Pith. sign in

REVIEW 2 major objections 6 minor 34 references

Instantaneous Failure, Repair and Mobility Rates for Markov Reliability Systems: A Wind-Farm application

T0 review · 2 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper introduces two new instantaneous rates, ROCOR and ROI, that complement ROCOF for Markov systems, and proves an identity that accounts for every transition as failure, repair, or internal movement.

desk verdict The theory is correct but elementary, and the wind-farm application rests on an unjustified matrix-logarithm embedding that invalidates the empirical claims as written. read the letter →

arxiv 2506.17280 v1 pith:DDDBWRET submitted 2025-06-14 eess.SY cs.SYphysics.app-phphysics.data-an

Guglielmo D'Amico , Filippo Petroni This is my paper · ORCID
classification eess.SYcs.SYphysics.app-phphysics.data-an MSC 60J2760K1090B25
keywords MarkovprocessesreliabilityROCOFROCORrateofinoccurrencetotalmobilitywindfarmwind-speedstates
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 argues that the Rate of Occurrence of Failures (ROCOF), alone, gives an incomplete instantaneous picture of a Markov reliability system. It introduces two companion rates, the Rate of Occurrence of Repairs (ROCOR), which counts transitions from failure states back to working states, and the Rate of Inoccurrence (ROI), which counts transitions that stay inside the current working or failure subset, and it proves explicit formulas for both in terms of the generator matrix. It then packs the three rates into a Total Mobility Rate (TMR) and proves the identity $\operatorname{roi}(t)+\operatorname{rof}(t)+\operatorname{ror}(t)=\operatorname{tmr}(t)$, so every infinitesimal transition is accounted as failure, repair, or internal movement. On hourly wind-speed data from 18 wind farms, the rates separate sites that have similar Weibull fits into persistence-driven and transition-driven reliability profiles, which would matter for scheduling maintenance and sizing fast reserves. If the claim is right, reliability analysis gains a direction-sensitive, time-resolved toolkit that static summaries cannot provide.

What carries the argument

The load-bearing object is the generator matrix $Q=(q_{h,j})$ of a continuous-time Markov chain, whose off-diagonal entries are instantaneous transition rates and whose row sums are zero. Partitioning $Q$ according to the working/failure split gives the block structure that separates failure transitions ($W\to F$), repair transitions ($F\to W$), and internal transitions (within $W$ or within $F$); each new rate is then an inner product of the unconditional occupancy probabilities $p_h(t)=\sum_i\alpha_i(e^{tQ})_{i,h}$ with the corresponding exit-intensity vector. The connective mechanism is the counting identity $N(t)=N_f(t)+N_r(t)+N_i(t)$, which turns the bookkeeping of every transition into three directionally distinct rates and yields the total mobility identity. In the empirical part, the same generator is recovered by embedding the estimated hourly transition matrix into continuous time through a matrix logarithm.

What would settle it

Re-estimate the rates without the embedding step by treating the hourly transition counts directly as discrete-time estimators of mobility, or by computing $Q$ from an explicit embedding algorithm and checking whether $e^{Q}$ reproduces the estimated transition matrix within sampling error. If the Gansu-versus-Muppandal difference in long-run total mobility ($\sim0.30$ versus $\sim0.11\,\mathrm{h}^{-1}$) collapses or reverses under this re-estimation, the claim that the indicators distinguish the two sites would be put in doubt.

Watch

Extended reading notes

Core claim

For a finite-state continuous-time Markov chain with state space $E$ partitioned into working states $W$ and failure states $F$, the paper defines ROCOR as the instantaneous expected rate of transitions $F\to W$, and ROI as the instantaneous expected rate of transitions within $W$ or within $F$. It derives the closed forms $\operatorname{ror}(t)=\sum_{i\in E}\sum_{f\in F}\sum_{w\in W}\alpha_i(e^{tQ})_{i,f}q_{f,w}$ and an analogous sum for ROI, using only the generator $Q$ and the initial distribution $\alpha$. Proposition 7 states the decomposition $\operatorname{roi}(t)+\operatorname{rof}(t)+\operatorname{ror}(t)=\operatorname{tmr}(t)=\sum_{i}\sum_{h}\alpha_i(e^{tQ})_{i,h}\sum_{j\neq h}q_{h,j}$, and the paper notes TMR is independent of the partition into working and failed states, recovering the long-run mobility index of the literature as its limit. Applied to hourly wind-speed records discretized into eleven states, with working states chosen as the operational wind classes, the indicators are computed for eighteen wind farms and claimed to expose 'reliability logics' that static Weibull parameters and ROCOF alone do not reveal.

