introduction to repairable systems stk4400 -- spring 2011
TRANSCRIPT
![Page 1: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/1.jpg)
Introduction to repairable systemsSTK4400 – Spring 2011
Bo Lindqvist
http://www.math.ntnu.no/∼bo/[email protected]
Bo Lindqvist Introduction to repairable systems
![Page 2: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/2.jpg)
Definition of repairable system
Ascher and Feingold (1984):
“A repairable system is a system which, after failing to performone or more of its functions satisfactorily, can be restored to fullysatisfactory performance by any method, other than replacement ofthe entire system”.
Bo Lindqvist The trend-renewal process
![Page 3: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/3.jpg)
Ascher and Feingold’s mission in 1984
Ascher and Feingold presented the following example of a “happy”and “sad” system:
Their claim:
Reliability engineers do not recognize the difference between thesecases since they always treat times between failures as i.i.d. and fitprobability models like Weibull.
Use nonstationary stochastic point process models to analyze re-pairable systems data!
Bo Lindqvist The trend-renewal process
![Page 4: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/4.jpg)
Today: Recurrent events extensively studied
q q q
0 T1 T2 · · · TN τ
Observe events occurring in time
Applications: engineering and reliability studies, public health,clinical trials, politics, finance, insurance, sociology, etc.
Reliability applications:
breakdown or failure of a mechanical or electronic system
discovery of a bug in an operating system software
the occurrence of a crack in concrete structures
the breakdown of a fiber in fibrous composites
Warranty claims of manufactured products
Bo Lindqvist The trend-renewal process
![Page 5: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/5.jpg)
Important aspects for modelling and analysis
q q q
0 T1 T2 · · · TN τ
Trend in times between events?
Renewals at events?
“Randomness” of events?
Dependence on covariates?
Unobserved heterogeneity (“frailty”, “random effects”) amongindividual processes?
Dependence between event process and the censoring at τ?
Bo Lindqvist The trend-renewal process
![Page 6: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/6.jpg)
Typical data format
q q q
0 T11 T21 · · · TN11 τ1
...
q q q
0 T1j T2j · · · TNj j τj
...
q q q
0 T1m T2m · · · TNmm τm
Covariates if available (fixed or time-varying):
Xj(t); j = 1, 2, . . . ,m
Bo Lindqvist The trend-renewal process
![Page 7: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/7.jpg)
Proschan (1963): The classical “aircondition data”
Times of failures of aircondition system in a fleet of Boeing 720airplanes
Bo Lindqvist The trend-renewal process
![Page 8: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/8.jpg)
Nelson (1995): Valve seat data
Times of valve-seat replacements in a fleet of 41 diesel enginesD
ieselE
ngin
e
Time
4140393837363534333231302928272625242322212019181716151413121110987654321
8007006005004003002001000
Event Plot for Valve Seat Replacements
Bo Lindqvist The trend-renewal process
![Page 9: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/9.jpg)
Barlow and Davis (1977): Tractor data
Bo Lindqvist The trend-renewal process
![Page 10: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/10.jpg)
Basic models for repairable systems
q q q
0 T1 T2 · · · TN τ
RP(F ): Renewal process with interarrival distribution F .
Defining property:
Times between events are i.i.d. with distribution F
NHPP(λ(·)): Nonhomogeneous Poisson process with intensityλ(t).
Defining property:
1 Number of events in (0, t] is Poisson-distributed with
expectation∫ t
0λ(u)du
2 Number of events in disjoint time intervals are stochasticallyindependent
Bo Lindqvist The trend-renewal process
![Page 11: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/11.jpg)
Point process modelling of recurrent event processes
q q q
6
φ(t|Ft−)
t
0 T1 T2 · · · TN τ
Ft− = history of events until time t.
Conditional intensity at t given history until time t,
φ(t|Ft−) = lim∆t↓0
Pr(failure in [t, t + ∆t)|Ft−)
∆t
Bo Lindqvist The trend-renewal process
![Page 12: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/12.jpg)
Special cases: The basic models
NHPP(λ(·)):
φ(t|Ft−) = λ(t)
so conditional intensity is independent of history.Interpreted as “minimal repair” at failures
RP(F ) (where F has hazard rate z(·)):
φ(t|Ft−) = z(t − TN(t−))
so conditional intensity depends (only) on time sincelast event.Interpreted as “perfect repair” at failures
Between minimal and perfect repair? Imperfect repair models.
