Page 1
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 84 | P a g e
Probabilistic Behaviour of A Redundant Complex System with
Imperfect Switching, Environmental Common Cause and Human
Error Effects under Head-Of-Line Repair Discipline
K. Uma Maheswari#1
, A. Mallikarjuna Reddy#1
and R. Bhuvana Vijaya#2
#1
Department of Mathematics, Sri Krishnadevaraya University, Ananthapuramu-515003, A.P., India #2
Department of Mathematics, J.N.T.U., Ananthapuramu-515001, A.P., India
ABSTRACT Series-parallel systems are made up of combination several series & parallel configuration to obtain system
reliability down into homogeneous subsystems. This is very simple to analyse a simple series-parallel system
without any error effects, But very difficult to analyse a complex system with various error effects. This paper
we presented mathematical analysis of a Redundant Complex System with Imperfect switching, Environmental
and Common Cause and Human error effects under Head of Line Repair Discipline.
Keywords: Imperfect switching, Reliability, Availability, MTTF, State Transition diagram.
I. INTRODUCTION Kontoleon & Konoleon [2] had considered a
system subject to partial and catastrophic failures to
be repaired at a single service station and also
assumed exponential distribution for repair. Gupta
and Mittal [3] considered a standby redundant system
and two types of failures to be repaired at a single
service station with general repair time distribution
under pre-emptive resume repair Discipline
incorporating environmental effects. Further, Gupta
and Agarwal [1] developed a model with two types of
failure under general repair time distribution and
under different repair discipline. Consequently
Agarwal, Mittal et. al., [4] have solved the model of a
Parallel redundant complex system with two types of
failures with Environmental effect under preemptive
Resume repair Discipline. Mittal, Gupta et. al., [1,3]
have assumed constant repair rate due to
environmental failure and perfect switching over
device of a standby redundant system under pre-
emptive resume and repeat discipline respectively.
But it is not always possible that the switch is perfect.
It may have some probability of failure. So the
authors have initiated the study keeping in mind the
practical aspect of imperfect switch. Also repair due
to environmental effects is considered to follow
general time distribution.
In this paper, we consider a complex system
consisting of three independent, repairable
subsystems A, B & C. Subsystem B is comprised of
two identical units B1 & B2. The subsystem C is in
standby redundancy to be switched into operation
when both the identical units B1 & B2 of subsystem B
fails, through an imperfect switching over device. For
the smooth operation of the system, the subsystems B
or C are of vital importance. The failure of the
subsystem C results into non-operative state of the
system. At the time of installation, the units of
subsystem B have the same failure rate but due to
adverse environmental effects, the failure rate of the
subsystem C increases by the time, it is switched into
operation. Subsystem A has minor failure and
subsystem C has major and minor failures. Minor
failure reduces the efficiency of the system causing a
degraded state while major failure results into a non-
operable state of the system. The whole complex
system can also be in a total failure state due to
Human error or Environmental effects like
temperature, humidity etc. The repair rate due to
Human error or Environmental failure is considered
general. Laplace transforms of the time dependent
probabilities of the system being in various states
have been obtained by employing the techniques
using supplementary variable under Head-of line
Repair Discipline. The general solution of the
problem is used to evaluate the ergodic behaviour
and some particular case. At the end of the chapter
some numerical illustrations are given and various
reliability parameters have been calculated. The flow
of states is depicted in the state transition diagram.
RESEARCH ARTICLE OPEN ACCESS
Page 2
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 85 | P a g e
The following symbols are used in state transition diagram:
: Operable : Degarded
: Imperfect : : Failed
switch
Fig. 1. STATE TRANSITION DIAGRAM
II. NOTATIONS: 1 : Minor failure rate of the operating units B1
& B2 of subsystem B.
2(1<2): Minor failure rate of the
operating unit of subsystem C
1m : Minor failure of subsystem A.
2m : Major failure of subsystem B & C.
x: Elapsed repair time for both units of
subsystem B.
y: Elapsed minor repair time for the
subsystem A, B & C.
z: Elapsed major repair time for the
subsystem B & C.
v: Elapsed repair time for environmental
failure.
h: Elapsed repair time for human error
failure.