Load-bearing premise

The wind-farm analysis assumes that the one-hour transition matrix estimated from the data has a matrix logarithm giving a valid continuous-time generator $Q$ with nonnegative off-diagonal rates and zero row sums; the paper states no algorithm, existence condition, or validity check for this embedding, and if the computed $Q$ is not a valid generator then every continuous-time indicator in the empirical study is undefined.

Editorial extensions

If this is right

  • Operators can classify a wind farm by its reliability logic: a flat high TMR plateau with strong ROI signals persistence and allows longer maintenance windows, while short-lived ROCOR and ROCOF bursts signal transition-driven volatility that needs fast reserves.
  • The identity $\operatorname{roi}+\operatorname{rof}+\operatorname{ror}=\operatorname{tmr}$ gives a monitoring decomposition: any change in total mobility can be attributed to more failures, more repairs, or more internal churn, so each movement direction is separately tracked over time.
  • Because TMR does not depend on the working/failure partition, the same function can be used as a generic time-dependent mobility measure for any finite-state continuous-time Markov process.
  • Sites with nearly equal Weibull scale and shape parameters can differ by a factor of roughly three in long-run total mobility (about $0.30\,\mathrm{h}^{-1}$ versus $0.11\,\mathrm{h}^{-1}$ in the contrasted pair), so static wind fits are not a reliable proxy for operational dynamism.

Reading between the lines

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

  • Inference: the same counting decomposition should carry over to semi-Markov reliability models, replacing the exponential holding times and $e^{tQ}$ occupancy with the transition probabilities of a semi-Markov kernel, although the paper proves the formulas only for the Markov case.
  • Inference: because the empirical 'failures' are defined by wind-speed thresholds rather than actual component failures, the rates are a proxy for operational regime dynamics; comparing them with turbine downtime logs would test whether the inferred reliability logics match real outage behavior.
  • Inference: the strong dependence on the initial state shown in the transient curves suggests a practical forecasting use not developed in the paper: set the initial distribution to the current wind state of a site and read the resulting ROCOR and ROCOF peaks as short-term transition risk warnings.
  • Inference: the same partition-independent TMR could be applied outside reliability, for instance to health-state, credit-rating, or labor-market Markov chains, wherever one wants a time-varying summary of how often agents change category.
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

2 major / 6 minor

Summary. The paper defines two new instantaneous reliability indicators for finite-state continuous-time Markov systems, the Rate of Occurrence of Repairs (ROCOR) and the Rate of Inoccurrence (ROI), and combines them with the classical ROCOF into a Total Mobility Rate (TMR). The theoretical part derives closed-form expressions for these rates, proves the identity roi(t)+rof(t)+ror(t)=tmr(t), and applies the framework to hourly wind-speed data from 18 wind farms, claiming that the indicators distinguish 'persistence-driven' from 'transition-driven' reliability behaviors even when Weibull parameters are similar. The mathematical derivations in Sections 2 and 3 are internally consistent, but the empirical application rests on an under-specified matrix-logarithm embedding of an estimated discrete-time transition matrix into a continuous-time generator.

