal price and solar output; actions are buy, sell, or null; and the reward is the net profit from energy trades. The DQN is trained by solving the Bellman optimality equation through gradient steps on sampled transitions. The argument's comparisons rest on three benchmarks: a rules-based buy-low/sell-high threshold controller tuned by a genetic algorithm, a receding-horizon controller using LSTM price forecasts, and a linear programming upper bound solved with perfect price foresight (neglecting efficiency and self-discharge terms, then evaluated with them).

What would settle it

Re-run the DQN experiments with a one-step price lag, replacing the current λt in the state with λt−1, and compare profit ratios to the approximate optimum; if the average drops well below 61%, the reported performance depends on non-causal price information.

Watch

Extended reading notes

Core claim

The paper claims that a deep Q-network (DQN) operating a grid-scale battery co-located with photovoltaic generation is a practical near-optimal controller. Using only the current and past four hours of locational marginal prices and solar output as observations, the RL agent earns on average 61% of the profit of an approximate perfect-foresight linear-program benchmark across three California pricing nodes, two seasons, and three PV sizes, with the best seeds reaching 96%; it also outperforms a genetic-algorithm-tuned rules-based controller and a receding-horizon controller that uses LSTM price forecasts on average. The authors further claim that RL dispatch shifts solar energy toward high-demand periods and, because each node learns its own local price dynamics, produces more diverse battery-state trajectories than the benchmarks, reducing the risk of synchronized ramping from many systems acting alike.

Load-bearing premise

The controller is assumed to act on the current interval's locational marginal price, but real-time prices are typically published after the interval, so the experiment tests a controller that may rely on future information.

Editorial extensions

If this is right

  • A DQN controller can be deployed without a separate forecasting module, since it learns price and solar dynamics directly from local historical data.
  • In settings where locational marginal prices are hard to predict, RL beats receding-horizon control, while the reverse can hold when forecasts are good.
  • Because RL responses vary with local price dynamics, many RL controllers trained on their own nodes will superimpose less than a fleet running identical rules, lowering the risk of grid-level ramping events.
  • The reward function only maximizes profit, yet the resulting dispatch shows higher cross-correlation between net load and demand than the benchmarks, so profit-seeking alone can partially align batteries with demand.

Reading between the lines

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

  • If the real-time locational marginal price is not available at decision time, the reported 61% should be treated as an upper bound for a strictly causal controller; retraining with a one-interval lag would quantify the gap.
  • The observed diversity of dispatch suggests a system-level benefit of individualized RL policies: multiple batteries sharing one network could have their aggregate charging ramps tested directly against a common-policy fleet.
  • Since the approximate optimum ignores storage efficiency and self-discharge, incorporating those losses into the benchmark would lower the denominator and increase the attained ratio, so the headline is conservative in that direction.
  • One could test the solar-alignment finding by checking whether the cross-correlation advantage persists when the agent is trained on prices without a clear daily periodicity, such as a wind-dominated grid region.
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

3 major / 6 minor

Summary. The paper studies deep reinforcement learning (DQN) for the real-time operation of a grid-scale battery co-located with solar generation, using 15-minute CAISO real-time LMP and solar data at three California nodes. The RL agent is compared against a rules-based buy-low/sell-high controller, a sell-only baseline, receding-horizon control with an LSTM price forecast, and a perfect-foresight approximate-optimal LP benchmark. The central empirical claim is that RL achieves on average 61% (up to 96%) of the approximate-optimal profit, outperforms the other realistic controllers on average, shifts solar to higher-demand periods, and produces more diverse dispatch across locations.

