Академический Документы
Профессиональный Документы
Культура Документы
Availability and cost analysis of an engineering system involving subsystems in series configuration
Anuj Kumar, Sangeeta Pant, S. B. Singh,
Article information:
To cite this document:
Anuj Kumar, Sangeeta Pant, S. B. Singh, (2017) "Availability and cost analysis of an engineering system involving
subsystems in series configuration", International Journal of Quality & Reliability Management, Vol. 34 Issue: 6, doi: 10.1108/
IJQRM-06-2016-0085
Permanent link to this document:
http://dx.doi.org/10.1108/IJQRM-06-2016-0085
Downloaded on: 25 April 2017, At: 23:19 (PT)
References: this document contains references to 0 other documents.
To copy this document: permissions@emeraldinsight.com
The fulltext of this document has been downloaded 5 times since 2017*
Access to this document was granted through an Emerald subscription provided by emerald-srm:543096 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.
Abstract.
Purpose: The Purpose of this paper is the evaluation of various reliability measures like
engineering system incorporating waiting time to repair. The considered system consist two
subsystems, namely, A and B connected in series. The subsystem B has two identical units in
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
standby arrangement. Each unit of the subsystem has two modes i.e. normal efficiency or failed.
The two standby units of the subsystem B are connected by an imperfect switching. The system
family of copula.
Findings: Numerical examples with a way to highlight the important results have been appended
at last. Numerical calculation shows that availabilty and reliability of the system is decreasing
with respect to time when failure rates are fixed at different values. Finally, cost analysis of
system reveals that the expected profit decreases with increase in service cost.
switching has been taken into conideration, which is consistent with actual failures of switching
by assuming two different types of failure between adjacent transition. It is evaluated with the
Key words: Availability, Reliability, Mean Time to failure (M.T.T.F.), profit function.
As far as planning, design and operations of any complex system are concerned the reliability
measures are the foremost concern. Reliability measures are very effective and efficient tool for
probabilistic risk assessment in system design, operation, and maintenance. The determination of
subsystems has received much attention over the years. Gupta and Sharma (1993), Singh and
Goel (1996), Jain et al. (2002), Jain et al. (2004), Rehmert and Nachlas (2009), Ghasemi et al.
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
(2010), Singh et al. (2011) develop different mathematical models and computed various
reliability measures such as availability, mean time to failure, reliability of complex engineering
system with different assumptions. In many cases it is not always possible to have an immediate
repair facility. Gopalan and Naidu (1982) discussed reliability system in which the repair facility
exists at random interval of time as it is not immediately available after detection of failure. Pirie
and Bendell (1984) have extended the study on the model taken by Bendell and they proposed
failures are assumed to be dependent on the actual operating time as well as elapsed calendar
time. The model was applied in the situation where all the system components are regularly
switched between the ‘on’ and ‘off’ position. Pan (1997) discussed two types of imperfect
switching system: (1) Non repairable system with only one standby component, one switch and
one sensor. (2) A non-repairable system with two standby components, one switch and one
sensor. They considered three modes of switch failure: (1) Failure under energized condition. (2)
Initially reliability models were based on the assumption that only one type of repair/failure is
possible between two adjacent states. However, there are many situations where more than one
repair/failure is possible between two adjacent states, and, when this happen, the system is
studied by copula. The use of copula in reliability analysis has been extensively studied by
Gennheimer (2002), Nelsen (2006), kumar and Singh (2008), Brunel et al. (2010) and Ram and
Singh (2010). Copula provides coupling of different distributions. Kumar and Singh (2008) has
considered a complex engineering system in which two different types of repairs possible
between two adjacent states and he successfully employing the concept of copula for the
determination of various reliability measures. Later Ram and Singh (2008) and Ram et al.
(2013), solved a complex system and evaluated various reliability measures with the help of
copula. Singh and Rawal (2014) have done the cost analysis of a complex system consisting two
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
subsystems, in series configuration and handling by a human operator. Kumar et al. (2015) and
Pant et al.(2015) have done the reliability optimization of some complex systems by using nature
The present contribution is structured as follows. Section 2 gives the description of the
mathematical model. It provides the details of assumptions, describes the formulation and
solution of the mathematical model and asymptotic behavior of the system. Section 3 presents
the particular cases of reliability measures, and Section 4 gives the conclusions of the proposed
analysis. The block diagram and transition state diagram of the system is shown in Figure 1 and
Figure 2 respectively.
