pith. sign in

arxiv: 2604.06692 · v1 · submitted 2026-04-08 · 📡 eess.SY · cs.SY· math.OC

A Markov Decision Process Framework for Enhancing Power System Resilience during Wildfires under Decision-Dependent Uncertainty

Xinyi Zhao , Prasanna Raut , Chaoyue Zhao , Alexandre Moreira This is my paper

Pith reviewed 2026-05-10 17:58 UTC · model grok-4.3

classification 📡 eess.SY cs.SYmath.OC
keywords Markov Decision Processpower system resiliencewildfire mitigationsafety power shutoffsdistribution networksapproximate dynamic programmingdecision-dependent uncertainty
0
0 comments X

The pith

A Markov Decision Process optimizes power switching actions during wildfires to minimize total operational costs.

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

This paper develops a state-based decision framework using Markov Decision Processes to schedule safety power shutoffs in distribution networks facing wildfire threats. Network topologies are treated as Markov states whose transitions depend on both external weather conditions and the internal effects of power flows. The aim is to reduce the sum of operational expenses, including costs from de-energized loads and equipment risks, across the full duration of a fire event. An approximate dynamic programming method based on post-decision states addresses the large state and action spaces that arise in realistic grids. Tests on 54-bus and 138-bus systems illustrate that the model produces workable policies for different network sizes.

Core claim

Representing network topologies as Markov states with transitions driven by exogenous weather and endogenous power flow dynamics allows an MDP formulation that optimizes switching sequences to minimize total operational costs throughout a wildfire event; the resulting policies are computed efficiently via approximate dynamic programming on post-decision states.

What carries the argument

Markov Decision Process in which network topologies are states and transitions combine weather-driven exogenous changes with power-flow-driven endogenous changes, solved by approximate dynamic programming on post-decision states.

If this is right

  • The framework produces time-varying shutoff schedules that trade off immediate load loss against long-term ignition and damage costs.
  • Approximate dynamic programming on post-decision states renders the approach computationally feasible for systems up to at least 138 buses.
  • The same state representation supports repeated re-optimization as new weather observations arrive.
  • Cost-minimizing policies differ across grid topologies, showing the value of tailoring decisions to each network's configuration.

Where Pith is reading between the lines

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

  • The same MDP structure could be adapted to other time-evolving hazards such as storms or heat waves that also alter line failure probabilities.
  • Embedding real-time sensor data into the transition probabilities would allow the model to update policies without full re-solving.
  • Extending the state space to include crew locations or repair resources could turn the framework into a joint resilience and restoration planner.

Load-bearing premise

That the uncertainty in network conditions during a wildfire can be captured sufficiently well by Markov state transitions that depend only on weather and power flows.

What would settle it

Deploy the computed shutoff policy on a live distribution feeder during an actual wildfire and compare the realized total costs and ignition events against the model's predicted minimum.

Figures

Figures reproduced from arXiv: 2604.06692 by Alexandre Moreira, Chaoyue Zhao, Prasanna Raut, Xinyi Zhao.

Figure 1
Figure 1. Figure 1: Line availability probability for the 54-bus network [PITH_FULL_IMAGE:figures/full_fig_p008_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Final network topology under the DDU method at hour [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: Distribution of total daily load shedding over 20-hour [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Distribution of total daily load shedding over 20-hour [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
read the original abstract

Wildfires pose an increasing threat to the safety and reliability of power systems, particularly in distribution networks located in fire-prone regions. To mitigate ignition risk from electrical infrastructure, utilities often employ safety power shutoffs, which proactively de-energize high-risk lines during hazardous weather and restore them once conditions improve. While this strategy can result in temporary load loss, it helps prevent equipment damage and wildfire ignition development in the system. In this paper, we develop a state-based decision-making framework to optimize such switching actions over time, with the goal of minimizing total operational costs throughout a wildfire event. The model represents network topologies as Markov states, with transitions influenced by both exogenous weather conditions and endogenous power flow dynamics. To address the computational challenges posed by the large state and action spaces, we propose an approximate dynamic programming algorithm based on post-decision states. The effectiveness and scalability of the proposed approach are demonstrated through case studies on 54-bus and 138-bus distribution systems, showcasing its potential for enhancing wildfire resilience across different grid configurations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

1 major / 0 minor

Summary. The paper develops an MDP framework for optimizing safety power shutoff actions in distribution networks during wildfires to minimize total operational costs. Network topologies are represented as discrete Markov states whose transitions depend on exogenous weather conditions and endogenous power-flow quantities; an approximate dynamic programming algorithm using post-decision states is proposed to solve the resulting large-scale problem. Effectiveness is illustrated via case studies on 54-bus and 138-bus test systems.

