ArticlePDF Available

A Monte Carlo approach to estimation of G-renewal process in warranty data analysis

Authors:
  • The Ford Motor Company

Figures

Content may be subject to copyright.
.
1
A Monte Carlo Approach to Estimation of
G-Renewal Process in Warranty Data Analysis
Mark P. Kaminskiy and Vasiliy V. Krivtsov
1. Introduction
For many years, the most commonly used models for the failure process have been the renewal process
(RP) and the nonhomogeneous Poisson process (NHPP). In the framework of the repairable system applications,
RP is used to model the situations with restoration to "good-as-new" condition (perfect repair assumption),
meanwhile NHPP is applied to the situations with the “same-as-old” restoration (minimal repair assumption). In a
sense, these two assumptions can be considered as extreme ones from both theoretical and practical standpoints.
In order to avoid this “extremism”, several generalizing models have been introduced in recent years. References
include Brown & Proschan (1982), Kijima & Sumita (1986), Filkenstein, (1993), Lindqvist (1999). Among these
models, the G-Renewal Process (GRP) introduced by Kijima & Sumita (1986) is very attractive, since it covers
the intermediate "better-than-old-but-worse-than-new" repair assumption and results in a G-renewal equation,
which is a generalization of the well-known ordinary renewal equation. Unfortunately, a closed form solution of
the equation is unavailable, which makes the respective statistical estimation challenging.
The objective of this paper is limited to statistical estimation of the parameters of G-Renewal Process,
based on typical one-dimensional warranty (grouped failure time) data. The Monte Carlo approach considered
below can be also applied to parameter estimation of Kijima’s Model I and II (Kijima, 1989).
2. G-Renewal Process
Kijima and Sumita (1986) introduced a G-Renewal Process, which can model restoration conditions
ranging from "good-as-new" to "same-as-old". The GRP is introduced using the notion of virtual age.
Let A
n
be the virtual age of a system immediately after the nth repair. If A
n
= y, then the system has the
time to the (n + 1)th failure X
n+1
, which is distributed according to the following cumulative distribution function
(CDF):
)(1
)()(
) ( yF
yFyXF
yAXF
n
+
==
, (1)
where F(X) is the CDF of the time-to-first-failure (TTFF) distribution of a new system. The sum
=
=
n
i
in
XS
1
,
with S
0
= 0, is called the real age of the system.
In the framework of the GRP it is assumed that the nth repair can remove the damage incurred only
during the time between the (n –1)th and the nth failures, so that the respective virtual age after the nth repair is
A
n
= A
n – 1
+ q X
n
= q S
n
, n =1, 2, . . .
where q is the parameter of rejuvenation (or repair effectiveness parameter) and the virtual age of a new system
A
0
= 0, so that the TTFF is distributed according to F(t|0)
F(t).
The time between the first and second failures is distributed according to (1) with A
1
= qX
1.
Respectively, the time between the second and third failure is distributed according to (1) with A
2
= q ( X
1
+ X
2
),
and so on.
.
2
It is clear that for q = 0, the considered process coincides with an ordinary RP, thus, modeling the
"good-as-new" repair assumption. With q = 1, a system is restored to the "same-as-old" condition, which is
similar to NHPP. The case of 0 < q < 1 corresponds to the intermediate "better-than-old-but-worse-than-new"
repair assumption. Finally, with q > 1, the virtual age A
n
> S
n
, i.e., the repair damages the system to a higher
degree than it was just before the respective failure, which corresponds to the "worse-than-old" repair
assumption.
The expected number of failures in (0, t], which is called a cumulative intensity function, is given by a
solution of G-renewal equation (Kijima, et al., 1988):
,)|()()0|()(
0 0
τττ
τ
ddxxxgxhgtH
t
∫ ∫
+=
(2)
where
,0,,
)(1
)(
) (
+
= xt
qxF
qxtf
xtg
is such that h(t) = d(H(t))/dt, g(t|0) = f(t), and F(t) and f(t) are the CDF and PDF of the TTFF distribution.
