Pith. sign in

REVIEW 5 major objections 4 minor 29 references

Energy Management of Airport Service Electric Vehicles to Match Renewable Generation through Rollout Approach

T0 review · 5 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read A customized rollout controller for airport electric baggage vehicles schedules charging to match renewable generation and cuts daily operating cost by about 10% versus a greedy charging heuristic.

desk verdict The new ASEV dynamics model is the real contribution; the rollout application is standard and the scalability claim is unsubstantiated, but the paper deserves a major revision. read the letter →

arxiv 1908.07902 v1 pith:Q5EN3PMJ submitted 2019-08-21 eess.SY cs.SY

classification eess.SYcs.SY
keywords airportserviceelectricvehicleenergymanagementrolloutalgorithmnear-optimalcontrolrenewablegenerationmatchingbatterychargingstateofchargestochasticworkload
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

The paper tries to establish that the hard problem of scheduling a fleet of airport service electric vehicles (ASEVs) — when to charge each battery and which vehicle should serve which flight — admits a near-optimal solution that is actually computable. It proposes a new ASEV dynamics model that mixes discrete work/charge/idle states, a continuous battery state of charge, and stochastic service times, and then applies a customized rollout algorithm that uses two simple charging heuristics as base policies. The central empirical claim is that this rollout policy cuts total daily operating cost by about 10% relative to a greedy "charge whenever not full" heuristic on a 25-vehicle Bristol Airport case study for both summer and winter days. The point of caring: electrified airport ground fleets are carbon-intensive per vehicle-hour, and aligning their charging with cheap solar and off-peak grid tariffs reduces both cost and emissions. If true, the method makes matching airport EV load to renewable generation a practical, online decision rule rather than an intractable dynamic program.

What carries the argument

The load-bearing object is the customized rollout algorithm combined with the ASEV dynamics model. The model represents each vehicle by a discrete state $q$ (negative means working with $q$ stages remaining, 0 idling, 1 charging), a continuous state of charge, and a cycle-to-failure count; service times are discrete random variables from a truncated normal distribution. At each stage the rollout algorithm enumerates the set $N(S_t)$ of feasible next fleet states and uses an approximate cost-to-go $\tilde{J}$, computed by simulating the better of two heuristics to the end of the day, to select the next state that minimizes $\tilde{J}$. The work this does is to replace the exact dynamic-programming cost-to-go with a heuristic approximation, making a one-step lookahead decision that captures future charging opportunities while remaining online.

What would settle it

Run the rollout on a small fleet (say four vehicles) where exact dynamic programming is feasible and count the feasible next states at a busy midday stage for 25 vehicles; if the rollout cost exceeds the exact optimum by a wide margin, or if the enumeration cannot finish within a 5-minute stage, the near-optimality or tractability claim fails.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that the ASEV energy-management problem can be solved near-optimally by a custom rollout algorithm. Starting from a current fleet state, the algorithm enumerates the feasible next states reachable by choosing charge, idle, or work for each vehicle, estimates the remaining cost-to-go from each candidate by running the better of two heuristics (one that charges only when renewable energy is available, one that charges as soon as possible), and moves to the candidate with the lowest estimated total cost. Repeating this at each 5-minute stage yields an online control strategy that schedules both luggage transport and battery charging. In the Bristol Airport case studies the strategy reduces total daily cost by about 10% compared with greedy charging, achieves this by buying much less grid energy during peak tariff periods, and adapts to flight cancellations on short notice.

Load-bearing premise

The algorithm's tractability depends on being able to enumerate the full set of feasible next fleet states at each 5-minute stage; the paper does not prune or decompose the action space, and with 25 vehicles and three actions per vehicle this candidate set can become astronomically large.