Significance. If the central modeling assumptions hold, the work supplies a computationally tractable, state-based policy for utilities facing decision-dependent wildfire risk, directly addressing a growing operational challenge. The use of post-decision states to mitigate the curse of dimensionality is a constructive algorithmic contribution, and the demonstration across two differently sized distribution systems provides initial evidence of scalability.

major comments (1)
  1. Abstract and modeling section: The claim that the MDP correctly optimizes shutoff policies under decision-dependent uncertainty rests on the unvalidated premise that topology transitions are Markovian and fully captured by weather plus power-flow dynamics. The 54-bus and 138-bus case studies rely on synthetic transition probabilities with no reported calibration, sensitivity analysis, or comparison against physics-based fire-spread models; without such evidence the optimality guarantees and resilience improvements cannot be assessed.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the detailed and constructive feedback. The concern about validating the Markovian transitions and the use of synthetic probabilities is well-taken; we clarify our modeling rationale and commit to targeted revisions that strengthen the presentation without altering the core contribution.

read point-by-point responses

  1. Referee: Abstract and modeling section: The claim that the MDP correctly optimizes shutoff policies under decision-dependent uncertainty rests on the unvalidated premise that topology transitions are Markovian and fully captured by weather plus power-flow dynamics. The 54-bus and 138-bus case studies rely on synthetic transition probabilities with no reported calibration, sensitivity analysis, or comparison against physics-based fire-spread models; without such evidence the optimality guarantees and resilience improvements cannot be assessed.

    Authors: The Markov property is a deliberate modeling choice: the state vector explicitly includes the current network topology (as a discrete Markov state) together with exogenous weather conditions, so that the transition kernel depends on both weather and the endogenous power-flow quantities that result from the chosen switching action. This structure directly encodes decision-dependent uncertainty. We acknowledge that the numerical case studies employ illustrative transition probabilities chosen to demonstrate scalability rather than calibrated from field data. In the revised manuscript we will (i) add an explicit subsection on calibration approaches that leverage historical wildfire, weather, and outage records, (ii) include a sensitivity analysis that perturbs the transition probabilities over plausible ranges and reports the resulting policy and cost variations, and (iii) cite representative physics-based fire-spread models (e.g., those based on Rothermel or cellular automata) while clarifying that our framework is intended to accept transition probabilities generated by such models. Because the paper focuses on the decision-making algorithm rather than fire physics, a full empirical comparison lies outside the present scope; the added discussion will make this boundary explicit. revision: partial

Circularity Check

0 steps flagged

MDP modeling framework for wildfire resilience contains no circular reductions

full rationale

The paper introduces an MDP in which network topologies are treated as discrete Markov states whose transitions depend on exogenous weather and endogenous power-flow quantities. This representation is an explicit modeling assumption rather than a derived result obtained by fitting parameters to data or by self-referential definition. The subsequent approximate dynamic programming algorithm is a standard solution technique for the resulting large-scale MDP and does not rely on any fitted prediction that is forced by construction. Case studies on the 54-bus and 138-bus systems serve only to illustrate computational scalability under synthetic transition probabilities; they do not close a loop in which an output is redefined as an input. Consequently the claimed optimization framework remains self-contained and independent of its own outputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the assumption that the power system during a wildfire can be modeled as a Markov Decision Process with states representing network topologies and transitions driven by weather and power flow; no free parameters or invented entities are explicitly listed in the abstract.

axioms (1)
  • domain assumption Network topologies can be represented as Markov states with transitions influenced by both exogenous weather conditions and endogenous power flow dynamics.
    Directly stated in the abstract as the basis for the state-based decision-making framework.

pith-pipeline@v0.9.0 · 5493 in / 1229 out tokens · 47116 ms · 2026-05-10T17:58:01.503197+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

14 extracted references · 14 canonical work pages

  1. [1]

    Human exposure and sensitivity to globally extreme wildfire events,

    D. M. Bowman, G. J. Williamson, J. T. Abatzoglou, C. A. Kolden, M. A. Cochrane, and A. M. Smith, “Human exposure and sensitivity to globally extreme wildfire events,”Nature ecology & evolution, vol. 1, no. 3, p. 0058, 2017

    work page 2017

  2. [2]

    arXiv preprint arXiv:2501.17880 , year=

    S. T. Seydi, “Assessment of the january 2025 los angeles county wildfires: A multi-modal analysis of impact, response, and population exposure,”arXiv preprint arXiv:2501.17880, 2025

    work page arXiv 2025

  3. [3]