Bo Lindqvist The trend-renewal process
![Page 13: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/13.jpg)
Trend Renewal Process – TRP(BL, Elvebakk and Heggland 2003)
Trend function: λ(t) (cumulative Λ(t) =∫ t
0 λ(u)du)
Renewal distribution: F with expected value 1 (foruniqueness)
-
-
��
���
AAAAU
@@
@@R
q q q
t0 T1 T2 T3
0 Λ(T1) Λ(T2) Λ(T3)
TRP(F , λ(·))
RP(F )
SPECIAL CASES:
NHPP: F is standard exponential distribution
RP: λ(t) is constant in t
Bo Lindqvist The trend-renewal process
![Page 14: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/14.jpg)
Motivation for TRP: Well known property of NHPP
-
-
��
���
AAAAU
@@
@@R
q q q
t0 T1 T2 T3
0 Λ(T1) Λ(T2) Λ(T3)
NHPP(λ(·))
HPP(1)
Bo Lindqvist The trend-renewal process
![Page 15: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/15.jpg)
Point process formulation of TRP:
Conditional intensity of TRP(F , λ(·)):
φ(t|Ft−) = z(Λ(t) − Λ(TN(t−)))λ(t)
where z(·) is hazard rate of F
Recall special cases:
NHPP: z(·) ≡ 1, implies φ(t|Ft−) = λ(t)
RP: λ(·) ≡ 1, implies φ(t|Ft−) = z(t − TN(t−))
Bo Lindqvist The trend-renewal process
![Page 16: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/16.jpg)
Comparison of NHPP, TRP and RP:Conditional intensities
0 0.5 1 1.5 2 2.5 3 3.50
1
2
3
4
5
6
7
8
9
10
Time
Con
ditio
nal i
nten
sity
Example: Failures observed at times 1.0 and 2.25.Conditional intensities for NHPP (solid); TRP (dashed); RP(dotted)
Bo Lindqvist The trend-renewal process
![Page 17: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/17.jpg)
Conditional intensities of TRP with DFR renewaldistribution and increasing trend function
0 0.5 1 1.5 2 2.5 3 3.50
10
20
30
40
50
60
Failures observed at times 1.0 and 2.25.Renewal: F ∼ Weibull, shape 0.5Trend: λ(t) = t2
Bo Lindqvist The trend-renewal process
![Page 18: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/18.jpg)
Interpretation of Λ(t)
-
-
��
���
AAAAU
@@
@@R
q q q
t0 T1 T2 T3
0 Λ(T1) Λ(T2) Λ(T3)
TRP(F , λ(·))
RP(F )
By using results from renewal theory:
limt→∞
N(t)
Λ(t)= 1 (a.s.)
limt→∞
E (N(t))
Λ(t)= 1
since F is assumed to have expected value 1.
Bo Lindqvist The trend-renewal process
![Page 19: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/19.jpg)
Alternative models: Cox’ modulated renewal model
Cox (1972):
φ(t|Ft−) = z(t − TN(t−))λ(t)
Compare to TRP:
φ(t|Ft−) = z(Λ(t) − Λ(TN(t−)))λ(t)
Bo Lindqvist The trend-renewal process
![Page 20: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/20.jpg)
Alternative models: The model by Lawless and Thiagarajah
Lawless and Thiagarajah (1996):
φ(t|Ft−) = eθ′h(t)
where h(t) is a vector of observable functions, depending on t andpossibly also on the history Ft−, and θ is a vector of parameters.