Editorial extensions

If this is right

  • An airport operating under similar two-tier tariffs and PV profiles can expect roughly a 10% daily cost reduction from rollout control instead of greedy charging.
  • The charging load shifts toward cheap solar and off-peak grid windows, making the fleet's load curve friendlier to local renewable consumption and to the grid.
  • The receding-horizon rollout re-plans after flight cancellations, so schedules need not be frozen a day ahead.
  • Battery degradation cost rises because the controller allows deeper discharges, even though total cost falls; the net benefit depends on the relative weight of battery replacement cost.

Reading between the lines

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

  • The same rollout template — discrete task assignment, SoC dynamics, and two base heuristics — should transfer to other centrally dispatched electric fleets such as port equipment, warehouse robots, and shuttle buses whose tasks are stochastic and whose charging is price-responsive.
  • The paper does not bound the optimality gap; comparing rollout against an exact solution for a small fleet would show how much of the 10% saving comes from the rollout itself rather than from the underlying heuristics.
  • A safety-SoC constraint that prevents the renewable-matching heuristic from becoming infeasible might preserve most of the PV benefit while avoiding the failure seen in the summer scenario.
  • Because the saving comes partly from deferring charge to off-peak hours, the method's value depends on tariff spread and PV availability; a sensitivity study across tariff ratios would show where rollout ceases to beat greedy charging.
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

5 major / 4 minor

Summary. The paper proposes a dynamics model for airport service electric vehicles (ASEVs) and a rollout-based near-optimal control strategy that jointly decides battery charging and luggage-transport task assignment. The model represents each ASEV by a discrete state (working, charging, idling), a continuous SoC, and a battery-cycle-failure count, with stochastic service-time workload modeled by a discretized truncated normal distribution. The control problem is formulated as a finite-horizon cost minimization over energy purchase, battery degradation, and a terminal SoC penalty. A customized rollout algorithm is built on two base heuristics (renewable-matching and greedy charging), enumerating all feasible next states at each stage and selecting the state with the minimum approximate cost-to-go. Case studies at Bristol Airport with 25 ASEVs on typical summer and winter days report a total daily cost reduction of approximately 10% relative to the greedy-charging heuristic, and a scenario with flight cancellation is used to show adaptability to schedule changes.

Significance. If the central claims hold, the paper would make a useful contribution by addressing a genuinely new application area — centrally controlled airport ground fleets — and by adapting rollout, a classic approximate dynamic programming technique, to a hybrid stochastic system. The ASEV dynamics model with discrete working/charging/idle modes, continuous SoC, and stochastic workload is a reasonable and clearly described formulation. The cost model, including renewable-matching incentives and battery degradation, is plausible and aligned with current interest in grid-friendly EV charging. However, the paper's main quantitative claims — that rollout "effectively overcomes the curse of dimensionality" and yields a "near-optimal" strategy — are not substantiated by the evidence actually presented. The algorithm's tractability for a 25-vehicle fleet is not demonstrated, the treatment of stochasticity is underspecified, and the absence of an optimality benchmark means the reported cost savings over the base heuristic are expected consequences of rollout's construction rather than evidence of near-optimality.

