|
RT&A
2013,
# 1(28) Vol.8
B. Chandrasekar, T.A. Sajesh
RELIABILITY MEASURES OF SYSTEMS
WITH LOCATION-SCALE ACBVE COMPONENTS
Block
and Basu (1974) proposed an absolutely continuous
bivariate exponential distribution whose marginals are
weighted average of exponentials. Chandrasekar and
Sajesh (2010) considered location, scale and
location-scale families arising out of absolutely
continuous bivariate exponential (ACBVE) distribution
with equal marginals and derived the minimum risk
equivariant estimators of location, scale and
location-scale parameters. In this paper we consider a
two-component system when failure times follow
location-scale ACBVE distribution with equal marginals.
We obtain the reliability performance measures of
two-component parallel, series and standby systems. Also
we provide the UMVUE, the MLE and the MREE of these
reliability performance measures.
G. F. Kovalev, M. A. Rychkov
WIND-HYDRO POWER SYSTEM AS AN
EXAMPLE OF DIVERSIFICATION OF DISTRIBUTED GENERATION
The
paper considers using renewable wind energy for
electricity generating. The system is characterized by
high reliability and ecological purity. The authors
briefly present the main methodological principles of
choosing the parameters of the considered wind-hydro
power system.
Mustafa Kamal, Shazia Zarrin, Arif-Ul-Islam
STEP STRESS ACCELERATED LIFE
TESTING PLAN FOR TWO PARAMETER PARETO DISTRIBUTION
In
Accelerated life testing if the accelerated test stress
level is not high enough then many of the test items
will not fail during the available time and one has to
be prepared to handle a lot of censured data. To avoid
such type of problems, a better way is step-stress ALT.
In Step-stress ALT all test items are first tested at a
specified constant stress for a specified period of time
and then Items which are not failed will be tested at
next higher level of stress for another specified time
and so on until all items have failed or the test stops
for other reasons. In this paper simple step stress
pattern of ALT assuming that the lifetime of a product
at any constant level of stress follow a two parameter
Pareto distribution is considered. The maximum
likelihood and asymptotic confidence interval estimate
of the parameters are obtained. Optimal step stress ALT
plan is proposed by minimizing the asymptotic variance
of the MLE of the 100 Pth percentile of the lifetime
distribution at normal stress condition. A simulation
study is also performed to analyse the performance of
parameter estimates.
Igor Ushakov
OPTIMAL REDUNDANCY IN SYSTEMS WITH
MULTI-LEVEL UNITS
Method
of Universal Generating Function (UGF) was introduced in
many papers. Here we give an example how method of UGF
can be implemented to solution problems of optimal
redundancy for systems consisting of multi-level units.
V. M. Chacko, M. Manoharan
MEAN RESIDUAL LIFE CRITERIA OF
FIRST PASSAGE TIME OF SEMI-MARKOV PROCESS BASED ON TOTAL
TIME ON TEST TRANSFORMS
Mean
residual life criteria of first passage time of
semi-Markov process is considered. Properties of
transition probability functions when using scaled Total
Time on Test (TTT) transform for some criteria of mean
residual life are discussed. Application to Multistate
reliability system is also addressed.
Smagin V.A.
COMPLEX DELTA – FUNCTION
Smagin
V.A. the brief review of a history of introduction of
delta - function on a complex plane. The proof of the
mathematical form of complex delta - function is given.
The example of application of complex delta - function
for a presence of stationary value alternation casual
process of accumulation with the information income and
the charge is given.
Mykhailo D. Katsman, Viktor K. Myronenko, Nikolaj I.
Adamenko,
PROBABILISTIC MODEL OF ECOLOGICAL
CONSEQUENCES OF RAILROAD ACCIDENTS
The
paper discusses the processes of inertial reacting and
self-regulation of the environment impacted by hazards
of railway accidents involving dangerous goods and the
queuing system Markovian model is proposed to determine
the probable consequences of such accidents development.
K. Muralidharan, Arti Khabia
INLIER PRONESS IN NORMAL
DISTRIBUTION
Inliers
in a data set are subset of observations not necessarily
all zeroes, which appears to be inconsistent with the
remaining data set. They are either the resultant of
instantaneous or early failures usually encountered in
life testing, financial, clinical trial and many other
studies. We study the estimation of inliers in Normal
distribution. The masking effect problem for correctly
identifying the inliers is also discussed. An
illustration and a real life example is presented with
detailed discussions.
K. Balaji Rao, M.B. Anoop, Nagesh R. Iyer
Application of Chebyshev- and
MarKov-type inequalities in structural engineering
This paper aims at
bringing out the usefulness of Chebyshev- and Markov-
type inequalities in structural engineering design
decision making. By examining whether the bounds
arising from Chebyshev - type inequality (associated
with these are weak upper bound probabilities) enclose
the respective experimental values for deflections of
six ferrocement I-beams and web shear fatigue life of a
steel plate girder it is inferred that the bounds and
the associated probabilities estimated are realistic and
hence can be used in structural engineering design
decision making. The paper also presents some recent
developments in application of Markov type inequalities
(which are due to Steliga and Szynal (2010)) for
estimation of bounds on probability of an event sought.
The importance of such bounds in structural engineering
applications is brought out. It is shown from the
results of Monte Carlo simulation that the bounds on
probability of an event, estimated using the method
presented by Steliga and Szynal, are sharp. One of the
important advantages of the bounds presented by Steliga
and Szynal (2010) is that the original (hidden/internal)
random variable need not have well defined moments.
Possible engineering applications are also pointed out.
|