B(x) : Repair rate of subsystem B
)(1
ym : Repair rate of minor failure of
subsystems A, B & C.
)(2
zm : Repair rate of major failure of
subsystems B & C.
: Failure due to environmental effects.
: Failure due to human error.
: Failure due to common cause.
(v) : Repair rate due to environmental
effects.
(h) : Repair rate due to Human error.
(c) : Repair rate due to common cause
b, R1 Probability of the successful operation
of the switching over device
and constant repair rate of
switching over device.
III. ASSUMPTIONS (1) Initially, at time t = 0 the system is in operable
state, i.e., it operates in its normal efficiency
state
(2) The subsystem B consists of two identical units
B1 & B2.
(3) Switching device is imperfect.
(4) When the system starts functioning the
subsystems A & B will operate and C is in stand
by.
(5) When the system starts functioning both the
subsystems B & C have the same failure rate 1
as minor failure, but as the time passes due to the
adverse environmental effects, the failure rate of
the stand by unit i.e., subsystems C increase to 2
which is the minor failure of subsystem C by the
time it is needed to operate.
(6) Subsystem A has no major failures.
(7) During the degraded state of the system due to
minor failure in subsystems A, B & C, major
failures may also occur in subsystems B & C.
(8) All the failures are distributed exponentially.
(9) All the repair follows general time distribution.
Page 3
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 86 | P a g e
(10) The system fails completely due to
environmental effects or due to major failures of
the subsystems B & C or due to the failure
caused by Human error.
(11) Repair of the subsystem C is taken only in failed
state.
(12) During the repair of the subsystem C in failed
state, the failed units of subsystem B are also
repaired.
STATE PROBABILITIES DESCRIPTION:
(i) )(0, tPa : Probability that the system is in
operable state at the time t, where a=0, 0, 1,
2.
(ii) ),(0,3 txP : Probability that the system is the
failed state and is under repair with elapsed
repair time in the interval (x, x+).
(iii) ),(1, tyP ma : Probability that the system is in
degraded state and is under repair with
elapsed repair time lying in the interval (y,
y+) where a = 0,1,2.
(iv) ),(2, tzP ma : Probability that the system is
the failed state and is under repair with
elapsed repair time in the interval (z, z+)
where a = 0,1,2.
(v) ),( tvPE : Probability that the system is in
failed state due to Environmental failure and
is under repair with elapsed Repair time in
the interval (v, v+).
(vi) )(tPr : Probability that the system is in
failed state due to failure of the switch in
switching the subsystem C when subsystem
B completely fails at any time t.
(vii) ),( thPh : Probability of the system is in
failed state due to Human error and is under
repair with elapsed repair time in the
interval (h, h+).
(viii) ),( tuPv : Probability of by the system is in
failed state due to common cause error and
is under repair with elapsed repair time in
the interval (u, u+).