The closed form solution of (2) is not available, and even numerical solutions are difficult to obtain,
since the equation contains a recurrent infinite system (Finkelstein, 1997). A Monte Carlo based solution is,
however, possible and was discussed by Kaminskiy and Krivtsov (1998).
3. Warranty Data
Typical one-dimensional warranty data are collected as results of observations on a large population of
identical repairable units. The population size, N
0
, is known and it can be assumed as constant in time (the
number of lost units, if any, is negligible with respect to N
0
). The following table provides an example of real
warranty data, which are analyzed in the next section.
Table 1. Example of Warranty Data for a Repairable System. Population Size, N
0
= 100000.
Month in Service, t 3 6 9 12 15 18 21 24 27
Cumulative Number of
Failures Per System, H
emp
(t)
0.03
0.09
0.14
0.24
0.38
0.54
0.70
0.90
1.17
4. Estimation Procedure
Based on the warranty data, the empirical cumulative intensity function, H
emp
(t), is calculated as
0
)(
)( N
tN
tH
i
ie mp
=
, t
i
< t
i+1
, i = 1, 2, . . . , n (3)
where N(t) is the cumulative number of failures in (0, t].
Denote a solution of G-renewal equation (2) obtained by Monte Carlo simulations by
Η
mc
(t) = f(F(
τ|α
), q, t)), (4)
where F(
τ|α
) is a given time-to-first-failure CDF with unknown vector of parameters,
α
.
Using (3) and (4), the least squares estimates of GRP parameters
α
and q can be obtained as a solution
of GRP parameters
α
and q can be obtained as a solution of the following optimization problem:
ΗΗ
=
n
i
imciemp
q
tqFt
1
2
,
)),),|(()((min
ατ
α
.
3
5. Examples
5.1
Simulated Data
The empirical cumulative intensity function, H
emp
(t), was obtained by simulating a GRP with a Weibull
distributed TTFF (shape parameter,
β
= 1.5 and scale parameter,
Θ
= 1) and the GRP rejuvenation parameter,
q = 0.5 using N
0
= 100 realizations over the observation interval, T = 5
Θ
.
Estimates of
β, Θ
, and q were obtained based on n
0
= 1000 realizations of GRP as follows:
0.49
^
,00.1
^
1.48,
^
=== q
θβ
. Tables 2 - 3 show the sample correlation and covariance matrices for the
obtained estimates of GRP parameters for 30 simulated empirical cumulative intensity functions, H
emp
(t).
Table 2. Sample Correlation Matrix
β
ββ
β
Θ
ΘΘ
Θ
q
β
ββ
β
1.000
Θ
ΘΘ
Θ
0.702 1.000
q 0.079 -0.523 1.000
Table 3. Sample Covariance Matrix
β
ββ
β
Θ
ΘΘ
Θ
q
β
ββ
β
2.6 10
-3
Θ
ΘΘ
Θ
1.9 10
-3
2.8 10
-3
q 2.3 10
-4
-1.6 10
-3
3.2 10
-3
With the empirical cumulative intensity function simulated for N
0
= 100000 realizations (which is a
more typical case for a company concerned with mass production) and using n
0
= 1000000, the estimation
procedure returns the original GRP parameters with close to zero variance.
5.2
Real Data
The warranty data collected on a system during first 18 months (see Table 1) were used for estimation of
GRP parameters. The Weibull distribution with the shape parameter,
β
, and the scale parameter,
Θ
, was assumed
as the underlying TTFF distribution. The solid line in Figure 1 represents the least squares fit from a family of G-
renewal functions simulated in the following parameter domain: {1 <
β
< 2, 10 <
Θ
< 50, 0 < q < 1}.
Figure 1. A Lest Square Fit of G-Renewal Function
0.0
0.2
0.4
0.6
0.8
1.0
1.2
0 5 10 15 20 25 30 35
time in months
number of failures per system
Actu al data
GRP, q = 0.7
GRP, q = 1
ORP
.