Here we have considered a complex system consisting of two subsystems A and B which
connected in series. The subsystem B consists of two identical units in standby arrangement
(Figure 1). Each unit of the subsystem has two modes namely normal efficiency or failed. The
whole system can fail, if one of the subsystems A and B fails or due to environmental
failure/critical human error. Dhillon (1989), Dhillon and Liu (2006) define human error as the
failure to perform a specified task that results in disruption of scheduled operations or result in
damaging equipment. Repair is not immediately available after detection of failure and repair
The two standby units of the subsystem B are connected by an imperfect switching. All the
repairs follow general time distributions whereas all the failures follow exponential time
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
distributions.
In contrast to earlier models here authors have taken an important aspect of switching which is
consistent with actual failures of switching by assuming two different types of failure between
adjacent transition states S3 and S4 and it is evaluated with the help of the Gumbel–Hougaard
family of copula. At the same time by employing Supplementary variable technique[Cox (1995)
and Oliveira et al. (2005)] and Laplace transformation technique; various transition state
probabilities, Availability, reliability and mean time to failure (M.T.T.F) of the system have been
evaluated. Some particular cases also have been discussed for different value of failure rates. The
(2) Each unit has two modes namely- normal efficiency or failed.
(4) The system can fail on the failure of either subsystem, or environmental failure or human
error.
(5) Switching over device is imperfect.
(6) Joint probability distribution of imperfect switching over device from state S3 to S4 is
(7) All repairs follow general time distribution whereas all failures follow exponential time
distribution.
(8) After repair, the system works like a new one and never damage anything.
(9) The system waits for repair if repair facility is not available.
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
[ ]
1
= exp {ln(log b x )} θ
( )
+ ln a θ θ (1)
By probability considerations and continuing arguments, we can obtain the following set of
∂ ∞ ∞
∂t + λ + λ1 + λE1 + λh1 PO,O,S (t) = ∫ PF,O,S (z, t)β (z)dz + ∫ PE (K,t)βE (K)dK
0 0
∞ ∞
+ ∫ PO, F ,F ( y, t )β1 ( y)dy + ∫ PH ( N , t )β H ( N )dN (2)
0 0
∂
{
∂t + a + exp (ln(logb x) ) + (ln a )
θ θ 1/ θ
}
PO, F ,S (t ) = λPO,O, S (t )
(3)
∂ ∂
∂x + ∂t + β ( x) PO, F ( x, t ) = 0 (4)
∂ ∞
∂t + λ ′ + λE2 + λh2 PO, F ,O (t ) = aPO, F , S (t ) + ∫ PO, F ( x, t ) β ( x)dx (5)
0
∂ ∂
+ + β1 ( y ) PO , F , F ( y, t ) = 0 (6)
∂y ∂t
∂ ∂
∂z + ∂t + β ( z ) PF ,O , S ( z , t ) = 0 (7)
∂ ∂
∂K + ∂t + β E ( K ) PE ( K , t ) = 0 (8)
∂ ∂
∂N + ∂t + β H ( N ) PH ( N , t ) = 0 (9)
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
Boundary Conditions
[
PO , F (0, t ) = exp {ln (log b x )}θ + (ln a )θ ]
1/θ
PO , F , S (t )
(10)
On taking Laplace transforms of equations (2) through (14) and using (15), we obtained
∞ ∞
+ ∫ PE (K, s)βE (K)dK + ∫ PO, F, F ( y, s)β1( y)dy
0 0
∞
+ ∫ PH ( N , s ) β H ( N )dN (16)
0
{
s + a + exp (ln(log b x) ) + (ln a )
θ θ 1/ θ
}
PO, F , S ( s ) = λPO ,O ,S ( s )
(17)
∂
∂x + s + β ( x) PO, F ( x, s ) = 0 (18)
[s + λ ′ + λE2 + λh2 ]PO, F ,O (s) = aPO, F ,S (s) + ∞∫ PO, F ( x, s)β ( x)dx (19)
0
∂
+ s + β1 ( y ) PO , F , F ( y, s ) = 0 (20)
∂y
∂