IV. FORMULATION OF THE
MATHEMATICAL MODEL: Using continuity arguments and probability
considerations, we obtain the following difference
differential equations governing the stochastic
behaviour of the complex system, which is discrete is
space and continuous in time:
0
00
,2
0
0
,0
0
,0
0
0,30,01
)(),(
)(),()(),()(),(
)(),()(),(
)(),()(
22
2211
12
xdutuP
xdhthPdzztzPdxvtvP
dzztzPdyytyP
dxxtxPtPdt
d
C
HmmE
mmmm
Bmm
(1)
0
00
,2
0
0
,1
0
,1
0
0,30,11
)(),(
)(),()(),()(),(
)(),()(),(
)(),()(
22
2211
12
xdutuP
xdhthPdzztzPdxvtvP
dxztzPdyytyP
dxxtxPtPdt
d
C
HmmE
mmmm
Bmm
(2)
Page 4
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 87 | P a g e
0
,21
0,110,010,22
)(),()(
)()()(
11
21
dyytyPtPR
tPbtPbtPdt
d
mmr
mm
(3)
0
,1
0,1
0
,00,010,0
1
),(
)(),()()(
)1()(2
1
1
1
dytyP
tPdytyPtPbtP
btPRdt
d
m
m
r
(4)
0),()(2 0,3
txPx
xtB (5)
0),()(112 ,01
tyPy
ytmmm (6)
0),()(112 ,11
tyPy
ytmmm (7)
0),()(112 ,22
tyPy
ytmmm (8)
0),()(22 ,0
tzPz
ztmm (9)
0),()(22 ,1
tzPz
ztmm (10)
0),()(222 ,2
tzPz
ztmm (11)
0),()(2
thPh
htH (12)
0),()(2
tvPv
vtE (13)
0),()(2
tuPu
utC (14)
BOUNDARY CONDITIONS
0
,220,220,3 ),()(),0(1
dytyPtPtP m (15)
)(),0( 0,0,0 11tPtP mm (16)
)(),0( 0,1,1 11tPtP mm (17)
0
1,11
0
,010,2,2 )(),(),()(),0(1121
tPRdytyPbdytyPbtPtP rmmmm (18)
0
,00,0,0 ),()(),0(1222
dytyPtPtP mmmm (19)
Page 5
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 88 | P a g e
0
,10,1,1 ),()(),0(1222
dytyPtPtP mmmm (20)
0
,20,2,2 ),()(),0(1222
dytyPtPtP mmmm (21)
0
,2
0
,1
0
,00,20,10,0
),(),(
),()()()(),0(
11
1
dytyPdytyP
dytyPtPtPtPtP
mm
mE
(22)
0
,2
0
,1
0
,00,20,10,0
),(),(
),()()()(),0(
11
1
dytyPdytyP
dytyPtPtPtPtP
mm
mC
(23)
)(),0( 0,2 tPtPH (24)
INITIAL CONDITIONS
P0,0(0) = 1 and all other states probabilities are zero at time t = 0.
from 6
0),()(112 ,01
tyPy
ytmmm
Taking Laplace transforms on both sides
)(),(
),(
12
1
1
1
,0
,0
yy
SsyP
syPy
mm
m
m
Integrating ‘y’ b/w the limits ‘0’ to ‘y’
y
mm dyyys
mm esPsyP 0112
11
)(][
,0,0 ),0(),(
from equation (16)
)(),0( 0,0,0 11tPtP mm
Taking Laplace transforms on both sides
y
mm yyys
mm esPsyP 0112
11
)(][
0,0,0 )(),(
On simplification
)()(),( 10,0,0 11sVsPsyP mm (25)
where
1
1
1
2
21)(1
)(m
mm
s
sssV
from equation (7)
0),()(112 ,11
tyPy
ytmmm
Taking Laplace transforms on both sides
Page 6
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 89 | P a g e
)(),(
),(
12
1
1
1
,1
,1
yssyP
syPy
mm
m
m
Integrating ‘y’ between the limits ‘0’ to ‘y’
y
mm dyyys
mm esPsyP 0112
11
)(][
,1,1 ),0(),(
from equation 16
)(),0( 0,1,0 11tPtP mm
Taking Laplace transforms on both sides
)(),0( 0,1,1 11sPsP mm
y
mm dyyys
mm esPsyP 0112
11
)(][
0,1,1 )(),(