4
The obtained estimates of GRP parameters are
0.70
^
,42
^
1.8,
^
=== q
θβ
. The estimated G-renewal
function shows a good fit to the data not only in the interval (0, 18] months (used for estimation) but also in the
remaining interval (18, 30] months (obtained by prediction), see Figure 1. The figure also shows the extreme
repair conditions modeled by the RP (q = 0) and the GRP (q = 1).
It is reasonable to conclude that the approach considered above is not only practically applicable for
estimation of the GRP parameters, but also for prediction of the G-renewal function, which is often essential in
warranty data analysis.
References
1. M. Brown, and F. Proshan, "Imperfect Maintenance" - in Crowley J. and Johnson R. (ed) Survival Analysis.
Vol. 2, 1982, pp 179-188
.
2. M. S. Filkenstein, "The Concealed Age of Distribution Function and the Problem of General Repair",
Journal of Statistical Planning and Inference, # 65, 1997, pp. 315-321.
3. M. P. Kaminskiy and V.V. Krivtsov, "A Monte Carlo Approach to Repairable System Reliability Analysis" -
in Probabilistic Safety Assessment and Management, Springer-Verlag London Ltd, 1998, pp. 1063-1068.
4. M. Kijima and N. Sumita, "A Useful Generalization of Renewal Theory: Counting Process Governed by
Non-negative Markovian Increments", Journal of Applied Probability, # 23, 1986, pp. 71-88.
5. M. Kijima, "Some Results for Repairable Systems with General Repair", Journal of Applied Probability, #
26, 1989, pp. 89-102.
6. H. Lindqvist, "Statistical Modeling and Analysis of Repairable Systems" - in Statistical and Probabilistic
Models in Reliability, Birkhauser, Berlin, 1999, pp. 3-25.
... The authors developed MAPLE language code for the same. Kaminskiy and Krivtsov (2006) proposed Monte Carlo approach for parameter estimation of GRP in 2006. Pascuala and Ortega (2006) proposed a failure rate model as given in Eq. (10), based on an improvement factor P considering three actions i.e. repair, overhaul and replacement, wherein repair and overhaul are considered as minimal and imperfect maintenance, respectively. ...
Article
Most of the systems are treated as repairable in automobile, nuclear and aviation industries, and often subjected to imperfect maintenance. Reliability assessment of such systems is generally done by observing their failures which could be the consequences of bad environmental conditions, and thus, the mathematical modeling of failure data along with other risk factors becomes essential. The available literature provides mainly two types of models to deal with such situations: (1) traditional imperfect maintenance (TIM) models, and (2) proportional hazard (PH) models (PHM). These are two major categories of models for reliability modelling and analysis of lifetime data and are treated separately. To model and analyze practical industrial problems, at times it is unavoidable to use and combine TIM with PH models. Presently, the literature lacks in providing an organized literature survey on the development of both TIM and PH models together summarized at one place. Considering this as a research gap, we present an extensive literature survey, from 1965 to 2020, on the development of (1) TIM models, and (2) PH models. Through this work, our objective is to create a new paradigm for the academicians to appreciate overall development of above stated categorized models for future work. Most importantly, the yearwise organized framework of the presented paper will contribute as a ready reference for the researchers to appreciate the developments in both the fields.
... Considering the complexity of the GRP model and the MC simulation, the computation is quite involved and time consuming. Kaminisky and Krivtsov [48] apply MC simulation to real warranty data to estimate GRP parameters. Authors recognize the complexities and the difficulties of developing a mathematically tractable probabilistic model to the GRP and discuss an alternative maximum likelihood estimation (MLE) approach to solve the GRP model without offering any solution. ...