This model allows the use of various dynamic covariates, such as
number of previous failures of the system, N(t−)
time since last event, t − TN(t−)
Bo Lindqvist The trend-renewal process
![Page 21: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/21.jpg)
Alternative models: The model by Pena and Hollander
Pena and Hollander (2004):
φ(t|Ft−, a) = a λ0(E(t)) ρ(N(t−); α) g(x(t), β)
where
a is a positive unobserved random variable (frailty)
λ0(·) is a baseline hazard rate function
E(t) is “effective age” process, a predictable process modelingimpact of performed interventions after each event. Possiblechoices
Minimal intervention on repair: E(t) = tPerfect intervention on repair: E(t) = t − TN(t−)
Imperfect repair: E(t) = time since last perfect repairAge reduction models: E(t) is reduced by a certain factor ateach repair
ρ(·; ·) can for example be on geometric form, ρ(k ; α) = αk
g(x(t), β) models impact of covariates, for example onCox-form
Bo Lindqvist The trend-renewal process
![Page 22: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/22.jpg)
CONDITIONAL ROCOF BY MINIMAL REPAIR (NHPP) ANDPERFECT REPAIR (RENEWAL PROCESS)
167
![Page 23: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/23.jpg)
SIMPLE EXAMPLE WITH THREE SYSTEMS
Sys. 1: � � �
0 S11 = 5 S21 = 12 S31 = 17 τ1 = 20
Sys. 2: � �
0 S12 = 9 S22 = 23 τ2 = 30
Sys. 3: �
0 S13 = 4 τ3 = 10
�Proj: � � � � � �
4 5 9 12 17 230 t
Y(t):Y (t) = 3 Y (t) = 2 Y (t) = 1
168
![Page 24: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/24.jpg)
COMPUTATIONS FOR THE NELSON-AALEN ESTIMATOR
t 1/Y (t) 1/Y (t)2 W (t) V arW (t) SDW (t)
4 1/3 1/9 1/3 1/9 0.33335 1/3 1/9 2/3 2/9 0.47149 1/3 1/9 1 1/3 0.577412 1/2 1/4 3/2 7/12 0.763817 1/2 1/4 2 5/6 0.912923 1 1 3 11/6 1.3540
169
![Page 25: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/25.jpg)
ESTIMATED W (t) with 95% confidence limits (Nelson-Aalen)
0 5 10 15 20 250
1
2
3
4
5
6
t
W(t
)
170
![Page 26: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/26.jpg)
Simple Example With 3 Systems
174
![Page 27: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/27.jpg)
Nelson-Aalen estimator for Cumulative ROCOF W (t)
1. Order all failure times as t1 < t2 < . . . tn.
2. Let dj(ti) = # events in system j at ti.
3. Let d(ti) =∑m
j=1dj(ti) = # events in all systems at ti.
4. Let Yj(t) ={
1 if system j is under observation at time t0 otherwise
5. Let Y (t) =∑m
j=1Yj(t) = # systems under observation at time t .
Then
Under general assumptions: W (t) =∑ti≤t
d(ti)
Y (ti).
Assuming NHPP: Var W (t) =∑ti≤t
d(ti)
{Y (ti)}2
Under general assumptions (MINITAB): Var W (t) =
m∑j=1
{∑ti≤t
Yj(ti)
Y (ti)
[dj(ti) − d(ti)
Y (ti)
]}2
175
![Page 28: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/28.jpg)
Illustration of last formula for Simple NHPP Example(Compare with MINITAB Output):
Var W (4) =
{1
3
[0 − 1
3
]}2
+
{1
3
[0 − 1
3
]}2
+
{1
3
[1 − 1
3
]}2
=6
81= 0.27222
Var W (5) =
{1
3
[0 − 1
3
]+
1
3
[1 − 1
3
]}2
+
{1
3
[0 − 1
3
]+
1
3
[0 − 1
3
]}2
+
{1
3
[1 − 1
3
]+
1
3
[0 − 1
3
]}2
=6
81= 0.27222
176
![Page 29: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/29.jpg)
Var W (9) =
{1
3
[0 − 1
3
]+
1
3
[1 − 1
3
]+
1
3
[0 − 1
3
]}2
+
{1
3
[0 − 1
3
]+
1
3
[0 − 1
3
]+
1
3
[1 − 1
3
]}2
+
{1
3
[1 − 1
3
]+
1
3
[0 − 1
3
]+
1
3
[0 − 1
3
]}2
= 0
Var W (12) =
{1
3
[0 − 1
3
]+
1
3
[1 − 1
3
]+
1
3
[0 − 1
3
]+
1
2
[1 − 1
2
]}2
+
{1
3
[0 − 1
3
]+
1
3
[0 − 1
3
]+
1
3
[1 − 1
3
]+
1
2
[0 − 1
2
]}2
+
{1
3
[1 − 1
3
]+
1
3
[0 − 1
3
]+
1
3
[0 − 1
3
]}2
=1
8= 0.35362
177
![Page 30: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/30.jpg)
Simple Example With 3 Systems
Power Law NHPP Model: W (t;α, θ) = (t/θ)α
178
![Page 31: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/31.jpg)
Simple Example With 3 Systems
179
![Page 32: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/32.jpg)
RESIDUAL PROCESS: ”SIMPLE EXAMPLE”.Data points (and endpoints on axes) are transformed with the es-timated cumulative ROCOF,
W (t) = 0.0538 · t1.20
Sys. 1: � � �
0 0.3711 1.0612 1.6118 1.9589
Sys. 2: � �
0 0.7514 2.3166 3.1866
Sys. 3: �
0 0.2840 0.8527
Times between events, plus censored times at the end of eachaxis, are on the next slide anlysed by MINITAB as a set of censoredexponential variables.