Significance. If the claims hold, the paper would provide useful evidence that a model-free RL controller can capture much of the value of perfect-foresight battery operation without an explicit forecast module, and that it can yield beneficial demand alignment and action diversity. The manuscript has tangible strengths: the code is released, the evaluation uses a chronological train/test split, multiple locations, seasons, PV sizings, and 10 random seeds, and the comparison set includes a more advanced RHC baseline. However, three load-bearing issues—causal observability of the current-interval price, the construction of the upper bound, and the definition of the 61% average—need to be resolved before the empirical claims are reliable.

major comments (3)
  1. [Sections 3, 2.2, and 4] The controller state in Section 3 includes the "current and historical (up to four hours prior)" locational marginal price λt, and the rules-based benchmark in Section 2.2 acts on the "current" λ. Since the data are CAISO real-time 15-minute LMPs, whose settlement prices are published only after the interval has ended, the text does not establish that λt is causally available when the agent chooses the action for interval t. If λt is the settlement price for the current interval, both the RL policy and the rules-based controller have a one-interval lookahead, which biases the comparison against RHC (which must forecast future prices) and undermines the "without a forecast module" reading of the central claim. Please state the timing convention explicitly and, if the settlement price is used, re-run the experiments with the last available price or a pre-dispatch price; the same caveat applies to the current-interval solar power in the state.
  2. [Section 5.1 and Eq. (4)] The description of the ~OPT benchmark is internally ambiguous: Section 2.1 presents Eq. (4) as an upper bound with the dynamics from Eq. (1), but Section 5.1 says the optimization model "ignores battery efficiency and self-discharge terms, but its solution is evaluated with them." If the reported ~OPT profit is obtained by simulating the relaxed solution under efficiency and self-discharge, that simulated value is not guaranteed to be an upper bound on the true optimal profit. Please clarify whether the denominator in the 61% figure is the relaxed LP objective value or the simulated value, and if necessary recompute the ratio with a true upper bound.
  3. [Section 5.1 and Table 3] The headline "61% of the profit" is not defined. For Table 3, the ratio of summed RL profits to summed ~OPT profits is 906/1483 = 61%, but the average of the six per-scenario ratios is approximately 51%, and one scenario has negative profit for both RL and OPT. Please state the aggregation rule used for the headline and report per-scenario ratios or a confidence interval, because the choice of aggregation changes the empirical claim.
minor comments (6)
  1. [Section 2.2] The threshold variables sp and bp are named "sell price" and "buy price," but the listed rule is "If λ ≤ sp, buy" and "If λ ≥ bp, sell"; the naming appears reversed and should be corrected or clarified.
  2. [Figure 2 caption] The caption contains the typo "pndoes" for "pnodes."
  3. [Section 5.1] The sentence "There are find single seeds which achieve 97% of the optimal profit" is ungrammatical, and the 97% value is not tied to a specific table or figure.
  4. [Section 4] Solar data are described as 1-hour increments while LMPs and battery dynamics are 15-minute; please state how the solar series was interpolated or aggregated to the 15-minute time step.
  5. [Figure 8 caption] The caption says "Mean (line) and confidence interval (shaded) of cumulative profit of state-of-charge and LMP," but the figure panels show SOC and LMP over time, not cumulative profit; please re-word the caption.
  6. [Section 2.1, Eq. (4)] The nonnegativity constraint in Eq. (4) lists pstore_t, but this variable is not defined in the model; it should be removed or defined.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central claims are empirical benchmark results measured on held-out data, with no fitted parameter renamed as a prediction and no load-bearing self-citation.

full rationale

The paper's central comparison is an empirical benchmark study: a DQN policy trained to maximize cumulative profit is evaluated on chronologically held-out test data against a rules-based controller, receding-horizon control with an LSTM forecast, a sell-only baseline, and a perfect-foresight MILP upper bound. Each baseline is defined independently of the RL result. The approximate optimal benchmark (~OPT) is not circular: it ignores battery efficiency and self-discharge while the solution is evaluated with them, making it a genuine upper bound on the lossy problem rather than a quantity defined in terms of the RL policy. The rules-based buy/sell thresholds and DQN hyperparameters are tuned on the training split and then applied to the test split; this is standard model selection and evaluation, not fitting the target result. The 61% figure is a measured statistic over test periods, not a fitted parameter renamed as a prediction. No load-bearing premise is justified solely by a self-citation: the paper relies on external standard references for RL, dynamic programming, genetic algorithms, and receding-horizon control, and the authors do not cite their own prior work as authority for a contested claim. The potential concern that current-interval CAISO real-time LMPs may not be causally available at decision time is an experimental-validity or correctness issue, not an instance of circularity in which a result is equivalent to its inputs by construction or a fitted input is called a prediction. The derivation chain is self-contained with respect to the claims made.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The central claims rest on a standard MDP model for battery dispatch with a discrete action space, a profit reward based on LMPs, and two benchmark constructions that introduce tuning and modeling assumptions. No new physical entities are postulated. The most fragile assumptions are the observability of the current real-time LMP at decision time and the validity of the no-efficiency LP as an upper bound.