Article
Full-text available
Many models and methodologies are available to deal with imperfect repair for repairable systems. Initial attempts at modeling imperfect repair using the (p, q) rule that defined the two extremities of imperfect repair-perfect renewal and minimal repair were effectively extended by Kijima and Sumita. They developed the generalized renewal theory from the renewal theory in the context of imperfect repair and applied it to repairable systems with the concept of virtual age. Since this pioneering work, much of imperfect repair modeling literature builds up on Kijima's models based on the Generalized Renewal Process (GRP). The paper conducts a survey for imperfect repair of repairable systems using GRP based on arithmetic reduction of age (ARA) and arithmetic reduction of intensity (ARI) concepts in general and Kijima models in particular. In addition to the theoretical development of Kijima models and its extensions, the review highlights their applications such as designing maintenance policies based on the concept of ARA. The survey will be useful to researchers in providing a starting point to develop appropriate models that suit their application oriented requirements.
... Kaminskiy and Krivtsov [7] extended the Kijima models by allowing q > 1, so that the repair damages (ages) the system to a higher degree than it was just before the respective failure. Tanwar et al. [8] provides a comprehensive review of Kijima-type models, wherein they generalize them as the arithmetic age reduction models. ...
Article
There is a variety of models available for repairable systems with general repairs. Most popular are the Kijima models (reflecting the generalized renewal process of recurrent failures) and the Lam model (reflecting the geometric process). The Kijima models relating system's real and virtual ages can be thought of as the time shift transformation, whereas The Lam model – as the time scale transformation. In this paper, a new model is proposed that combines these two fundamental transformations within a new probabilistic formulation. Besides probabilistic aspect of the model, the maximum likelihood estimation of model parameters is discussed, and its performance is illustrated through several numerical examples using both simulated and real–life data. The efficient Monte Carlo calculation method is suggested for the expected number of recurrent events, unavailability, and the failure intensity function.
... Only in the case of q = 1, the closed-from solution to the g-renewal equation can be obtained. Therefore, different numerical methods, such as the Monte Carlo approach (Kaminskiy and Krivtsov 2006), are employed. However, numerical methods are computationally demanding, on the one hand, and provide solutions only within limited domains, on the other. ...
Article
Full-text available
The Kijima’s type 1 maintenance model, representing the general renewal process, is one of the most important in the reliability theory. The g-renewal equation is central in Kijima’s theory and it is a Volterra integral equation of the second kind. Although these equations are well-studied, a closed-form solution to the g-renewal equation has not yet been obtained. Despite the fact that several semi-empirical techniques to approximate the g-renewal function have been previously developed, analytical approaches to solve this equation for a wide class of underlying distributions is still of current interest. In this paper, a long-time asymptotic for the g-renewal rate is obtained for distributions with nondecreasing hazard functions and for all values of the restoration factor \(q\in [0,1]\). The obtained analytical result is compared with the numerical solutions for two types of underlying distributions, showing a good asymptotic match. The obtained approximate g-renewal rate is employed for maintenance optimization, considering the repair cost as a function of the restoration factor. Several numerical examples are performed in order to show the efficiency of our results.
... One of the simulations is proposed by Blischke and Murthy [1]. The simulations based on Monte Carlo (MC) method are proposed by Brown et al. [22], and Kaminskiy and Krivtsov [23] who apply it on G-renewal process. ...
Article
Full-text available
In the analysis of warranty, renewal functions are important in acquiring the expected number of failures of a nonrepairable component in a time interval. It is very difficult and complicated -if at all possible- to obtain a renewal function analytically. This paper proposes a numerical integration method for estimating renewal functions in the terms of renewal integral equations. The estimation is done through the Mean Value Theorem for Integrals (MeVTI) method after modifying the variable of the renewal integral equations. The accuracy of the estimation is measured by its comparison against the existing analytical approach of renewal functions, those are for Exponential, Erlang, Gamma, and Normal baseline failure distributions. The estimation of the renewal function for a Weibull baseline failure distribution as the results of the method is compared to that of the well-known numerical integration approaches, the Riemann-Stieljies and cubic spline methods. Keywords : Mean Value Theorem for Integrals, Renewal Functions, Renewal Integral Equations.
... When dealing with the model I of Kijima (1989), the closed form solution of the equation is not available, and even numerical solutions are extremely difficult to ob- tain because of the mathematical complexity of the g-renewal equation. Kaminskiy and Krivtsov (2006) proposed an approximate solution using the Monte Carlo (MC) simulation technique. ...