)()(),( 10,1,1 11sVsPsyP mm (26)
From equation 4
0
,1
0,1
0
,00,0
11
),(
)(),()(
)1()(2
1
1
dytyP
tPdytyPtP
btPRdt
d
m
m
r
Taking Laplace transforms on both sides
0
,1
0,1
0
,00,0
1
1
),(
)(),()(
2
)1()(
1
1
dysyP
sPdysyPsP
Rs
bsP
m
m
r
)]()(1()()(1[(2
)1(0,010,01
1
1
11sPsVsPsV
Rs
bmm
]))[(1[(2
)1(101
1
1
1PPsV
Rs
bm
])[( 102 PPsV (27)
where )(1[(2
)1()( 1
1
12 1
sVRs
bsV m
from equation 8
0),()(112 ,22
tyPy
ytmmm
Taking Laplace transforms on both sides
)(),(
),(
12
1
1
2
,2
,2
yssyP
syPy
mm
m
m
Integrating ‘y’ between the units ‘0’ to ‘y’
y
mm dyyys
mm esPsyP 0122
11
)(][
,2,2 ),0(),(
from equation 18
Page 7
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 90 | P a g e
0
1,11
0
,010,2,2 )(),(),()(),0(1111
tPRdytyPbdytyPbtPtP rmmmm
Taking Laplace transforms on both sides
)]()()[()()(),( 0,10,050,24,2 1sPsPsVsPsVsyP m
where
2
2
3
2
11)(1
)(m
mm
s
sssV (28)
1
)()( 34 msVsV
)()()( 211115 sVRsVmbsV
From equation (10)
0),()(22 ,1
tzPz
ztmm
Taking Laplace transforms on both sides
)(),(
),(
2
2
2
,1
,1
zsszP
szPy
m
m
m
Integrating ‘z’ between the limits ‘0’ to ‘z’
z
m dzzsz
mm esPszP 02
22
)(
,1,1 ),0(),(
From equation 19
0
,10,1,1 ),()(),0(1222
dytyPtPtP mmmm
Taking Laplace transforms on both sides
z
m dzzsz
mmmm edysyPsPszP 02
1222
)(
0
,10,1,1 ]),()([),(
)()(
)()()](1[),(
0,17
0,161,1 122
sPsV
sPsVsVszP mmm
(29)
where s
sssV
m )(1)( 2
6
)()](1[)( 617 12sVsVsV mm
From equation (12)
0),()(222 ,2
tzPz
tmm
Taking Laplace transforms on both sides
)(2),(
),(
2
2
2
,2
,2
zsszP
szPy
m
m
m
Integrating ‘z’ between the limits ‘0’ to ‘z’
z
m dzzsz
mm esPszP 02
22
)(
,2,2 ),0(),(
From equation 21.
Page 8
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 91 | P a g e
0
,20,2,2 ),()(),0(2222
dytyPtPtP mmmm
Taking Laplace transforms on both sides
z
m dzzsz
mmmm edysyPsPszP 02
1222
)(2
0
,20,2,2 ),()(),(
)()()(
)()()()()()](1[
0,085
0,1580,284
2
22
sPsVsV
sPsVsVsPsVsV
m
mm
)()()()()( 0,00,11029 sPsPsVsPsV (30)
where s
sssV
m )(21)( 2
8
)())(1()( 849 2sVsVsV m
)()()( 5810 2sVsVsV m
From equation (12)
0),()(2
thPh
htH
Taking Laplace transforms on both sides
)(2),(
),(
hsshP
shPh
H
H
Integrating ‘h’ between the limits ‘0’ to ‘h’
h
dhhsh
HH esPshP 0
)(2
),0(),(
From equation 22
)(),0( 0,2 tPtPH
Taking Laplace transforms on both sides
)(),0( 0,2 tPtPH
After simplification
)()(),( 0,211 sPsVshPH (31)
where s
sssV h )(21)(11
from equation 13
0),()(2
tvPv
vtE
Taking Laplace transforms on both sides
)(2),(
),(
vssvP
svPv
E
E
Integrating ‘v’ between the limits ‘0’ to ‘v’
v
dvvsv
EE esPsvP 0
)(2
),0(),(
From equation 14
Page 9
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 92 | P a g e