free parameters (3)
  • Rules-based buy/sell price thresholds (sp, bp) = not reported
    Tuned by a genetic algorithm on the training period (Section 2.2); the comparison depends on this baseline being well tuned.
  • DQN hyperparameters (exploration fraction decay, max gradient steps, training steps) = decay 0.9685, 200,000 training steps
    Chosen by random search (Section 4); the reported RL performance depends on these settings, though the paper averages over 10 seeds.
  • LSTM price-forecast training configuration = 100 epochs
    Used for receding-horizon control (Section 5.1); architecture and hyperparameters are not specified, so RHC's poor winter performance may partly reflect tuning.
assumptions (6)
  • domain assumption Battery dynamics follow equations (1)-(2) with constant efficiency η=0.93 and self-discharge only above 90% SoC.
    Used to define the MDP transition and the LP benchmark; not validated against a physical battery or degradation model.
  • domain assumption LMP is an adequate measure of the value of balancing service and demand alignment.
    Objective (3) maximizes profit using λt; Section 5.4 infers demand alignment from LMP-driven behavior.
  • domain assumption Any excess solar must be sold rather than curtailed.
    Stated in Section 2 after Eq. (2); affects the reward and the sell-only baseline.
  • standard math An optimal solution to the battery LP will charge/discharge at zero or maximum power, justifying the three discrete actions.
    Section 3 uses the extreme-point argument; this is plausible but not proven for the efficiency-adjusted model.
  • domain assumption The current real-time LMP is observable when the agent chooses an action.
    State and rules-based policy use current λt (Sections 2.2 and 3); the paper does not discuss whether settlement prices are available before the interval.
  • ad hoc to paper The ~OPT benchmark that ignores efficiency and self-discharge is an upper bound on optimal profit.
    Section 2.1 calls the LP an upper bound but Section 5.1 removes efficiency and self-discharge from the solved model, so the upper-bound property is not established.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Learning a local trading strategy: deep reinforcement learning for grid-scale renewable energy integration." pith.science (2026). https://pith.science/paper/PZYDMPW6

@misc{pith2026241115422,
  author       = {Pith},
  title        = {Pith review of: Learning a local trading strategy: deep reinforcement learning for grid-scale renewable energy integration},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/PZYDMPW6}},
  note         = {Machine review of arXiv:2411.15422}
}
read the original abstract

Variable renewable generation increases the challenge of balancing power supply and demand. Grid-scale batteries co-located with generation can help mitigate this misalignment. This paper explores the use of reinforcement learning (RL) for operating grid-scale batteries co-located with solar power. Our results show RL achieves an average of 61% (and up to 96%) of the approximate theoretical optimal (non-causal) operation, outperforming advanced control methods on average. Our findings suggest RL may be preferred when future signals are hard to predict. Moreover, RL has two significant advantages compared to simpler rules-based control: (1) that solar energy is more effectively shifted towards high demand periods, and (2) increased diversity of battery dispatch across different locations, reducing potential ramping issues caused by super-position of many similar actions.

Figures

Figures reproduced from arXiv: 2411.15422 by the authors.