∂z + s + β ( z ) PF ,O, S ( z , s ) = 0 (21)
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
∂
∂K + s + β E ( K ) PE ( K , s ) = 0 (22)
∂
∂N + s + β H ( N ) PH ( N , s ) = 0 (23)
[
PO, F (0, s ) = exp {ln (log b x )}θ + (ln a )θ ]
1/θ
PO , F , S ( s ) (24)
Now on integrating equation (16) through (23) and using (24) through (28) one may obtain
1
PO,O, S ( s ) = (29)
γ (s)
−1
PO, F , S ( s ) =
1
{
⋅ λ s + a + exp (ln(log b x) )θ + (ln a )θ
γ (s)
}
1/θ
(30)
−1
PO, F ( s ) =
1
{
⋅ s + a + exp (ln(logb x) )θ + (ln a )θ
γ ( s )
}1/θ
×
{
× exp (ln(logb x) )θ + (ln a )θ } θ .D ( s )
1/
x (31)
−1
1
{
s + a + exp (ln(logb x) )θ + (ln a )θ }
1/θ
⋅λ
PO, F ,O ( s ) = ⋅ ×
γ (s) [
s + λ ′ + λE2 + λh2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
} θ ⋅ S (s)
1/
x (32)
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
−1
1
{
λλ ′ ⋅ s + a + exp (ln(logb x) )θ + (ln a )
}
θ 1/θ
PO, F , F ( s ) = ⋅ ×
γ (s) [s + λ ′ + λE2 + λh2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
} θ ⋅ S (s).D (s)
1/
x 1 (33)
λ1
PF ,O, S ( s ) = ⋅ DZ ( s ) (34)
γ (s)
−1
1
{
λλ E 2 ⋅ s + a + exp (ln(logb x) )θ + (ln a )θ
1 / θ
×
}
PE ( s) = ⋅ λ E1 +
γ (s)
[
s + λ ′ + λ E 2 + λ h2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
} θ ⋅ S (s) ⋅ D
1/
x E (s) (35)
−1
1
{
λλh2 ⋅ s + a + exp (ln(logb x))θ + (ln a )θ
1 / θ
×
}
PH (s) = ⋅ λh1 +
γ ( s)
[
s + λ ′ + λE2 + λh2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
1/θ
}
⋅ S x ( s ) ⋅ DH ( s )
(36)
−1
{
λλ ′ ⋅ s + a + exp (ln(logb x) )θ + (ln a )θ
1/θ
}
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
− ×
[
s + λ ′ + λ E2 + λh2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
1/θ
}
⋅ S x ( s ).S1 ( s )
−1
{
λλ E2 ⋅ s + a + exp (ln(logb x) )θ + (ln a )θ
1 / θ
×
}
− λE1 +
[
s + λ ′ + λE2 + λh2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
1/θ
}
⋅ S x ( s ) ⋅ S E ( s )
−1
λλh2
{
⋅ s + a + exp (ln(logb x) )θ + (ln a )
}
θ 1/θ
− λh1 + ×
[
s + λ ′ + λE2 + λh2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
1/θ
}
⋅ S x ( s ) ⋅ S H ( s )
1 − Si ( s )
And Di ( s) = ∀ i. (37)
s
Using Abel’s lemma viz. lim s →0 sF ( s) = limt →∞ A(t ) = F ( say ) , provided the limit on R.H.S exists, in
equation (29) through (36), the time independent probabilities are obtained as follows:
1
PO,O, S = (39)
γ ′(0)
−1
PO, F , S =
1
{
⋅ λ a + exp (ln(logb x) )θ + (ln a )θ
γ ′(0)
1/θ
} (40)
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
PO, F =
{
exp (ln(logb x) )θ + (ln a )θ } θ ⋅ λ ⋅η
1/
x (41)
γ ′(0)
1 λ
PO, F ,O = ⋅ (42)
γ ′(0) (λ ′ + λ E2 + λh2 )
1 λλ ′
PO , F , F = ⋅ η1 (43)
γ ′(0) (λ ′ + λ E2 + λ h2 )
λ1
PF ,O, S = ⋅ηZ (44)
γ ′(0)
1 λλ E2
PE = ⋅ λE1 + ⋅ηE
′
γ (0) ′ [
s + λ + λE2 + λh2 ]
(45)
1 λλh2
PH = ⋅ λh1 + ⋅ηH
γ ′(0) [
λ ′ + λE2 + λh2 ] (46)
Where γ ′(0) =
d
γ ( s) (47)
ds s =0
βE
S E (s) = etc. in equation (29) through (36), we get the following results
(s + β E )
1
PO ,O, S ( s ) = (49)
E (s)
−1
PO, F , S ( s ) =
1
{
⋅ λ s + a + exp (ln(logb x) )θ + (ln a )θ
E ( s)
1/θ
} (50)
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
PO, F ( s ) =
1
E ( s)
{
⋅ exp (ln(logb x) )θ + (ln a )θ
1/θ
}
×
−1
{
× λ s + a + exp (ln(logb x) )θ + (ln a )θ
1/θ
} 1
(s + β )
(51)
−1
{
λ s + a + exp (ln(logb x) )θ + (ln a )θ
1/θ
}
⋅
1
PO, F ,O ( s ) = ×
E (s) [
s + λ ′ + λE2 + λh2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
1/θ
⋅ }β
(s + β )
(52)
−1
1
{
λλ ′ s + a + exp (ln(logb x) )θ + (ln a )θ
1/θ
}
PO, F , F ( s ) = ⋅ ×
E (s) [
s + λ ′ + λE2 + λh2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
1/θ
⋅} β
.