180
![Page 33: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/33.jpg)
181
![Page 34: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/34.jpg)
182
![Page 35: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/35.jpg)
VALVESEAT DATA
183
![Page 36: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/36.jpg)
VALVESEAT DATA
185
![Page 37: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/37.jpg)
VALVESEAT DATA
186
![Page 38: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/38.jpg)
188
![Page 39: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/39.jpg)
189
![Page 40: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/40.jpg)
Grampus- data: Plot of (Ti, Ti+1) to investigate whether timesbetween failures can be assumed independent. The figure doesnot indicate a correlation between successive times.
190
![Page 41: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/41.jpg)
191
![Page 42: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/42.jpg)
192
![Page 43: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/43.jpg)
PROFILE LIKELIHOOD FOR BETA(”SIMPLE EXAMPLE”)β = 1.20, λ = 0.0538.
0 0.5 1 1.5 2 2.5 3 3.5 4−26
−25
−24
−23
−22
−21
−20
−19
beta
prof
ile
193
![Page 44: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/44.jpg)
CONNECTION BETWEEN LAMBDA OG BETA(”SIMPLE EXAMPLE”)β = 1.20, λ = 0.0538.
0.5 1 1.5 2 2.50
0.2
0.4
0.6
0.8
1
1.2
beta
lam
bda
194
![Page 45: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/45.jpg)
195
![Page 46: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/46.jpg)
196
![Page 47: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/47.jpg)
197
![Page 48: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/48.jpg)
198
![Page 49: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/49.jpg)
Valveseat Data
199
![Page 50: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/50.jpg)
Valveseat Data
200
![Page 51: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/51.jpg)
TTT-analysis Simple Example
Row STTT ID Scaled1 12 1 0,200002 15 1 0,250003 27 1 0,450004 34 1 0,566675 44 1 0,733336 53 1 0,883337 60 1 1,00000
Parameter Estimates
Standard 95% Normal CIParameter Estimate Error Lower UpperShape 1,25093 0,511 0,249996 2,25186Scale 0,238749 0,160 -0,0746105 0,552109
Trend Tests
MIL-Hdbk-189 Laplace’s Anderson-DarlingTest Statistic 9,59 0,12 0,24P-Value 0,697 0,906 0,977DF 12
201
![Page 52: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/52.jpg)
Scaled Failure Number
Sca
led
To
talTim
eo
nTe
st
1,00,80,60,40,20,0
1,0
0,8
0,6
0,4
0,2
0,0
Parameter, MLE
Shape Scale
1,25093 0,238749
Total Time on Test Plot for Simple ExampleSystem Column in ID
202
![Page 53: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/53.jpg)
TTT-analysis of Valve Seat Data
Parametric Growth Curve: C1
Model: Power-Law ProcessEstimation Method: Maximum Likelihood
Parameter Estimates
Standard 95% Normal CIParameter Estimate Error Lower UpperShape 1,39706 0,202 1,00184 1,79229Scale 0,0626023 0,026 0,0119179 0,113287
Trend Tests
MIL-Hdbk-189 Laplace’s Anderson-DarlingTest Statistic 68,72 2,03 3,17P-Value 0,032 0,043 0,022DF 96
203
![Page 54: Introduction to repairable systems STK4400 -- Spring 2011](https://reader033.vdocuments.net/reader033/viewer/2022060120/6292b8b15c3e53543354411a/html5/thumbnails/54.jpg)
Scaled Failure Number
Sca
led
To
talTim
eo
nTe
st
1,00,80,60,40,20,0
1,0
0,8
0,6
0,4
0,2
0,0
Parameter, MLE
Shape Scale
1,39706 0,0626023
Total Time on Test Plot for Valve Seat Data
204