Figure 1
Figure 1. Information flow in an RL environment. The grid-scale BESS problem can be formulated as a finite-horizon discrete-time MDP. The time-inhomogeneous MDP consists of a quintuple (T, S, A, r, {Pt}t). T is the number of time steps. A state st ∈ S at time t contains both the state-of-charge (Et) of the battery as well as exogenous variables derived from real-world datasets, which are the current and historical (up to four… view at source ↗
Figure 2
Figure 2. Profiles for two generator pnodes based in Los [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. Solar and LMP profile for Santa Cruz pnode. [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Mean (line) and confidence interval (shaded) [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]
Figure 5
Figure 5. Figure 5: The average hourly daily LMP (red, left axis) [PITH_FULL_IMAGE:figures/full_fig_p007_5.png]
Figure 6
Figure 6. Figure 6: The average energy stored in the battery across [PITH_FULL_IMAGE:figures/full_fig_p007_6.png]
Figure 7
Figure 7. Figure 7: Profits from selling solar immediately (Solar-sell), storing solar then selling (Solar-battery), and buying from grid then selling (Grid-battery). We used a small PV sizing (average solar power: 9.1MW). to its low degrees-of-freedom (i.e., one can only tune its buy and…
Figure 8
Figure 8. Figure 8: Mean (line) and confidence interval (shaded) [PITH_FULL_IMAGE:figures/full_fig_p008_8.png]
Figure 9
Figure 9. Figure 9: Cross-correlation between net load and demand. Demand is normalized so that periods of high demand have relative positive demand while low demand have negative demand. The average solar power is 9.1MW. The bottom plots zoom-in into the fourth day. 21-06), morning to no…
Figure 10
Figure 10. Figure 10: The mean (middle horizontal) line, extremal [PITH_FULL_IMAGE:figures/full_fig_p009_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 29 canonical work pages

  1. [1]

    An overview of genetic algorithms: Part 1, fundamentals

    David Beasley, David R Bull, and Ralph Robert Martin. An overview of genetic algorithms: Part 1, fundamentals. University computing, 15(2):56–69, 1993

    work page 1993

  2. [2]

    Dynamic programming

    Richard Bellman. Dynamic programming. science, 153(3731):34–37, 1966

    work page 1966

  3. [3]

    Deep reinforcement learning-based energy storage arbitrage with accurate lithium-ion battery degradation model

    Jun Cao, Dan Harrold, Zhong Fan, Thomas Morstyn, David Healey, and Kang Li. Deep reinforcement learning-based energy storage arbitrage with accurate lithium-ion battery degradation model. IEEE Transactions on Smart Grid, 11(5):4513–4521, 2020

    work page 2020

  4. [4]

    Microgrid reliability modeling and battery scheduling using stochastic linear programming

    Goncalo Cardoso, Michael Stadler, Afzal Siddiqui, Chris Marnay, Nicholas DeForest, Ana Barbosa-P ´ovoa, and Paulo Ferr ˜ao. Microgrid reliability modeling and battery scheduling using stochastic linear programming. Electric power systems research, 103:61–69, 2013

    work page 2013

  5. [5]

    The challenges of achieving a 100% renewable electricity system in the united states

    Paul Denholm, Douglas J Arent, Samuel F Baldwin, Daniel E Bilello, Gregory L Brinkman, Jaquelin M Cochran, Wesley J Cole, Bethany Frew, Vahan Gevorgian, Jenny Heeter, et al. The challenges of achieving a 100% renewable electricity system in the united states. Joule, 5(6):1331–1352, 2021

    work page 2021

  6. [6]

    Data-driven decision making in power systems with probabilistic guarantees: Theory and applications of chance-constrained optimization

    Xinbo Geng and Le Xie. Data-driven decision making in power systems with probabilistic guarantees: Theory and applications of chance-constrained optimization. Annual reviews in control, 47:341–363, 2019

    work page 2019

  7. [7]

    Long short-term memory

    Alex Graves and Alex Graves. Long short-term memory. Supervised sequence labelling with recurrent neural networks, pages 37–45, 2012

    2012

  8. [8]

    Deep-learning-and reinforcement-learning-based profitable strategy of a grid-level energy storage system for the smart grid

    Gwangwoo Han, Sanghun Lee, Jaemyung Lee, Kangyong Lee, and Joongmyeon Bae. Deep-learning-and reinforcement-learning-based profitable strategy of a grid-level energy storage system for the smart grid. Journal of Energy Storage , 41:102868, 2021

    work page 2021

Show all 34 references
  1. [9]