Significance. If the empirical embedding is made rigorous, the paper offers a genuinely useful complement to ROCOF: the decomposition into failure, repair, and within-subset transition rates gives a time-dependent, direction-sensitive picture of Markov reliability systems, and the TMR identity (25) is a clean, correct summary. The proof of Theorem 6 is carefully developed, Proposition 7 is correct, and Remark 3 properly connects TMR to the known Geweke-Marshall-Zarkin mobility index. The wind-farm case study addresses a real operational problem and could be valuable for comparing sites, but the current empirical claims are conditional on an unvalidated numerical step and on the absence of any uncertainty quantification.

major comments (2)
  1. [Section 3.2] The embedding of the estimated hourly transition matrix P into a continuous-time generator Q via a matrix logarithm is the load-bearing step for every empirical result in the paper, but Section 3.2 gives no algorithm, no existence conditions, and no verification that the computed Q satisfies q_ij >= 0 for i != j and zero row sums. Without a valid generator, equations (6), (8), (10), (24), and (25), together with Figures 5–10, are not rates of any continuous-time Markov process, and the persistence/transition classification in Sections 3.5 and 3.6 is unsupported. The concern is concrete: Figure 3 shows that hourly jumps of more than two wind-speed classes are 'virtually absent', so if the estimated P contains exact zero entries, P = e^Q cannot hold for any irreducible Q because e^{tQ} has strictly positive entries for every t > 0. The approximation 'P ≈ e^{QΔt}' therefore needs an explicit error criterion, and the non-embeddable case must be handled explicitly, for example by a projection method, by direct estimation of Q from holding times, or by reporting discrete-time analogues of the indicators.
  2. [Sections 3.5–3.6] The headline empirical claim that the indicators 'distinguish between sites with similar long-term wind profiles' is presented through point estimates only. In particular, Section 3.6 contrasts Gansu (TMR_infinity about 0.30 h^-1) with Muppandal (about 0.11 h^-1) as evidence of fundamentally different reliability profiles, but the paper provides no confidence bands, bootstrap intervals, or sensitivity analysis for any of the estimated rate curves. Because Q is itself estimated from a finite sample, the displayed differences could in principle be driven by estimation error, so the classification needs at least a bootstrap or simulation-based assessment before the applied conclusion can be accepted.
minor comments (6)
  1. [Section 2.1, Remark 1] The asymptotic formulas contain notation errors: 'rof(∞)=⟨α L, Q_W⟩' should be '⟨L_W, Q_W⟩', and 'L=[L_W, L_W]' should be '[L_W, L_F]'; the formulas for ror(∞) and roi(∞) should use L_F and L_W consistently.
  2. [Section 3.3 and Figure 5] The text says the process starts in the calmest wind class 0–2 m/s, which is state 0, while Figure 5's caption says 'state 1'; state 1 corresponds to 2–4 m/s. Please reconcile this inconsistency.
  3. [Section 3.2] The statement that the final incomplete segment of each day is 'ignored' appears to discard the legitimate one-hour transition from hour 23 to hour 0 of the next day. If the goal is to keep Δt = 1 h, those transitions should be included, or the exclusion should be justified.
  4. [Section 3.6] The claim that Gansu and Muppandal 'both have shape factors near k≈2.5' is inaccurate: Table 1 lists k = 2.09 for Gansu and k = 2.85 for Muppandal. Either revise the wording or choose sites with closer shape parameters.
  5. [Section 2, Definitions 3 and 5] The term 'rate of inoccurrence' may mislead readers, since ROI counts transitions within the working and failure subsets rather than the absence of transitions; consider a more descriptive term or an explicit remark clarifying that ROI measures mobility inside the two subsets.
  6. [General] The paper does not state how the matrix logarithm was computed or whether code and data are available; a reproducibility statement and a description of the numerical method (for example, a Schur–Parlett algorithm) would be needed for the empirical figures to be auditable.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the new rate measures are explicit functions of a given Markov generator Q, TMR is defined as their sum, and the empirical indicators are descriptive transforms of the fitted Markov chain rather than predictions equivalent to their inputs.

