TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.6, Jun
e
201
4, pp. 4505 ~ 4
5
1
1
DOI: 10.115
9
1
/telkomni
ka.
v
12i6.538
9
4505
Re
cei
v
ed
De
cem
ber 2
5
, 2013; Re
vi
sed
Febr
uary 22,
2014; Accept
ed March 6, 2
014
Improved Fuzzy Evalu
a
tion Model Analysis o
f
Nuclear
Power Plant Operational Safety Performance
Dong
xiao Ni
u*, Zong
y
un
Song, Jinpe
ng Liu
Schoo
l of Econ
omics an
d Man
agem
ent, North Chi
na Electri
c
Po
w
e
r Un
iver
sit
y
,
Bei No
ng Stree
t, Beijing 1
0
2
2
06,Ch
ina
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: niud
x@12
6.com
A
b
st
r
a
ct
T
he eva
l
u
a
tion
on n
u
cl
ear
po
w
e
r plant
oper
ation
a
l s
a
fety perfor
m
a
n
ce
h
ad gr
eat si
gnifi
cance
o
n
w
hether the n
u
cle
a
r pow
er pla
n
t oper
ated
safely or
not. Currently the
acade
mic liter
atures on n
u
cl
e
a
r
pow
er p
l
a
n
t o
p
e
ratio
nal
safet
y
perfor
m
anc
e
are
rare.
Impr
oved
F
u
zz
y
ev
alu
a
tion
mo
del
w
h
ich
intro
d
u
c
e
confid
enc
e lev
e
l h
ad
be
en
us
ed i
n
to th
e ev
a
l
uati
on
of
nuc
le
ar pow
er
pla
n
t
oper
ation
a
l s
a
f
e
ty perfor
m
a
n
c
e
.
The article b
u
il
t safety performa
n
ce i
ndic
a
tor system
a
n
d
further establi
s
hed i
m
porta
n
c
e leve
l eval
u
a
ti
o
n
matrix w
h
ich s
how
ed the in
d
e
xes rel
a
tive i
m
p
o
rtanc
e, an
d establ
ishe
d perfor
m
a
n
ce e
v
alu
a
tion
matri
x
w
h
ich repres
en
ted ind
i
cators i
m
p
a
ct on op
er
ation
a
l e
ffect. F
r
om the i
m
p
o
r
tance lev
e
l ev
alu
a
tion
matrix
the
w
e
igh a
nd c
o
nfide
n
ce
of in
dicators c
an
b
e
ga
in
ed
and
from th
e p
e
rformanc
e ev
al
uatio
n
matrix
th
e
eval
uatio
n
mat
r
ix can be
gai
ned. T
he pr
eli
m
i
nary ev
alu
a
tion res
u
lt an
d
synthetic con
f
idenc
e can b
e
obtai
ne
d by multiply
in
g eval
u
a
tion
matrix by
indica
t
o
r w
e
ig
h and co
nfid
en
ce, and then th
e final ev
alu
a
ti
on
result can b
e
a
c
hiev
ed.
Ke
y
w
ords
: nuclear power plant, safety per
forma
n
ce i
ndic
a
tor, confide
n
ce l
e
vel, fu
zz
y
eva
l
uati
on
mod
e
l
Copy
right
©
2014 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
In the pre
s
e
n
t era which
the energy is
extremely
use
d
, nucl
e
a
r
safety ha
s
been a
controversial
topic in en
ergy field. From the
Thre
e
Mile Island i
n
the USA to the Che
r
n
obyl
disa
ster in
the F
o
rm
er
Soviet Unio
n
and
the
Ja
pan'
s F
u
kushima n
u
cl
ea
r lea
k
eve
n
t, an
increa
sing
nu
mber of
peo
ple h
a
ve
real
ized
t
hat th
e
safety
perfo
rmance
of the
nu
clea
r
po
wer
station n
o
t o
n
ly rep
r
e
s
ent
the ente
r
p
r
ise’
s sc
ie
nce
and te
chn
o
l
ogy level, b
u
t also
ha
s
a
signifi
cant influen
ce in our
cou
n
ty’s polity, ec
onomy a
nd livelihood
s. Therefore, to evaluate the
operational
safety perfo
rm
ance of
nu
cle
a
r
po
wer stat
i
on i
s
n
o
t
so
much
to m
e
e
t
the ne
ed
a
s
to
be com
pelle
d
by situation.