0),()(2
tuPu
utC
Taking Laplace transforms on both sides
)(),(
),(
ussuP
suPv
C
C
Integrating ‘u’ between the limits ‘0’ to ‘u’
u
duusu
CC esPsuP 0
)(2
),0(),(
From equation 22
0
,2
0
,1
0
,00,20,10,0
),(),(
),()()()(),0(
11
1
dytyPdytyP
dytyPtPtPtPtP
mm
mE
Taking Laplace transforms on both sides
0
,2
0
,1
0
,00,20,10,0
),(),(
),()()()(),0(
11
1
dysyPdysyP
dysyPsPsPsPsP
mm
mE
After simplification
)())()](1[
)()()]()(1[
)()()]()(1[)(
0,2124
0,11251
0,01251
1
1
sPsVsV
sPsVsVsV
sPsVsVsVsP
m
mE
)]()()[()()()( 0,10,0140,213 sPsPsVsPsVsPE (32)
where
s
sssV
)(21)(12
)()](1[)( 12413 sVsVsV
)()]()(1[)( 125114 1sVsVsVsV m
From equation 23
0
,2
0
,1
0
,00,20,10,0
),(),(
),()()()(),0(
11
1
dytyPdytyP
dytyPtPtPtPtP
mm
mC
Taking Laplace transforms on both sides
0
,2
0
,1
0
,00,20,10,0
),(),(
),()()()(),0(
11
1
dysyPdysyP
dysyPsPsPsPsP
mm
mC
After simplification
Page 10
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 93 | P a g e
)()()](1[
)()()]()(1[
)()()]()(1[)(
0,2124
0,11251
0,01251
1
1
sPsVsV
sPsVsVsV
sPsVsVsVsP
m
mC
)]()()[()()()( 0,10,0160,215 sPsPsVasPsVsPC (33)
where
s
sssV
)(21)(12
)()](1[)( 12415 sVsVsV
)()]()(1[)( 125116 1sVsVsVsV m
From equation 5
0),()(2 0,3
txPx
xtB
Taking Laplace transforms on both sides
)(2),(
),(
0,3
0,3
xssxP
sxPx
B
Integrating ‘x’ between the limits ‘0’ to ‘x’
x
B dxxsx
esPsxP 0
)(2
0,30,3 ),0(),(
From equation 14
0
,220,220,3 ),()(),0(1
dytyPtPtP m
Taking Laplace transforms on both sides
0
,220,220,3 ),()(),0(1
dysyPsPsP m
Substituting the value of P3,0(0,s) in the above equation we get
)]()()[()()(),( 0,00,1190,2180,3 sPsPsVsPsVsxP (34)
where
s
sssV B )(21)(17
)()](1[)( 174218 sVsVsV
)()()( 175219 sVsVsV
From equation 3
0
,21
0,110,010,22
)(),()(
)()()(
11
21
dyytyPtPR
tPbtPbtPdt
d
mmr
mm
Taking Laplace transforms on both sides
0
,21
0,110,010,22
)(),()(
)()()(
11
21
dyysyPsPR
sPbsPbsPS
mmr
mm
Page 11
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 94 | P a g e
Substituting the values of )(sPr , ),(,2 1 symP in the above equation we get
)]()()[()( 0,10,0220,2 sPsPsVsP (35)
)()()()( 2521120 121 mmm sSsVsVRbsV
)(
)()(
2
221
1211
12
mmmm
mm
sS
ssV
)(
)()(
21
2022
sV
sVsV
from equation 2
0
00
,2
0
0
,1
0
,1
0
0,30,11
)(),(
)(),()(),()(),(
)(),()(),(
)(),()(
22
2211
12
xdutuP
xdhthPdzztzPdxvtvP
dzztzPdyytyP
dxxtxPtPdt
d
C
HmmE
mmmm
Bmm
Taking Laplace transforms on both sides
0 0 0
,2
00
,1
0
,1
0
0,30,11
)(),()(),()(),(
)(),()(),()(),(
)(),()(
22
2211
12
dxusuPdxhshPdzzszP
dxvsvPdzzszPdyysyP
dxxsxPsPS
CHm
Emmmm
Bmm
m
Substituting the values of
),(0.3 sxP , ),(1.1 syP m , ),(
2.1 szP m , ),(2.2 szP m , ),( svPE , ),( shPH , ),( suPC