1
( s + β ) ( s + β1 )
(53)
λ1 1
PF ,O, S ( s ) = ⋅ (54)
E (s) (s + β z )
−1
1
{
λλ E 2 s + a + exp (ln(log b x) )θ + (ln a )θ
}
1/θ
PE ( s ) = ⋅ λE + ×
E (s) 1
[
s + λ ′ + λE 2 + λh2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
1/θ
⋅ } β
⋅
1
( s + β ) ( s + β E )
(55)
−1
1
{
λλh2 s + a + exp (ln(logb x))θ + (ln a )θ
1 / θ
×
}
PH (s) = ⋅ λh1 +
[ ]
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
{
× a + exp (ln(logb x) )θ + (ln a )θ }
1/θ
.
β 1
(s + β ) s + β
]. (56)
−1
βz
{
λλ ′ s + a + exp (ln(logb x ) )θ + (ln a )θ
}1/θ
Where E ( s) = s + λ + λ1 + λE1 + λh1 − λ1 − ×
(s + β z ) [
s + λ ′ + λ E 2 + λ h2]
{
× a + exp (ln(logb x) )θ + (ln a )θ }
1/θ
⋅
β β1
.
( s + β ) ( s + β1 )
−1
{
λλ E 2 s + a + exp (ln(log b x ) )θ + (ln a )θ
}1/θ
− λ E1 + ×
[
s + λ ′ + λ E 2 + λ h2 ]
{
× a + exp (ln(logb x) )θ + (ln a )θ
1/θ
⋅ } β
⋅
βE
( s + β ) ( s + β E )
−1
{
λλh2 s + a + exp (ln(logb x) )θ + (ln a )θ
1 / θ
×
}
− λh1 +
[
s + λ ′ + λE2 + λh2 ]
{
× a + exp (ln(logb x ) )θ + (ln a )θ
1/θ
⋅
+
}
β
β
⋅
βH
+ βH )
(57)
( s ) ( s
1 λa
Pup ( s ) = 1 + (58)
( s + λ + λ1 + λE1 + λh1 ) ( s + 1)( s + λ ′ + λE2 + λh2 )
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
− (λ + λ1 + λ E1 + λh1 )t − (λ + λ E 2 + λh2 )t
Pup (t ) = (1 + B )e + Ae − t + Ce (59)
Where
aλ
A= (60)
(λ + λ1 + λE1 + λh1 − 1)(λ ′ + λE2 + λh2 − 1)
aλ
B= (61)
(1 − λ − λ1 − λ E1 − λh1 )(λ ′ + λE2 + λh2 − λ − λ1 − λE1 − λh1 )
aλ
C= (62)
(1 − λ ′ − λE2 − λh2 )(λ + λ1 + λE1 + λh1 − λ ′ − λE2 − λh2 )
1
Also R (s) = (63)
( s + λ + λ1 + λE1 + λh1 )
− (λ + λ1 + λ E1 + λh1 )t
R (t ) = e (64)
The unreliability of the system can be obtain by the expression 1 − Pup (t ) , i.e.,
λ1 = 0.05, λE1 = λE 2 = 0.01 and λh1 = λh2 = 0.04 in equation (59) & (64), one may obtain
One may obtain different values of Pup (t ) and R(t ) from equation (66) & (67), for t = 0, 1, 2, 3, 4,
Taking all repairs zero and the limit as ‘s’ tending towards zero in equation (58) for the
1
M .T .T .F = lim s →0 Pup ( s ) = (68)
(λ + λ1 + λE1 + λh1 )
One may obtain variation of M.T.T.F. with respect to failure rates as shown in Figure 5 just by
setting, λ = 0.01, λ1 = 0.05, λE1 = 0.01 and λh1 = 0.04 , and varying λ , λ1 , λE1 , λh1 one by one
respectively 0.01, 0.02, 0.03, 0.04, 0.05, 0.06, 0.07, 0.08, 0.09 in equation (68).