    Data-driven battery operation for energy arbitrage using rainbow deep reinforcement learning

    Daniel JB Harrold, Jun Cao, and Zhong Fan. Data-driven battery operation for energy arbitrage using rainbow deep reinforcement learning. Energy, 238:121958, 2022

    2022

  2. [10]

    Adaptation in natural and artificial systems: an introductory analysis with applications to biology, control, and artificial intelligence

    John H Holland. Adaptation in natural and artificial systems: an introductory analysis with applications to biology, control, and artificial intelligence . MIT press, 1992

    1992

  3. [11]

    Deep-reinforcement-learning-based capacity scheduling for pv-battery storage system

    Bin Huang and Jianhui Wang. Deep-reinforcement-learning-based capacity scheduling for pv-battery storage system. IEEE Transactions on Smart Grid, 12(3):2272–2283, 2020

    2020

  4. [12]

    The economic and reliability impacts of grid-scale storage in a high penetration renewable energy system

    Samuel C Johnson, Dimitri J Papageorgiou, Michael R Harper, Joshua D Rhodes, Kevin Hanson, and Michael E Webber. The economic and reliability impacts of grid-scale storage in a high penetration renewable energy system. Advances in Applied Energy, 3:100052, 2021

    2021

  5. [13]

    Storage futures study: Grid operational impacts of widespread storage deployment

    Jennie Jorgenson, A Will Frazier, Paul Denholm, and Nate Blair. Storage futures study: Grid operational impacts of widespread storage deployment. Technical report, National Renewable Energy Lab.(NREL), Golden, CO (United States), 2022

    2022

  6. [14]

    Elecsim: Monte-carlo open-source agent-based model to inform policy for long-term electricity planning

    Alexander Kell, Matthew Forshaw, and A Stephen McGough. Elecsim: Monte-carlo open-source agent-based model to inform policy for long-term electricity planning. In Proceedings of the Tenth ACM International Conference on Future Energy Systems , pages 556–565, 2019

    2019

  7. [15]

    Energy storage arbitrage under day-ahead and real-time price uncertainty

    Dheepak Krishnamurthy, Canan Uckun, Zhi Zhou, Prakash R Thimmapuram, and Audun Botterud. Energy storage arbitrage under day-ahead and real-time price uncertainty. IEEE Transactions on Power Systems , 33(1):84–93, 2017

    2017

  8. [16]

    Receding horizon control: model predictive control for state models

    Wook Hyun Kwon and Soo Hee Han. Receding horizon control: model predictive control for state models . Springer Science & Business Media, 2005

    2005

  9. [17]

    Use of battery storage systems for price arbitrage operations in the 15-and 60-min german intraday markets

    Dennis Metz and Jo ˜ao Tom ´e Saraiva. Use of battery storage systems for price arbitrage operations in the 15-and 60-min german intraday markets. Electric Power Systems Research, 160:27–36, 2018

    2018

  10. [18]

    Human-level control through deep reinforcement learning

    V olodymyr Mnih, Koray Kavukcuoglu, David Silver, Andrei A Rusu, Joel Veness, Marc G Bellemare, Alex Graves, Martin Riedmiller, Andreas K Fidjeland, Georg Ostrovski, et al. Human-level control through deep reinforcement learning. nature, 518(7540):529–533, 2015

    2015

  11. [19]

    Energy dispatch schedule optimization and cost benefit analysis for grid-connected, photovoltaic-battery storage systems

    A Nottrott, Jan Kleissl, and Byron Washom. Energy dispatch schedule optimization and cost benefit analysis for grid-connected, photovoltaic-battery storage systems. Renewable Energy, 55:230–240, 2013

    2013

  12. [20]

    Optimal operation of battery energy storage under uncertainty using data-driven distributionally robust optimization

    Seyed Shahin Parvar and Hamidreza Nazaripouya. Optimal operation of battery energy storage under uncertainty using data-driven distributionally robust optimization. Electric Power Systems Research , 211:108180, 2022

    2022

  13. [21]