major comments (5)
  1. [Section 4, Step 2 (Eqs. (19)-(20))] The enumeration of the set N(S_t) of all feasible next states is not shown to be tractable for the 25-ASEV fleet used in the case study. With three control decisions per vehicle, the unfiltered product set has 3^25 ≈ 8.5×10^11 elements; constraints (9)-(12) and (16)-(17) only exclude vehicles that are already committed to work or at SoC bounds, leaving a potentially enormous set of free vehicles. The paper provides no bound, pruning rule, decomposition, or complexity analysis for N(S_t). Without a tractable enumeration, the central claim that the rollout "effectively overcomes the curse of dimensionality" is unsupported, and the algorithm as described may not be executable for the reported case studies.
  2. [Section 4, Steps 1-2 and Section 3.1] The stochastic workload is not integrated into the rollout algorithm in a well-defined way. Transition (11) uses the random variable w_jt, but the rollout description does not state whether J(S) is evaluated using expected values, Monte Carlo samples, scenario trees, or a single realization. The approximate cost-to-go of a candidate next state depends on the future workload distribution, and without an explicit stochastic rollout formulation the numerical results in Section 5 are not reproducible and the algorithm's behavior under uncertainty is not rigorously specified.
  3. [Section 5, Scenarios 1-3] The simulation study is missing several key input parameters and is not reproducible from the manuscript: the truncated-normal workload parameters (μ, σ, a, b) are never specified; the two-tier grid tariff shown qualitatively in Fig. 6 is not given numerically; the battery-degradation cost coefficient is only cited to [23] without a value or formula; and the PV generation profiles are referenced to external websites but not provided in a usable form. In addition, each scenario is a single run of a stochastic process, so the reported 10% cost reduction lacks error bars or a statistical comparison across multiple workload realizations.
  4. [Section 4, Figs. 4 and 5] The pseudo-code for the two base heuristic strategies is referenced as Fig. 4 and Fig. 5, but the figures are not present in the manuscript text provided for review. Since the rollout algorithm's construction depends on the exact behavior of these heuristics, the algorithm cannot be implemented or verified from the description alone. The heuristics must be specified in the text, either as pseudo-code or as a precise step-by-step procedure.
  5. [Section 5 and Section 6] The claim that the rollout algorithm yields a "near-optimal" control strategy is not supported by any comparison with an optimal solution or a lower bound. The approximately 10% cost improvement over the greedy-charging heuristic is a built-in property of rollout: rollout with a base heuristic is designed to improve upon that heuristic, so observing an improvement does not demonstrate closeness to optimality. To justify the near-optimal label, the authors should compare against an exact dynamic-programming solution on a small fleet size, or provide a formal bound on the rollout cost-to-go approximation.
minor comments (4)
  1. [Eq. (1)] The notation in the truncated normal density is inconsistent: the text states that a and b are the upper and lower bounds, respectively, but the conditions in (1) use "b < w < a" and the later integration range [b,a] implies that b is the lower bound and a the upper bound. Please correct the definitions and the conditions in (1).
  2. [Section 4, Eq. (20)] The sentence "S_t, N(S_t), and J(S) are defined in (19)" is slightly inaccurate; J(S) is defined in the text accompanying (19), not in the equation itself. Please rephrase to avoid ambiguity.
  3. [Section 5, Scenario 3] The statement that the rollout algorithm is "proven to be adaptive" is too strong based on a single flight-cancellation scenario; "demonstrated" or "illustrated" would be more appropriate.
  4. [Section 5, Tables I-III] The tables show control decisions for one example ASEV, but the formatting is difficult to read. Please present the tables with clear column headers (e.g., Start time, End time, Activity) and consider aligning rows for charging and working periods.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the ASEV cost model, rollout construction, and case-study comparison form a self-contained chain; the 10% saving is a simulated comparison, not a construction-level identity.

full rationale

The derivation chain is: (i) Eqs. (4)-(8) define an explicit cost model from energy prices, renewable availability, battery degradation, and terminal SoC; (ii) Eqs. (3), (9)-(18) define independent ASEV state dynamics; (iii) Section 4 constructs a rollout controller whose approximate cost-to-go is obtained by simulating one of two heuristics; and (iv) Section 5 evaluates the resulting policy under this cost model. The central quantitative claim—a roughly 10% cost reduction over the greedy-charging heuristic—is not a prediction derived from an input; it is the output of simulations on Bristol Airport data using the independently specified cost equations. Rollout is, by design, a one-step improvement scheme over a base heuristic, which makes an improvement over the base plausible, but the paper does not define the improvement into the answer: the costs are computed from the cost model and from the state trajectories, not inferred from the algorithm definition alone. No parameter is fitted to the reported outputs. Citations to Bertsekas [20] and the battery-degradation source [23] are external and not author-overlapping, so no self-citation chain supports the result. The nearest concerns are validation and complexity issues rather than circularity: the label 'near-optimal' is not checked against an exact optimum, and Section 4 Step 2's enumeration of N(S_t) is not accompanied by a complexity bound or pruning analysis for 25 ASEVs. These are legitimate correctness or reproducibility caveats, but they do not make any equation equivalent to its own input. Hence no circular step can be identified under the required standard.