full rationale

The mathematical core of the paper is self-contained conditional on a continuous-time Markov generator Q and an initial distribution α. Definitions 1, 3, and 5 introduce ROCOF, ROCOR, and ROI as time-derivatives of expected counting processes; Theorems 2, 4, and 6 evaluate these as explicit algebraic functions of Q and (e^{tQ}); Proposition 7 is an algebraic identity among these already-derived expressions, and TMR is then defined to equal that sum. The only imported mathematical result, Ding-hua's ROCOF formula in Theorem 2, is external and used as a benchmark, and the proof of Theorem 6 invokes Lemma 1 of Yeh (1997), also external. The authors' self-citations, mostly about sequential interval reliability and higher-order ROCOF, are contextual literature review and are not load-bearing premises of the new derivations. The empirical section estimates a discrete-time transition matrix P and then chooses a generator Q satisfying P ≈ e^{QΔt}; the mobility indicators are then computed from Q, so they are descriptive summaries of the same fitted data rather than predictions of independent outcomes. Thus there is no fitted-parameter-renamed-as-prediction step. A genuine missing-support concern does exist—Section 3.2 asserts the matrix-logarithm embedding with no algorithm, existence conditions, or treatment of non-embeddable P, and the near-zero off-diagonal entries of the empirical P are in tension with exact embeddability—but that is a correctness and validity issue, not a circularity of the derivation, and it does not raise the circularity score.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The theoretical formulas require only standard Markov machinery. The empirical claims additionally assume a valid continuous-time embedding and a hand-chosen W/F partition. No new physical entities are posited.

free parameters (2)
  • Wind-speed bin width (state discretization) = 2 m/s
    Each of the 11 states spans 2 m/s from 0 to at least 20 m/s. The choice affects the empirical transition matrix P, the generator Q, and every indicator value; no sensitivity analysis is given.
  • Working-state set W = W = {2,...,8} (states corresponding to 4-18 m/s)
    Chosen by hand as a proxy for the turbine operating range. This determines the decomposition into ROCOR and ROI, though TMR is invariant to the partition.
assumptions (4)
  • domain assumption The wind-speed process is a first-order time-homogeneous Markov chain on 11 states.
    Stated in Section 3.2; not tested for higher-order dependence, seasonality, or regime changes over the 2016-2025 window.
  • domain assumption The estimated hourly transition matrix P is embeddable into a continuous-time generator Q via a matrix logarithm such that P is approximately e^{Q*1h}.
    Invoked in Section 3.2 to compute time-continuous indicators; no conditions for the existence or validity of Q are provided.
  • domain assumption Wind-speed bins 0, 1, 9, 10 are failure states and bins 2 through 8 are working states.
    Section 3.3 classifies states based on typical turbine power curves; this partition is an assumption, not derived from turbine failure data.
  • standard math Standard Markov chain calculus (Kolmogorov equations, matrix exponential) and Ding-hua's ROCOF formula are correct.
    Used throughout Section 2; standard background for the definitions and proofs.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Instantaneous Failure, Repair and Mobility Rates for Markov Reliability Systems: A Wind-Farm application." pith.science (2026). https://pith.science/paper/DDDBWRET

@misc{pith2026250617280,
  author       = {Pith},
  title        = {Pith review of: Instantaneous Failure, Repair and Mobility Rates for Markov Reliability Systems: A Wind-Farm application},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/DDDBWRET}},
  note         = {Machine review of arXiv:2506.17280}
}
read the original abstract

The Rate of Occurrence of Failures (ROCOF) is a widely utilized indicator for assessing a system's performance over time, yet it does not fully disclose the instantaneous behavior of a system. This paper introduces new measures to complement the ROCOF, providing a more comprehensive understanding of system reliability, particularly for Markov systems. We define the Rate of Occurrence of Repairs (ROCOR), which quantifies the system's instantaneous tendency to transition from failure to working states, and the Rate of Inoccurrence (ROI), which measures the propensity to remain within the current subset of states (either working or failure) without transitioning out. Explicit expressions for the computation of these rates are derived for Markov systems. Furthermore, a Total Mobility Rate (TMR) is proposed, integrating these individual rates to capture the overall dynamism of the system. The utility of these new indicators is demonstrated through a significant real-world application to wind farm management. The results from the wind farm study show that ROCOR, ROI, and TMR, when used in conjunction with ROCOF, reveal nuanced operational dynamics and reliability characteristics that are not discernible from static measures like Weibull parameters or ROCOF alone. These indicators can distinguish between sites with similar long-term wind profiles by identifying different "reliability logics," such as persistence-driven versus transition-driven behaviors. This enriched, time-dependent perspective provides valuable information for maintenance scheduling, operational strategies, and risk assessment, ultimately enhancing the ability to manage complex systems effectively.