Suppose that the service facility is always available, then the expected profit during the interval
[0, t[ is
t
E p (t ) = K1 ∫ Pup (t ) dt − K 2t (69)
0
For the same set of parameter of equation (66) & (67) one can obtain equation (70).
Therefore
[ ]
E p (t ) = K1 − 7.048e (− 0.11t ) − 0.0073e (− t ) − 3.623e (− 0.06 )t + 10.678 − K 2t (70)
Setting revenue cost ( K1 ) at 1 and varies service cost ( K 2 ) as 0.01, 0.05, 0.10, 0.15, 0.20, 0.25
and 0.5 respectively and varying t = 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 and 10 unit of time in equation (70),
one may get expected profit as a function of time as shown in Figure 6. Table 3 that again in this
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
4. Conclusion
Figure 3 & Figure 4 provide information about the changes in availability and reliability of the
complex engineering system with respect to time when failure rates are fixed at different values.
It is clear from the data presented in Table 2 that availability as well as reliability of the system
decreases as ‘t’ increases and it became steady after a sufficient long interval of time so the
future behavior of a complex system is easily predictable for any given set of parametric values.
Table 3 and Figure 5 yields the M.T.T.F. of the system with respect to variation in λ / λ E1 , λ1 and
λh1 respectively, when the other parameters have been taken as constant. From these graphs, one
can conclude that the Mean time to failure of the system decreases smoothly with each failure
rate.
Table 4 and Graphs in Figure 6 demonstrate the relation between time ‘t’ and expected profit
‘EP(t)’ when the revenue cost (K1) is fixed at 1 and service cost (K2) varies as 0.01, 0.05, 0.10,
0.15, 0.20, 0.25, and 0.5. One can clearly observe that expected profit decreases very far with
respect to time when the service cost is fixed at a very high value 0.5. So one may analyze the
behavior of profit function with respect to time for different set of values of parameters and it
5. References
Dhillon, B. S. (1989), “Human errors: A review”, Microelectronic Reliability, Vol. 29, No.
3, pp. 299–304.
Gennheimer, H. (2002), “Model Risk in copula based default pricing models, working paper
series”, Swiss Banking Inst., Univ. of Zurich and NCCR, FINRISK, Zurich, Switzlerland,
Ghasemi, A., Yacout, S., and Ouali, M. S. (2010), “Evaluating the reliability function and the
mean residual life for equipment with unobservable states”, IEEE Transaction on Reliability,
standby system subject to inspection”, Microelectronics Reliability, Vol. 22, No.4 , , pp.
699-705.
Gupta, P.P., Sharma, M.K. (1993), “Reliability and MTTF evaluation of a two duplex-unit
standby system with two types of repair,” Microelectronic Reliability, Vol. 33, No. 3, pp.
291-295.
Jain, M., Kumar, A. and Sharma, G. C. (2002), “Maintenance Cost analysis for Replacement
Model with Perfect/Minimal Repair”, International Journal of Engineering, Vol. 15, No. 2,
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
pp. 161-168.
Repairable System with Degraded Failure”, International Journal of Engineering, Vol. 17,
Kumar, A. and Singh, S.B. (2008), “Reliability analysis of an n-unit parallel standby system
under imperfect switching using copula”, Computer Modeling and New Technologies, Vol.