Assumptions & free parameters 5 free parameters · 8 assumptions · 0 invented entities

The model rests on a small set of engineering assumptions typical of EV scheduling papers. No novel physical entities are introduced. The main unresolved choices are the workload distribution parameters and the battery degradation coefficient, neither of which is specified numerically in the paper.

free parameters (5)
  • Workload distribution parameters (mu, sigma, a, b)
    The truncated normal model in Eq. (1)-(2) requires these values for each flight's service time. They are not stated in Section 5, so the stochastic workload is irreproducible.
  • Grid tariff schedule = two-tier prices shown in Fig. 6, numeric values absent
    Energy cost depends on the off-peak and peak grid prices, which are only presented as a figure, not as numbers.
  • PV generation time series = external websites [25] and [26]
    Summer and winter PV output profiles are cited to URLs but not included; the renewable matching behavior depends on them.
  • Battery degradation cost coefficient = from [23]
    The linear degradation coefficient in Eq. (5) is borrowed from Bordin et al. without stating its value, affecting the cost trade-off.
  • Flight schedule = Bristol Airport typical day in 2018 [24]
    The schedule of 88 arrivals and 86 departures is referenced to a website and not reproduced, but all workload is derived from it.
assumptions (8)
  • domain assumption The time to serve a flight follows a truncated normal distribution, and the discretized probability is computed from the parent CDF.
    Section 3.1 justifies the normal as a default; Eq. (1) as printed has an inverted CDF denominator, so the distribution is not correctly specified.
  • domain assumption Charging power is constant at 22 kW and battery efficiency is 90%.
    Section 5 sets these values; real fast chargers taper as SoC approaches the cap, which is ignored.
  • domain assumption Each flight is served by exactly one ASEV, and once working, an ASEV cannot be reassigned until the job completes.
    Transitions (11)-(17) force a working ASEV to finish; no job splitting or multi-trailer service is modeled.
  • domain assumption At each stage, renewable energy is consumed first, and grid energy tops up the remainder at a time-of-use tariff.
    Eq. (4) applies priority to renewable energy, which is an accounting rule rather than a physical constraint.
  • domain assumption Battery degradation cost is a linear function of discharge energy, SoC, and cycles to failure.
    Eq. (5) uses a linear f with coefficient from [23]; nonlinear aging effects are not modeled.
  • domain assumption Any SoC shortfall at the end of the day is penalized linearly at the final grid price.
    Eq. (7) imposes a terminal cost to encourage full charge by day end.
  • domain assumption PV output, grid tariffs, and the flight schedule are known exactly when the day's control is computed.
    The case study uses deterministic profiles, despite the workload model being stochastic; forecast errors are not considered.
  • domain assumption Service times for different flights are independent draws.
    Section 3.1 defines a distribution per flight with no correlation between flights or stages.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Energy Management of Airport Service Electric Vehicles to Match Renewable Generation through Rollout Approach." pith.science (2026). https://pith.science/paper/Q5EN3PMJ

@misc{pith2026190807902,
  author       = {Pith},
  title        = {Pith review of: Energy Management of Airport Service Electric Vehicles to Match Renewable Generation through Rollout Approach},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/Q5EN3PMJ}},
  note         = {Machine review of arXiv:1908.07902}
}
read the original abstract