Figures

Figures reproduced from arXiv: 2506.17280 by the authors.

Figure 1
Figure 1. Geographical distribution of the 18 wind farms analysed. • Spatial coverage – 18 on-shore and off-shore facilities located in 10 countries on five continents (United States, Chile, China, Australia, United Kingdom, India, Kenya, Italy, Mongolia, Belgium). • Temporal coverage – 1 January 2016 – 21 April 2025 (81 576 hourly records per site, 𝑁 = 1 468 368 observations in total). • Variable – 𝑊 𝑆50𝑀 (m∕s); original MER… view at source ↗
Figure 2
Figure 2. Hourly wind-speed histograms (bars) and fitted Weibull densities (lines) for each site. The resulting matrix is displayed in [PITH_FULL_IMAGE:figures/full_fig_p011_2.png] view at source ↗
Figure 3
Figure 3. Hourly one–step transition–probability matrix 𝑃 = ( 𝑝𝑖𝑗) obtained by classifying each observation into the 11 wind-speed states (2 m s−1 bins, 0–≥20 m s−1). The strong main diagonal and its two nearest neighbours show that hour-to-hour changes seldom exceed ±2 m s−1, while the nearly symmetric pattern indicates the absence of any persistent upward or downward drift over the 2016–2025 window. Page 16 of 15 [PITH_FUL… view at source ↗
Figures from the paper (7 more)
Figure 4
Figure 4. Figure 4: Continuous-time generator matrix 𝐐 = ( 𝐪𝐢𝐣) (units: h−1) derived from the embedded chain in [PITH_FULL_IMAGE:figures/full_fig_p017_4.png]
Figure 5
Figure 5. Figure 5: Temporal evolution of the four mobility indicators starting from a calm (failure) state (state 1). Sites differ markedly in their early repair dominance, timing of crossover, and long-run volatility, revealing distinct reliability dynamics not visible from static analy…
Figure 6
Figure 6. Figure 6: Mobility indicators starting from a lower working state (state 2, 4–6 m/s). The early dynamics are dominated by failures, with delayed and site-dependent recoveries. Long-run mobility reflects the system’s ability to stabilize within productive classes. Page 18 of 15 …
Figure 7
Figure 7. Figure 7: Mobility indicators starting near the center of the working range (state 6, 12–14 m/s). The system exhibits balanced, low-volatility dynamics with persistent ROI and mild transitions, highlighting intra-class motion rather than threshold crossings [PITH_FULL_IMAGE:fig…
Figure 8
Figure 8. Figure 8: Mobility indicators starting in an upper failure state (state 10, >20 m/s). The early evolution is dominated by attempts to re-enter the working subset, with ROR spikes followed by renewed failures. Long-run mobility remains low and asymmetric across sites. Page 19 of …
Figure 9
Figure 9. Figure 9: 72-hour mobility indicators at Gansu Wind Farm (persistence-driven regime) [PITH_FULL_IMAGE:figures/full_fig_p020_9.png]
Figure 10
Figure 10. Figure 10: 72-hour mobility indicators at Muppandal Wind Farm (transition-driven regime) Page 20 of 15 [PITH_FULL_IMAGE:figures/full_fig_p020_10.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 33 canonical work pages

  1. [1]

    author W. R. Blischke , author D. P. Murthy , title Reliability: modeling, prediction, and optimization , publisher John Wiley & Sons , year 2011

    work page 2011

  2. [2]

    author V. S. Barbu , author N. Limnios , title Semi-Markov chains and hidden semi-Markov models toward applications: their use in reliability and DNA analysis , volume volume 191 , publisher Springer Science & Business Media , year 2009

    work page 2009

  3. [3]

    author I. A. Ushakov , title Probabilistic reliability models , publisher John Wiley & Sons , year 2012

    work page 2012

  4. [4]

    author M. V. Koutras , title On a markov chain approach for the study of reliability structures , journal Journal of Applied Probability volume 33 ( year 1996 ) pages 357--367

    work page 1996

  5. [5]