Kumar, A., Pant, S., Singh, S.B., (2015), “ Reliability Optimization of Complex System by
Using Cuckoos Search algorithm” accepted for publication in Mathematical Concepts and
Nelsen, R. B. (2006), “An Introduction to Copulas”, 2nd ed. New York: Springer-Verlag,
(2006).
Oliveira, E. A., Alvim, A. C. M., and Melo, P. F. F. (2005), “Unavailability analysis of
safety systems under aging by supplementary variables with imperfect repair”, Annals of
Pan, J.N. (1997), “Reliability prediction of imperfect switching systems subjected to multiple
Pant, S., Kumar, A., Kishor, A., Anand, D., & Singh, S.B. (2015), “Application of a Multi-
Problem”, In the proceeding of IEEE Int. Conf. on Next generation Computing Technologies,
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
Pirie, K. and Bendell, A. (1984), “Optimum component switching for systems that operate
Ram, M. and Singh, S.B., Singh .V.V. (2013), “Stochastic Analysis of a Standby System
with Waiting Repair Strategy”, IEEE transactions on systems, man, and cybernetics, systems,
Ram, M. and Singh, S.B. (2010), “Availability, MTTF and cost analysis of complex system
International journal of Quality & Reliability management, Vol. 27, No. 5, pp. 576-595.
Ram, M. and Singh, S.B. (2008), “Availability and Cost Analysis of a parallel redundant
complex system with two types of failure under preemptive-resume repair discipline using
IEEE Transaction on Systems, Man, Cybernetics. A, Syst., Humans, Vol. 39, No.1, pp. 272–
180.
Singh, J. and Goel, P. (1996), “Availability analysis of a standby complex system having
imperfect switching-over device”. Microelectronic Reliability, Vol. 35, No. 2, pp. 285-288.
Singh, S. B., Ram.M. and Chaube, S. (2011) , “Analysis of the reliability of a three
Singh, V.V., Rawal, D.K. (2014), “Availability, MTTF and Cost analysis of complex system
under Preemptive resume repair policy using copula distribution”, Pakistan journal of
Constant
λE1 / λE2 λh1 / λh2 Constant human error
environmental
rates.
failure rates.
Constant failure rate of
λ1 Constant failure rate
λ′ / λ main unit/standby unit
of subsystem A.
of subsystem B.
The probability that
at time‘t’, the system The probability that at
is in the “good time‘t’ system is in
PO , O , S (t ) / PO , F , O (t ) state/good state PO , F , S (t ) failed state due to
while main unit of failure of one unit of
subsystem B is subsystem B.
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
already failed.”
The probability that The probability that at
at time‘t’, the system time‘t’, system is in
is in failed state failed state due to
due to human error/ PO , F ( x, t ) ∆
failure of switch and
PH ( N , t )∆ / PE ( K , t )∆
environmental cause, elapsed repair time for
elapsed repair time this lie in the interval
lies in the interval (x, x+∆).
(N, N+∆)/ (K, K+∆).
The probability that at
General repair rate time‘t’, the system is
for the state of in the failed state due
β H (N ) / β E (K ) human error/ PO , F , F ( y , t ) ∆ to failure of subsystem
environmental B, elapsed repair time
failure. for the same lies in the
interval (y, y+∆).
The first order
probability that the
The probability that at
system will be
time‘t’, the system is
repaired completely
in the failed state
in the time interval
β ( z )∆ / β1 ( y )∆ PF ,O , S ( z , t ) ∆ due to failure of
(z, z+∆)/(y,
subsystem A, elapsed
y+∆),conditioned repair time for this lies
that it was not
in the interval (z, z+∆).
repaired completely
up to time z/y.
Constant/
logarithmic
a / log b x imperfect switching θ Parameter in [0,1].
rates over device
(b<1).
β (s) General repair rate Laplace transformation
A (s )
for switch. of function A (t).
Constant switching Laplace transform
a s
rate over device. variable.
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
Table 2. Availability & Reliability of System vs Time
0 1 1
1 0.901965 0.895834
2 0.815997 0.802519
3 0.739333 0.718924
4 0.670468 0.644036
5 0.608412 0.57695
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
6 0.552408 0.516851
7 0.501823 0.463013
8 0.456107 0.414783
9 0.414774 0.371577
10 0.377387 0.332871
Downloaded by Fudan University At 23:19 25 April 2017 (PT)
Table 4. Expected profit
0 0 0 0 0 0 0 0