Traditional diesel-based airport service vehicles are characterized by a heavy-duty, high-usage-frequency nature and a high carbon intensity per vehicle per hour. Transforming these vehicles into electric vehicles would reduce CO2 emissions and potentially save energy costs in the context of rising fuel prices, if a proper energy management of airport service electric vehicles (ASEVs) is performed. To perform such an energy management, this paper proposes a new customized rollout approach, as a near-optimal control method for a new ASEV dynamics model, which models the ASEV states, their transitions over time, and how control decisions affect them. The rollout approach yields a near-optimal control strategy for the ASEVs to transport luggage and to charge batteries, with the objective to minimize the operation cost, which incentivizes the charging of the ASEVs to match renewable generation. Case studies demonstrate that the rollout approach effectively overcomes the "curse of dimensionality". On both typical summer and winter days, the rollout algorithm results in a total cost approximately 10% less than that of the underlying "greedy charging" heuristic, which charges a battery whenever its state of charge is not the maximum. The rollout algorithm is proven to be adaptive towards flight schedule changes at short notice.

Figures

Figures reproduced from arXiv: 1908.07902 by the authors.

Figure 1
Figure 1. Flowchart of the models and the optimal control approach [PITH_FULL_IMAGE:figures/full_fig_p006_1.png] view at source ↗
Figure 2
Figure 2. presents an overview of the ASEV dynamics model [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. presents a state transition graph describing the relation among 𝑞௜௧, 𝑤௝௧, and 𝑢௜௧ [PITH_FULL_IMAGE:figures/full_fig_p012_3.png] view at source ↗
Figures from the paper (8 more)
Figure 4
Figure 4. Figure 4: Pseudo-code for heuristic i), i.e. the “renewable matching” heuristic. [PITH_FULL_IMAGE:figures/full_fig_p015_4.png]
Figure 5
Figure 5. Figure 5: Pseudo-code for heuristic ii), i.e. the “greedy charging” heuristic. [PITH_FULL_IMAGE:figures/full_fig_p015_5.png]
Figure 6
Figure 6. Figure 6: The two-tier tariff for one day. To validate the algorithm, this section performs three sets of simulations. The first set is a comparison among the two heuristic algorithms and the rollout algorithm on a typical summer day. The second set is a comparison among the two…
Figure 7
Figure 7. Figure 7: The PV generation power of the typical sunny summer day. [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]
Figure 8
Figure 8. Figure 8: The cost of Bristol Airport under Heuristic ii) and rollout algorithm on the sunny summer day. [PITH_FULL_IMAGE:figures/full_fig_p019_8.png]
Figure 9
Figure 9. Figure 9: The PV generation power of the typical sunny winter day. [PITH_FULL_IMAGE:figures/full_fig_p021_9.png]
Figure 10
Figure 10. Figure 10: The cost of Bristol Airport under Heuristic ii) and rollout algorithm on the winter sunny day. [PITH_FULL_IMAGE:figures/full_fig_p022_10.png]
Figure 11
Figure 11. Figure 11: The cost of Bristol Airport under Heuristic ii) and rollout algorithm when flights are cancelled at 10.45am. [PITH_FULL_IMAGE:figures/full_fig_p023_11.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

29 extracted references · 29 canonical work pages

  1. [23]

    A linear programming approach for battery degradation analysis and optimization in offgrid power systems with solar energy integration,

    C. Bordin, H. O. Anuta, A. Crossland, I. L. Gutierrez, C. J. Dent, and D. Vigo, "A linear programming approach for battery degradation analysis and optimization in offgrid power systems with solar energy integration," Renewable Energy, vol. 101, pp. 417-430, 2017/02/01/ 2017

  2. [1]

    Electric Vehicles,

    "Electric Vehicles," House of Parliament, 2010, Available: http://researchbriefings.files.parliament.uk/documents/POST-PN-365/POST-PN-365.pdf

  3. [2]

    Two-Stage Energy Management for Office Buildings With Workplace EV Charging and Renewable Energy,

    D. Wu, H. Zeng, C. Lu, and B. Boulet, "Two-Stage Energy Management for Office Buildings With Workplace EV Charging and Renewable Energy," IEEE Transactions on Transportation Electrification, vol. 3, no. 1, pp. 225-237, 2017