    Wildfire mitigation plans in power systems: A literature review,

    D. A. Z. Vazquez, F. Qiu, N. Fan, and K. Sharp, “Wildfire mitigation plans in power systems: A literature review,”IEEE Transactions on Power Systems, vol. 37, no. 5, pp. 3540–3551, 2022

    work page 2022

  4. [4]

    A review of public safety power shutoffs (psps) for wildfire mitigation: Policies, practices, models and data sources,

    C. Huang, et al, “A review of public safety power shutoffs (psps) for wildfire mitigation: Policies, practices, models and data sources,”IEEE Transactions on Energy Markets, Policy and Regulation, vol. 1, no. 3, pp. 187–197, 2023

    work page 2023

  5. [5]

    Population exposure to pre-emptive de-energization aimed at averting wildfires in northern california,

    J. T. Abatzoglou, C. M. Smith, D. L. Swain, T. Ptak, and C. A. Kolden, “Population exposure to pre-emptive de-energization aimed at averting wildfires in northern california,”Environmental Research Letters, vol. 15, no. 9, p. 094046, 2020

    work page 2020

  6. [6]

    Optimally scheduling public safety power shutoffs,

    A. Lesage-Landry, F. Pellerin, D. S. Callaway, and J. A. Taylor, “Optimally scheduling public safety power shutoffs,”Stochastic Systems, vol. 13, no. 4, pp. 438–456, 2023

    work page 2023

  7. [7]

    Distribution system operation amidst wildfire-prone climate conditions under decision-dependent line availability uncertainty,

    A. Moreira, F. Pianc ´o, B. Fanzeres, A. Street, R. Jiang, C. Zhao, and M. Heleno, “Distribution system operation amidst wildfire-prone climate conditions under decision-dependent line availability uncertainty,”IEEE Transactions on Power Systems, vol. 39, no. 5, pp. 6522–6538, 2024

    work page 2024

  8. [8]

    Enhancing power system operational resilience against wildfires,

    M. Abdelmalak and M. Benidris, “Enhancing power system operational resilience against wildfires,”IEEE Transactions on Industry Applica- tions, vol. 58, no. 2, pp. 1611–1621, 2022

    work page 2022

  9. [9]

    Decision-dependent uncertainty-aware distribution system planning un- der wildfire risk,

    F. Pianc ´o, A. Moreira, B. Fanzeres, R. Jiang, C. Zhao, and M. Heleno, “Decision-dependent uncertainty-aware distribution system planning un- der wildfire risk,”IEEE Transactions on Power Systems, 2025

    work page 2025

  10. [10]

    Markov decision process-based resilience enhancement for distribution systems: An approximate dynamic programming approach,

    C. Wang, P. Ju, S. Lei, Z. Wang, F. Wu, and Y . Hou, “Markov decision process-based resilience enhancement for distribution systems: An approximate dynamic programming approach,”IEEE Transactions on Smart Grid, vol. 11, no. 3, pp. 2498–2510, 2019

    work page 2019

  11. [11]

    Mdp-based dis- tribution network reconfiguration with renewable distributed generation: Approximate dynamic programming approach,

    C. Wang, S. Lei, P. Ju, C. Chen, C. Peng, and Y . Hou, “Mdp-based dis- tribution network reconfiguration with renewable distributed generation: Approximate dynamic programming approach,”IEEE Transactions on Smart Grid, vol. 11, no. 4, pp. 3620–3631, 2020

    work page 2020

  12. [12]

    A markov decision process to enhance power system operation resilience during hurricanes,

    M. Abdelmalak and M. Benidris, “A markov decision process to enhance power system operation resilience during hurricanes,” in2021 IEEE Power & Energy Society General Meeting (PESGM). IEEE, 2021, pp. 01–05. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2021 11

    work page 2021

  13. [13]

    Distributionally robust distribution network configuration under random contingency,

    S. Babaei, R. Jiang, and C. Zhao, “Distributionally robust distribution network configuration under random contingency,”IEEE Transactions on Power Systems, vol. 35, no. 5, pp. 3332–3341, 2020

    work page 2020

  14. [14]

    Distribution system operation amidst wildfire- prone climate conditions under decision-dependent line availability uncertainty - dataset,

    A. Moreira, F. Pianc ´o, B. F. dos Santos, A. Street, R. Jiang, C. Zhao, and M. Heleno, “Distribution system operation amidst wildfire- prone climate conditions under decision-dependent line availability uncertainty - dataset,” 2023. [Online]. Available: https://dx.doi.org/10. 21227/318q-5k50

    work page 2023