In the above equation we get
)()()()()( 0,0270,2260,1 sPsVsPsVsP (36)
where
)()()](1[()()](1[(
)()](1[()()(1()(
22 44
44223
sSsSsVsSsV
sSsVsSsVsV
mm
B
)()()()](
)(1[)()]()(1[)()()(
22
11
85
1515224
sSsVsSsV
sVsSsVsVsSsVsV
mm
mmB
)()](1[)()()()(21212 11124125 sSsVssmmsVSsV mmmmm
)(
)()(
25
2326
sV
sVsV
)(
)()(
25
2427
sV
sVsV
From equation 1
Page 12
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 95 | P a g e
0
00
,2
0
0
,0
0
,0
0
0,30,01
)(),(
)(),()(),()(),(
)(),()(),(
)(),()(
22
2211
12
xdutuP
xdhthPdzztzPdxvtvP
dzztzPdyytyP
dxxtxPtPdt
d
C
HmmE
mmmm
Bmm
Taking Laplace transforms on both sides
0 0 0
,2
00
,0
0
,0
0
0,30,01
)(),()(),()(),(
)(),()(),()(),(
)(),()(
22
2211
12
dxusuPdxhshPdzzszP
dxvsvPdzzszPdyysyP
dxxsxPsPS
CHm
Emmmm
Bmm
m
Substituting the values of
),(0.3 sxP , ),(1.0 syP m , ),(
2.0 szP m , ),(2.2 szP m , ),( svPE , ),( shPH , ),( suPC in the above
equation we get
1)()()()()()( 0,1240,2230,025 sPsVsPsVsPsP (37)
From equation 34
)]()()[(),( 0,10,0220,2 sPsPsVsxP
Substitute the value of )(0,1 sP in this equation we get
)(P)()()(P )()()()()( 0,027220,226220,0220,2 ssVsVssVsVsPsVsP
)()()(1
)](1)[()( 0,0
2622
27220,2 sP
sVsV
sVsVsP
)()(
)()( 0,0
28
290,2 sP
sV
sVsP (38)
where )()(1)( 262228 sVsVsV
)](1)[()( 272229 sVsVsV
From equation 36
)()()()()( 0,0270,2260,1 sPsVsPsVsP
Substitute the vale of )(0,2 sP in this equation we get
)()()(1
)()()()( 0,0
2226
2722260,1 sP
sVsV
sVsVsVsP
)()(
)()( 0,0
28
300,1 sP
sV
sVsP
where
)()()()( 27222630 sVsVsVsV
From equation 37
1)()()()()()( 0,1240,2230,025 sPsVsPsVsPsP
Page 13
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 96 | P a g e
Substitute the value of )(),( 0,10,2 sPsP in this equation we get
)(
)()(
31
280,0
sV
sVsP (39)
where )()()()()(P)()( 302429230,02531 sVsVsVsVssVsV
Substituting the value )(P 0,0 s in equation (37, 38) we get
)(
)()(
31
300,1
sV
sVsP (40)
)(
)()(
31
290,2
sV
sVsP (41)
Evaluation of operational Availability and Non-Availability
The Laplace transforms of the probability that the system is in operable up and down state at time ‘t’ can be
evaluated as follows.
)()]([1
)()]()([1)()]()([1
)()()()()()()(
0,24
0,151m0,051m
,2,1,00,20,10,0
11
111
sPsV
sPsVsVsPsVsV
sPsPsPsPsPsPsP mmmUP
(42)
)()()()()()()()()(222 ,2,1,00,3 sPsPsPsPsPsPsPsPsP CHrEmmmdawn
)()](()()()()([
))()()](()()()()([
0,2181513119
0,10,0191614107
sPsVsVsVsVsV
sPsPsVsVsVsVsV
(43)
ERGODIC BEHAVIOUR
Using Abel’s lemma is Laplace transform,
viz., )()(0
)(0
sayftft
ltssf
s
lt
, provided that the limit on the R.H.S exists, the time independent
up and down state probabilities are as follows.