  4. [3]

    Transactive Real-time Electric Vehicle Charging Management for Commercial Buildings with PV On-site Generation,

    Z. Liu, Q. Wu, M. Shahidehpour, C. Li, S. Huang, and W. Wei, "Transactive Real-time Electric Vehicle Charging Management for Commercial Buildings with PV On-site Generation," IEEE Transactions on Smart Grid, pp. 1-1, 2018. 26

  5. [4]

    Optimal Charging Control of Energy Storage and Electric Vehicle of an Individual in the Internet of Energy With Energy Trading,

    C. Lin, D. Deng, C. Kuo, and Y. Liang, "Optimal Charging Control of Energy Storage and Electric Vehicle of an Individual in the Internet of Energy With Energy Trading," IEEE Transactions on Industrial Informatics, vol. 14, no. 6, pp. 2570-2578, 2018

  6. [5]

    Model Predictive Charging Control of In- Vehicle Batteries for Home Energy Management Based on Vehicle State Prediction,

    A. Ito, A. Kawashima, T. Suzuki, S. Inagaki, T. Yamaguchi, and Z. Zhou, "Model Predictive Charging Control of In- Vehicle Batteries for Home Energy Management Based on Vehicle State Prediction," IEEE Transactions on Control Systems Technology, vol. 26, no. 1, pp. 51-64, 2018

  7. [6]

    Stochastic Optimal Energy Management of Smart Home With PEV Energy Storage,

    X. Wu, X. Hu, X. Yin, and S. J. Moura, "Stochastic Optimal Energy Management of Smart Home With PEV Energy Storage," IEEE Transactions on Smart Grid, vol. 9, no. 3, pp. 2065-2075, 2018

  8. [7]

    A Probabilistic Energy Management Scheme for Renewable-Based Residential Energy Hubs,

    M. Rastegar, M. Fotuhi-Firuzabad, H. Zareipour, and M. Moeini-Aghtaieh, "A Probabilistic Energy Management Scheme for Renewable-Based Residential Energy Hubs," IEEE Transactions on Smart Grid, vol. 8, no. 5, pp. 2217- 2227, 2017

Show all 29 references
  1. [8]

    Cooperative Distributed Demand Management for Community Charging of PHEV/PEVs Based on KKT Conditions and Consensus Networks,

    N. Rahbari-Asr and M. Chow, "Cooperative Distributed Demand Management for Community Charging of PHEV/PEVs Based on KKT Conditions and Consensus Networks," IEEE Transactions on Industrial Informatics, vol. 10, no. 3, pp. 1907-1916, 2014

  2. [9]

    Optimal Behavior of Electric Vehicle Parking Lots as Demand Response Aggregation Agents,

    M. Shafie-khah et al., "Optimal Behavior of Electric Vehicle Parking Lots as Demand Response Aggregation Agents," IEEE Transactions on Smart Grid, vol. 7, no. 6, pp. 2654-2665, 2016

  3. [10]

    Optimal Management for Parking-Lot Electric Vehicle Charging by Two-Stage Approximate Dynamic Programming,

    L. Zhang and Y. Li, "Optimal Management for Parking-Lot Electric Vehicle Charging by Two-Stage Approximate Dynamic Programming," IEEE Transactions on Smart Grid, vol. 8, no. 4, pp. 1722-1730, 2017

  4. [11]

    A Transactive Operating Model for Smart Airport Parking Lots,

    M. Maigha and M. L. Crow, "A Transactive Operating Model for Smart Airport Parking Lots," IEEE Power and Energy Technology Systems Journal, vol. 5, no. 4, pp. 157-166, 2018

  5. [12]

    Technology solutions to mitigate electricity cost for electric vehicle DC fast charging,