    Limnios , author G

    author N. Limnios , author G. Oprisan , title Semi-Markov processes and reliability , publisher Springer Science & Business Media , year 2001

    work page 2001

  6. [6]

    Csenki , title On the interval reliability of systems modelled by finite semi-markov processes , journal Microelectronics Reliability volume 34 ( year 1994 ) pages 1319--1335

    author A. Csenki , title On the interval reliability of systems modelled by finite semi-markov processes , journal Microelectronics Reliability volume 34 ( year 1994 ) pages 1319--1335

    work page 1994

  7. [7]

    author A. Csenki , title An integral equation approach to the interval reliability of systems modelled by finite semi-markov processes , journal Reliability engineering & system safety volume 47 ( year 1995 ) pages 37--45

    work page 1995

  8. [8]

    Georgiadis , author N

    author S. Georgiadis , author N. Limnios , title Interval reliability for semi-markov systems in discrete time , journal Journal de la Soci \'e t \'e Fran c aise de Statistique volume 155 ( year 2014 ) pages 152--166

    work page 2014

Show all 34 references
  1. [9]

    Georgiadis , author N

    author S. Georgiadis , author N. Limnios , title Nonparametric estimation of interval reliability for discrete-time semi-markov systems , journal Journal of Statistical Theory and Practice volume 10 ( year 2016 ) pages 20--39

    2016

  2. [10]

    author V. S. Barbu , author G. D’Amico , author T. Gkelsinis , title Sequential interval reliability for discrete-time homogeneous semi-markov repairable systems , journal Mathematics volume 9 ( year 2021 ) pages 1997

    2021

  3. [11]

    D’Amico , author T

    author G. D’Amico , author T. Gkelsinis , title On a mixed transient--asymptotic result for the sequential interval reliability for semi-markov chains , journal Mathematics volume 12 ( year 2024 ) pages 1842

    2024

  4. [12]

    Platis , author N

    author A. Platis , author N. Limnios , author M. Le Du , title Dependability analysis of systems modeled by non-homogeneous markov chains , journal Reliability Engineering & System Safety volume 61 ( year 1998 ) pages 235--249

    1998

  5. [13]

    Limnios , title Dependability analysis of semi-markov systems , journal Reliability Engineering & System Safety volume 55 ( year 1997 ) pages 203--207

    author N. Limnios , title Dependability analysis of semi-markov systems , journal Reliability Engineering & System Safety volume 55 ( year 1997 ) pages 203--207

    1997

  6. [14]

    D’Amico , author J

    author G. D’Amico , author J. Janssen , author R. Manca , title Initial and final backward and forward discrete time non-homogeneous semi-markov credit risk models , journal Methodology and Computing in Applied Probability volume 12 ( year 2010 ) pages 215--225

    2010

  7. [15]

    D’Amico , author R

    author G. D’Amico , author R. Manca , author F. Petroni , author D. Selvamuthu , title On the computation of some interval reliability indicators for semi-markov systems , journal Mathematics volume 9 ( year 2021 ) pages 575

    2021

  8. [16]

    Yang , author J

    author K. Yang , author J. Xue , title Continuous state reliability analysis , in: booktitle Proceedings of 1996 Annual Reliability and Maintainability Symposium , organization IEEE , year 1996 , pp. pages 251--257

    1996

  9. [17]

    D’Amico , title Age-usage semi-markov models , journal Applied Mathematical Modelling volume 35 ( year 2011 ) pages 4354--4366

    author G. D’Amico , title Age-usage semi-markov models , journal Applied Mathematical Modelling volume 35 ( year 2011 ) pages 4354--4366

    2011

  10. [18]

    author N. Limnios , title Reliability measures of semi-markov systems with general state space , journal Methodology and Computing in Applied Probability volume 14 ( year 2012 ) pages 895--917