)0(
)0()]0([1
)0(
)0()]0()0([1
)0(
)0()]0()0([1
)()()()()()()(
31
294
31
3051m
31
2851m
,2,1,00,20,10,0
11
111
V
VV
V
VVV
V
VVV
sPsPsPsPsPsPsP mmmUP
(44)
)()()()()()()()()(222 ,2,1,00,3 sPsPsPsPsPsPsPsPsP CHrEmmmdown
)0(
)0()]0()0()0()0()0([
)0(
)0(
)0(
)0()]0()0()0()0()0([
31
29181513119
31
30
31
28191614107
V
VVVVVV
V
V
V
VVVVVV
(45)
PARTICULAR CASE:- When repair follows exponential distribution setting
i
ii
ssS
)( and
ssS )( where i = B, m1, m2 in results (25) to (34) (39) (40) (41) one may get the various probabilities as
follows:
Page 14
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 97 | P a g e
)(
)()(
31
280,0
sg
sgsP
)(
)()(
31
300,1
sg
sgsP
)(
)()(
31
290,2
sg
sgsP
)]()()[()()()( 0,10,0190,2180,3 sPsPsgsPsgsP
)()()( 0,01,0 11sPsgsP mm
)()()( 0,11,1 11sPsgsP mm
)]()()[()( 0,10,02 sPsPsgsPr
)]()()[()()()( 0,10,050,24,2 1sPsPsgsPsgsP m
)()()( 0,07,0 2sPsgsP m
)()()( 0,17,1 2sPsgsP m
)]()()[()()()( 0,00,1100,29,2 2sPsPsgsPsgsP m
)()()( 0,211 sPsgsPH
)]()()[()()()( 0,10,0140,213 sPsPsgsPsgsPE
)]()()[()()()( 0,10,0160,215 sPsPsgsPsgsPC
where 1
115 )()(12
mmSsg
)](1[2
)1()( 1
1
12 1
sgRs
bsg m
1
23 )()(12
mmSsg
134 )()( msgsg
)()()( 211115 sgRsgmbsg
1
6 )()(2
mssg
)()](1[)( 617 12sgsgsg mm
)()(
2
2
8
m
m
ss
ssg
)()](1[)( 849 2sgsgsg m
)()]()( 8510 2sgsgsg m
)()(11
h
h
ss
ssg
)()(12
ss
ssg
)()](1[)( 12413 sgsgsg
)()]()(1[)( 125114 1sgsgsgsg m
)()](1[)( 12415 sgsgsg
Page 15
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 98 | P a g e
)()]()(1[)( 125116 1sgsgsgsg m
)()(17
B
B
ss
ssg
)()](1[)( 174218 sgsgsg
)()()( 175219 sgsgsg
1
2521120 )()()()(21
mm ssgsgRbsg
1
2
221
)(
)()(
1211
12
mmmm
mm
s
ssg
)(
)()(
21
2022
sg
sgsg
1
1
4
1
4
1
4
1
4223
)(
)()](1[)()](1(
)()](1[))()(1()(
222
s
ssgssg
ssgssgsg
mmm
BB
1
8
1
51
1
51
1
5224
)()()()]()(1[
)()]()(1[)()()(
2221
1
mmmm
mBB
ssgssgsg
ssgsgssgsg
1
1
1
24225
)()](1[)(
)()()(
2212111
12
mmmmmmm
mm
ssgs
sgssg
)(
)()(
25
2326
sg
sgsg
)(
)()(
25
2427
sg
sgsg
)()(1)( 262228 sgsgsg
)](1)[()( 272229 sgsgsg
)()()()( 27222630 sgsgsgsg
)()()()()()()( 30242923282531 sgsgsgsgsgsgsg
UP and Down state probabilities
The Laplace transforms of up and down state probabilities are as follows.