    M. Muratori et al., "Technology solutions to mitigate electricity cost for electric vehicle DC fast charging," Applied Energy, vol. 242, pp. 415-423, 2019/05/15/ 2019. 27

  6. [13]

    Two-stage stochastic optimization for cost-minimal charging of electric vehicles at public charging stations with photovoltaics,

    K. Seddig, P. Jochem, and W. Fichtner, "Two-stage stochastic optimization for cost-minimal charging of electric vehicles at public charging stations with photovoltaics," Applied Energy, vol. 242, pp. 769-781, 2019/05/15/ 2019

  7. [14]

    Balanced charging strategies for electric vehicles on power systems,

    S.-K. Moon and J.-O. Kim, "Balanced charging strategies for electric vehicles on power systems," Applied Energy, vol. 189, pp. 44-54, 2017/03/01/ 2017

  8. [15]

    An integrated optimisation approach to airport ground operations to foster sustainability in the aviation sector,

    M. Weiszer, J. Chen, and G. Locatelli, "An integrated optimisation approach to airport ground operations to foster sustainability in the aviation sector," Applied Energy, vol. 157, pp. 567-582, 2015/11/01/ 2015

  9. [16]

    Review of stochastic hybrid systems with applications in biological systems modeling and analysis,

    X. Li, O. Omotere, L. Qian, and E. R. Dougherty, "Review of stochastic hybrid systems with applications in biological systems modeling and analysis," EURASIP journal on bioinformatics & systems biology, vol. 2017, no. 1, pp. 8-8, 2017

  10. [17]

    Bradshaw Electric Vehicles

    (2016). Bradshaw Electric Vehicles. Available: https://www.airport-suppliers.com/supplier/bradshaw-electric- vehicles/

  11. [18]

    Smith Electric Vehicles

    (2016). Smith Electric Vehicles. Available: https://www.airport-technology.com/contractors/groundequipment/smith- electric/

  12. [19]

    Alke Electric Vehicles

    (2017). Alke Electric Vehicles. Available: https://www.alke.com/doc/alke-atx-electric-vehicles-catalog-eng.pdf

  13. [20]

    D. P. Bertsekas, Dynamic Programming and Optimal Control, 4th ed. Athena Scientific, 2017

  14. [21]

    R. S. Sutton and A. G. Barto, Reinforcement learning: an introduction, 2nd ed. The MIT Press, 2017

  15. [22]

    The Truncated Normal Distribution,

    J. Burkardt, "The Truncated Normal Distribution," Florida State University, 2014, Available: https://people.sc.fsu.edu/~jburkardt/presentations/truncated_normal.pdf

  16. [24]

    Bristol Airport Arrivals and Departures

    (2018, 16 Nov 2018). Bristol Airport Arrivals and Departures. Available: https://www.bristolairport.co.uk/arrivals- and-departures/arrivals 28

  17. [25]

    Weather & Local Environment

    (2019). Weather & Local Environment. Available: http://www.uq.edu.au/solarenergy/pv-array/weather

  18. [26]

    Live monitoring of the UK electricity National Grid

    (2019). Live monitoring of the UK electricity National Grid. Available: http://gridwatch.co.uk/

  19. [27]

    Understanding electric car charging

    (2018). Understanding electric car charging. Available: https://www.spiritenergy.co.uk/kb-ev-understanding-electric- car-charging

  20. [28]

    On Generation Schedule Tracking of Wind Farms With Battery Energy Storage Systems,

    X. Zhang, Y. Yuan, L. Hua, Y. Cao, and K. Qian, "On Generation Schedule Tracking of Wind Farms With Battery Energy Storage Systems," IEEE Transactions on Sustainable Energy, vol. 8, no. 1, pp. 341-353, 2017

  21. [29]

    Electric car boom in Europe starts this year

    (2019). Electric car boom in Europe starts this year. Available: https://pushevs.com/2019/01/28/electric-car-boom-in- europe-starts-this-year/

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.