    2012

  11. [19]

    author L. Yeh , title Calculating the rate of occurrence of failures for continuous-time markov chains with application to a two-component parallel system , journal Journal of the Operational Research Society volume 46 ( year 1995 ) pages 528--536

    1995

  12. [20]

    Ouhbi , author N

    author B. Ouhbi , author N. Limnios , title The rate of occurrence of failures for semi-markov processes and estimation , journal Statistics & probability letters volume 59 ( year 2002 ) pages 245--255

    2002

  13. [21]

    author G. D’Amico , title Rate of occurrence of failures (rocof) of higher-order for markov processes: analysis, inference and application to financial credit ratings , journal Methodology and Computing in Applied Probability volume 17 ( year 2015 ) pages 929--949

    2015

  14. [22]

    D’Amico , author F

    author G. D’Amico , author F. Petroni , title Rocof of higher order for semi-markov processes , journal Applied Mathematics and Computation volume 441 ( year 2023 ) pages 127719

    2023

  15. [23]

    D’Amico , author F

    author G. D’Amico , author F. Gismondi , title Recent developments in the computation of the rocof of multi-state systems and its applications , journal Reliability theory & applications volume 1 (77) (Vol.19) ( year 2024 )

    2024

  16. [24]

    Bruneau , author S

    author M. Bruneau , author S. E. Chang , author R. T. Eguchi , author G. C. Lee , author T. D. O'Rourke , author A. M. Reinhorn , author M. Shinozuka , author K. Tierney , author W. A. Wallace , author D. Von Winterfeldt , title A framework to quantitatively assess and enhance...

    2003

  17. [25]

    Sharma , author A

    author N. Sharma , author A. Tabandeh , author P. Gardoni , title Resilience analysis: A mathematical formulation to model resilience of engineering systems , journal Sustainable and Resilient Infrastructure volume 3 ( year 2018 ) pages 49--67

    2018

  18. [26]

    Billinton , author K

    author R. Billinton , author K. E. Bollinger , title Transmission system reliability evaluation using markov processes , journal IEEE Transactions on power apparatus and systems ( year 1968 ) pages 538--547

    1968

  19. [27]

    Banjevic , author A

    author D. Banjevic , author A. Jardine , title Calculation of reliability function and remaining useful life for a markov failure time process , journal IMA journal of management mathematics volume 17 ( year 2006 ) pages 115--130

    2006

  20. [28]

    author A. Csenki , title Joint interval reliability for markov systems with an application in transmission line reliability , journal Reliability Engineering & System Safety volume 92 ( year 2007 ) pages 685--696

    2007

  21. [29]

    Br \'e maud , title Gibbs fields and monte carlo simulation , journal Markov Chains: Gibbs Fields, Monte Carlo Simulation, and Queues ( year 1999 ) pages 253--322

    author P. Br \'e maud , title Gibbs fields and monte carlo simulation , journal Markov Chains: Gibbs Fields, Monte Carlo Simulation, and Queues ( year 1999 ) pages 253--322

    1999

  22. [30]

    Sadek , author N

    author A. Sadek , author N. Limnios , title Nonparametric estimation of reliability and survival function for continuous-time finite markov processes , journal Journal of Statistical Planning and Inference volume 133 ( year 2005 ) pages 1--21

    2005

  23. [31]

    author S. Ding-hua , title A new method for calculating the mean failure numbers of a repairable system during (0, t , journal Acta Mathematicae Applicatae Sinica volume 8 ( year 1985 ) pages 101--110

    1985

  24. [32]

    Yeh , title The rate of occurrence of failures , journal Journal of Applied Probability volume 34 ( year 1997 ) pages 234--247

    author L. Yeh , title The rate of occurrence of failures , journal Journal of Applied Probability volume 34 ( year 1997 ) pages 234--247

    1997

  25. [33]

    Geweke , author R

    author J. Geweke , author R. C. Marshall , author G. A. Zarkin , title Mobility indices in continuous time markov chains , journal Econometrica: Journal of the Econometric Society ( year 1986 ) pages 1407--1423

    1986

  26. [34]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.