Book
Replacement Models with Minimal Repair is a collection of works by several well-known specialists on the subject of minimal repair in replacement policies. It gives an exhaustive list of minimal repair models for the effective planning of minimal repair and maintenance actions. Written in an engaging style, Replacement Models with Minimal Repair balances complex mathematical models with practical applications. It is divided into six parts that cover: • mathematical modeling of minimal repair; • preventive maintenance models and optimal scheduling of imperfect preventive maintenance activities; • a new warranty servicing strategy with imperfect repair; • mathematical models combining burn-in procedure and general maintenance policies; • methods for parameters’ estimation of minimal repair models; and • product support. Replacement Models with Minimal Repair is for anyone with an interest in minimal repair and its impact on maintenance policies and strategies. It is a particularly useful resource for researchers, practitioners, and graduate students.
... Among the repairable systems, GRP is the attractive one for reliability analysis modelling, since it covers not only the RP and the NHPP, but also the intermediate "younger than old but older than new" repair assumption. GRP has been used in many applications, such as automobile industry [34] and oil industry [35]. ...
Article
Full-text available
Reliability is an important phase in durable system designs, specifically in the early phase of the product development. In this paper, a new methodology is proposed for complex systems’ design for reliability. Specific test and field failure data scarcity is evaluated here as a challenge to implement design for reliability of a new product. In the developed approach, modeling and simulation of the system are accomplished by using reliability block diagram (RBD) method. The generic data are corrected to account for the design and environment effects on the application. The integral methodology evaluates reliability of the system and assesses the importance of each component. In addition, the availability of the system was evaluated using Monte Carlo simulation. Available design alternatives with different components are analyzed for reliability optimization. Evaluating reliability of complex systems in competitive design attempts is one of the applications of this method. The advantage of this method is that it is applicable in early design phase where there is only limited failure data available. As a case study, horizontal drilling equipment is used for assessment of the proposed method. Benchmarking of the results with a system with more available failure and maintenance data verifies the effectiveness and performance quality of presented method.
... Considering the complexity of the GRP model and the MC simulation, the computation is quite involved and time consuming. Kaminisky and Krivtsov [48] apply MC simulation to real warranty data to estimate GRP parameters. Authors recognize the complexities and the difficulties of developing a mathematically tractable probabilistic model to the GRP and discuss an alternative maximum likelihood estimation (MLE) approach to solve the GRP model without offering any solution. ...
Article
Many models and methodologies are available to deal with imperfect repair for repairable systems. Initial attempts at modeling imperfect repair using the (p, q) rule that defined the two extremities of imperfect repair–perfect renewal and minimal repair were effectively extended by Kijima and Sumita. They developed the generalized renewal theory from the renewal theory in the context of imperfect repair and applied it to repairable systems with the concept of virtual age. Since this pioneering work, much of imperfect repair modeling literature builds up on Kijima's models based on the Generalized Renewal Process (GRP). The paper conducts a survey for imperfect repair of repairable systems using GRP based on arithmetic reduction of age (ARA) and arithmetic reduction of intensity (ARI) concepts in general and Kijima models in particular. In addition to the theoretical development of Kijima models and its extensions, the review highlights their applications such as designing maintenance policies based on the concept of ARA. The survey will be useful to researchers in providing a starting point to develop appropriate models that suit their application oriented requirements.
Chapter
An age-dependent repair model is considered in this paper. The notion of the “age” of the product and the degree of repair are used to define the virtual age of the product. Two problems are considered in this paper. In the first problem, the degree of repair is a stochastic process and is allowed to switch between a finite number of values due to various phenomena. Switching is assumed to happen according to the jumps of a homogeneous, finite-state Markov chain. We use hidden Markov models (HMM) to develop a recursion to estimate the conditional probability distribution of the degree of repair process. We also use the Expectation-maximization (EM) algorithm to update optimally the probability transitions of this process. In the second problem, the degree of repair is a random variable and belongs to a set of hypotheses hypothesis. At each epoch \(n,\) a list of \(M\) candidate models is available and the optimal one is chosen.