)()](1[
)()]()(1[)()]()([1
)()()()()()()(
0,24
0,1510,051m
,2,1,00,20,10,0
11
111
sPsg
sPsgsgsPsgsg
sPsPsPsPsPsPsP
m
mmmUP
(44)
)(][
)(][
)(][
)()()()()()()()()(
0,2181513119
0,119164107
0,0191614107
,2,1,00,3 222
sPggggg
sPggggg
sPggggg
sPsPsPsPsPsPsPsPsP CHrEmmmdown
RELIABILITY: Laplace transforms of the reliability of the system given as follows:
Page 16
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 99 | P a g e
))((1
1)( 111 111
DsBs
b
Bs
b
Cs
b
CsASsR
mmm
Taking inverse Laplace transforms of the above equation we get
))((
)()())(()(
)())((1)(
1
11
111
1
1111
111
DBBA
be
AC
b
AC
me
BDBA
b
AB
be
AC
b
ADAB
b
AB
b
ACetR
mDt
mCtmBt
mmmAt
M.T.T.F Mean time to system failure is the expected time to operate the system successfully which is given as
follows:
0
)(.. dttRFTMT
))(()()(1
111 1111
ABAD
b
AC
b
AC
m
AB
mb
A
mm
))((
1
)()(
1
))((
1
1
111
1
111
DADB
b
D
CA
b
ACCBABD
b
AB
b
B
m
mmm
where 121 mmA , 121 mmB
12mC , 22mD
Numerical Illustrations:
Reliability Analysis:
01.01 , 02.02 , 011.01m , 015.0
2m , = 0.01, = 0.02, = 0.03, b = 0.96 and for
different values of t in the equation (60) one may obtain the reliability of the system as given in fig 2. The
reliability of the system decreases slowly as the time period increases. It also depicts the reliability of the system
for a long time period.
M.T.T.F
1 = 0.01, 2 = 0.02, 01.01m , b = 0.96, = 0.02 and taking different values of
2m in the equation (61)
one may obtain the variations of M.T.T.F. of the system against the environmental failure rate shown in figure
3.
The variations in M.T.T.F w.r.to Environmental failure rate as the major failure rate of the subsystem A
increases. The series of curves Represents that MTTF decreases as the environmental failure increases
apparently.
Page 17
K. Uma Maheswari et al Int. Journal of Engineering Research and Applications www.ijera.com
ISSN : 2248-9622, Vol. 4, Issue 6( Version 2), June 2014, pp.84-100
www.ijera.com 100 | P a g e
0
0.2
0.4
0.6
0.8
1
1.2
0 5 10 15 20 25
Time (t)
Re
lia
bil
ity
Fig. 5.2.
Fig. 2 Fig. 3
Fig. 5.4
0
0.5
1
1.5
2
2.5
3
3.5
4
4.5
5
0 0.02 0.04 0.06 0.08 0.1 0.12Environmental Failure Rate
M.T
.T.F
.
Series1
Series2
Series3
Fig. 5.5.
Series 1 2 3
2.02m 4.0
2m 6.0
2m
Fig. 4 Series 1 2 3
2.01m
4.02m
6.03m
Fig.5
Fig. 5.6.
Fig.6
V. DISCUSSION In this paper we presented mathematical models
of imperfect switching, environmental, common
cause and human error effects, so probabilistic
behaviour considering various values of coefficient
of human error, common cause, environmental
effects respectively.
VI. CONCLUSIONS : The Reliability, MTTF curves are plotted in figures (2),
(3). From these graphs we observe that
i) The Reliability of the system decreases slowly as
the time period increases.
ii) The MTTF decreases as the environmental
failure increases
REFERENCES [1] Gupta, P.P. and Agarwal, S.C. : A parallel
redundant complex system with two types of
failure under different repair discipline, IEEE
Trans. Reliability, Vol R-32, 1983.
[2] Kontoleon, J.M. and Kontoleon, N., 1974.
Reliability Analysis of a system subjected to
partial & Catastrophic Failures, IEEE Trans.
On Rel., Vol. R-23, pp 277-278.
[3] Mittal, S.K., Gupta, Ritu and C.M. Batra,
2007. Probabilistic behavior of a Redundant
Complex System, With Imperfect Switching
and Environmental Effects under Head of Line
Repair Discipline, Bulletin of pure and
Applied Sciences, Vol. 6E(1), 2007, pp. no.43-
58.
[4] Mittal S.K., Agarwal, S.C., and Kumar Sachin,
2004. Operational Behaviour of a parallel
redundant complex system with two types of
failures with environmental effect under Pre-
Emptive Resume Repair Discipline, Bulletin of
Pure and Applied Sciences, Vol. 23E (No.2),
pp.253-263.