    Markov decision processes: discrete stochastic dynamic programming

    Martin L Puterman. Markov decision processes: discrete stochastic dynamic programming. John Wiley & Sons, 2014

    2014

  14. [22]

    Stable-baselines3: Reliable reinforcement learning implementations

    Antonin Raffin, Ashley Hill, Adam Gleave, Anssi Kanervisto, Maximilian Ernestus, and Noah Dormann. Stable-baselines3: Reliable reinforcement learning implementations. Journal of Machine Learning Research, 22(268):1–8, 2021

    2021

  15. [23]

    A survey and comparison of leading-edge uncertainty handling methods for power grid modernization

    Sahar Rahim and Pierluigi Siano. A survey and comparison of leading-edge uncertainty handling methods for power grid modernization. Expert Systems with Applications, 204:117590, 2022

    2022

  16. [24]

    Uncertainty parameters of battery energy storage integrated grid and their modeling approaches: A review and future research directions

    MS Reza, MA Hannan, Pin Jern Ker, M Mansor, MS Hossain Lipu, MJ Hossain, and TM Indra Mahlia. Uncertainty parameters of battery energy storage integrated grid and their modeling approaches: A review and future research directions. Journal of Energy Storage, 68:107698, 2023

    2023

  17. [25]

    Exploiting battery storages with reinforcement learning: a review for energy professionals

    Rakshith Subramanya, Seppo A Sierla, and Valeriy Vyatkin. Exploiting battery storages with reinforcement learning: a review for energy professionals. IEEE Access, 10:54484–54506, 2022

    2022

  18. [26]

    Reinforcement learning: An introduction

    Richard S Sutton and Andrew G Barto. Reinforcement learning: An introduction. MIT press, 2018

    2018

  19. [27]

    The economics of planning electricity transmission to accommodate renewables: Using two-stage optimisation to evaluate flexibility and the cost of disregarding uncertainty

    Adriaan Hendrik Van Der Weijde and Benjamin F Hobbs. The economics of planning electricity transmission to accommodate renewables: Using two-stage optimisation to evaluate flexibility and the cost of disregarding uncertainty. Energy Economics , 34(6):2089–2101, 2012

    2012

  20. [28]

    The expected revenue of energy storage from energy arbitrage service based on the statistics of realistic market data

    Sadegh Vejdan and Santiago Grijalva. The expected revenue of energy storage from energy arbitrage service based on the statistics of realistic market data. In 2018 IEEE Texas Power and Energy Conference (TPEC), pages 1–6. IEEE, 2018

    2018

  21. [29]

    Stochastic programming models in energy

    Stein W Wallace and Stein-Erik Fleten. Stochastic programming models in energy. Handbooks in operations research and management science , 10:637–677, 2003

    2003

  22. [30]

    Energy storage arbitrage in real-time markets via reinforcement learning

    Hao Wang and Baosen Zhang. Energy storage arbitrage in real-time markets via reinforcement learning. In 2018 IEEE Power & Energy Society General Meeting (PESGM), pages 1–5. IEEE, 2018

    2018

  23. [31]

    Optimal scheduling of energy storage under forecast uncertainties

    Zeyu Wang, Ahlmahz Negash, and Daniel S Kirschen. Optimal scheduling of energy storage under forecast uncertainties. IET Generation, Transmission & Distribution, 11(17):4220–4226, 2017

    2017

  24. [32]

    Learning from delayed rewards

    Christopher Watkins. Learning from delayed rewards. 1989

    1989

  25. [33]

    Uncertainty models for stochastic optimization in renewable energy applications

    A Zakaria, Firas B Ismail, MS Hossain Lipu, and Mahammad Abdul Hannan. Uncertainty models for stochastic optimization in renewable energy applications. Renewable Energy, 145:1543–1571, 2020

    2020

  26. [34]

    Arbitraging variable efficiency energy storage using analytical stochastic dynamic programming

    Ningkun Zheng, Joshua Jaworski, and Bolun Xu. Arbitraging variable efficiency energy storage using analytical stochastic dynamic programming. IEEE Transactions on Power Systems, 37(6):4785–4795, 2022

    2022

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.