Chapter
This presentation, which is mainly of an expository nature, is divided into two parts. The first part is concerned with age replacement age replacement models with minimal repair minimal repair and the second part deals with block replacement block replacement models with minimal repair. Each part focuses mainly on the mathematical modeling of the notion of minimal repair. To limit the scope of this survey, we decided to limit ourselves almost exclusively to research on papers where a cost function cost function is designed specifically and optimal replacement replacement times optimal replacement times that yield minimum cost are sought.
Article
Full-text available
: A repairable system can briefly be characterized as a system which is repaired rather than replaced after a failure. The most commonly used models for the failure process of a repairable system are nonhomogeneous Poisson processes (NHPP), corresponding to minimal repairs, and renewal processes (RP), corresponding to perfect repairs. The paper reviews models for more general repair actions, often called "better-than-minimal repair" models. In particular we study the class of so called trend-renewal processes (TRP), which has both the NHPP and the RP as special cases. Parametric inference in TRP models is considered, including cases with several systems involving unobserved heterogeneity. Trend testing is discussed when the null hypothesis is that the failure process is an RP. It is shown how Monte Carlo trend tests for this case can be made from the commonly used trend tests for the null hypothesis of a homogeneous Poisson process (e.g. the Laplace test and the Military Handbook test)...
Article
Let N(t) be a counting process associated with a sequence of non-negative random variables (Xj)1 ∞ where the distribution of Xn+1 depends only on the value of the partial sum Sn = ∑j = 1 n Xj. In this paper, we study the structure of the function H(t)=E[N(t)], extending the ordinary renewal theory. It is shown under certain conditions that h(t)=(d/dt)H(t) exists and is a unique solution of an extended renewal equation. Furthermore, sufficient conditions are given under which h(t) is constant, monotone decreasing and monotone increasing. Asymptotic behavior of h(t) and H(t) as t → ∞ is also discussed. Several examples are given to illustrate the theoretical results and to demonstrate potential use of the study in applications.
Article
In this paper, we develop general repair models for a repairable system by using the idea of the virtual age process of the system. If the system has the virtual age V n – 1 = y immediately after the ( n – l)th repair, the n th failure-time X n is assumed to have the survival function where is the survival function of the failure-time of a new system. A general repair is represented as a sequence of random variables A n taking a value between 0 and 1, where A n denotes the degree of the n th repair. For the extremal values 0 and 1, A n = 1 means a minimal repair and A n = 0 a perfect repair. Two models are constructed depending on how the repair affects the virtual age process: V n = V n – 1 + A n X n as Model 1 and V n = A n ( V n – 1 + X n ) as Model II. Various monotonicity properties of the process with respect to stochastic orderings of general repairs are obtained. Using a result, an upper bound for E [ S n ] when a general repair is used is derived.
Article
The notion of the concealed age of lifetime distribution functions is introduced. It is based on the physical principle (law) of reliability that treats the cumulative wear ∫x0λ(u)du, where λ(u) is a failure-rate function, as a main reliability characteristic of components performance in a changing environment. These changes can result in different stresses or in alteration of the parameter in the distribution function, for instance. This approach is applied for describing the renewal process with nonidentical, dependent cycles when each repair changes the parameter of the distribution function and decreases wear. The renewal equations are derived. Some directions for further mathematical investigations are discussed.
A Monte Carlo Approach to Repairable System Reliability Analysis
  • M P Kaminskiy
  • V V Krivtsov
M. P. Kaminskiy and V.V. Krivtsov, "A Monte Carlo Approach to Repairable System Reliability Analysis" -in Probabilistic Safety Assessment and Management, Springer-Verlag London Ltd, 1998, pp. 1063-1068.
Imperfect Maintenance
  • M Brown
  • F Proshan
M. Brown, and F. Proshan, "Imperfect Maintenance" -in Crowley J. and Johnson R. (ed) Survival Analysis. Vol. 2, 1982, pp 179-188.