The eval
uati
on of n
u
cl
ea
r p
r
og
ram i
n
our
county
mainly con
c
e
n
trates on
progra
m
’s
environ
ment,
eco
nomy, su
pply or te
ch
n
o
logy, but
rarely in the o
p
e
ration
al
safe
ty perform
an
ce.
To fill in the blank, the art
i
cle will establish
safety performance
indica
tor
syst
em (SPIs), and
evaluate nu
cl
ear p
r
og
ram
safety operation stat
e thro
ugh an
alyzin
g the indicators’ im
porta
n
c
e
and influe
nce
.
2. Dete
rmina
t
ion of Nucle
a
r Program Opera
t
ion SPIs
The internati
onal existin
g
nucl
ear
ope
rational
pe
rformance indi
ca
to
r system
h
a
s three
types: the SPIs develop
ed
and ap
plied b
y
Internationa
l Atom
ic Energy Agen
cy
(I
AEA), the SP
Is
publi
s
hed
by
Wo
rld A
s
so
ciation
of Nu
clea
r O
p
e
r
at
ors
(WANO)
and th
e SPIs gene
rali
ze
d
an
d
execute
d
by
Nu
clea
r Re
gu
latory
Com
m
i
ssi
on
(N
RC)
.
The SPIs developed by IAEA is a pyramid st
ructure whi
c
h takes
operational
safety
perfo
rman
ce
into con
s
ide
r
ation [1]. Th
e
SPIs i
s
g
r
a
d
ually detail
e
d
from
the to
p
to the
botto
m.
The to
p level
is op
eratio
n
a
l security att
r
ibute
s
; the
seco
nd l
e
vel i
s
gl
obal
level
pa
ramete
r; t
he
third level i
s
pra
c
tical
pa
rameter
and t
he pa
ram
e
ters in the la
st
level ca
n be
sup
e
rvised a
n
d
measured directly. The SPIs of
IAEA i
s
comprehensively and sy
st
emati
c
ally, but still short
of
pertinence and practicability.
The NRC’
s
SPIs can be
sepa
rated i
n
to thr
ee se
ctors: rea
c
tor safety perfo
rman
ce,
radiatio
n security, disa
ste
r
prevention
and p
r
ote
c
tion [2]. The
sup
e
rvisi
o
n
result ca
n
be
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4505 – 4
511
4506
combi
ned wit
h
perform
a
nce indi
ca
tor by NRC, and
the emphasis
will be put
on the part
wi
th
highe
r ri
sk in
orde
r to re
ali
z
e the b
a
lan
c
e betwe
en ri
sk and
ben
efit. Beside
s, in
orde
r to re
ali
z
e
experie
nce exchan
geme
n
t and build a
unified operat
ional pe
rformance evalu
a
tion crite
r
io
n,
WANO rel
e
a
s
ed
a repo
rt
about o
perational
safety
p
e
rform
a
n
c
e i
ndicator in
1
991 a
nd g
r
a
d
ually
es
tablis
hed a fairly c
o
mplete s
e
t of indicator s
y
s
t
em [3].
On the basi
s of the three indi
cator
system
, the art
i
cle
will
synt
hesi
z
e and i
nduce to
obtain a ne
w
SPIs as sh
own in Figure 1.
Figure 1. Nu
clear Op
eratio
nal Safe
ty Performa
nce Indicato
r Syste
m
3. Fuzzy
Sy
n
t
he
tic Ev
aluation Model
Bas
e
d on Co
nfidenc
e
Lev
e
l
3.1. Model Introduction
The fu
zzy ev
aluation
mod
e
l is on
e of
t
he mo
st fre
quently-u
se
d
evaluatio
n
method
s
applie
d in p
r
ogra
mme
r ev
aluation, the
main ide
a
of whi
c
h i
s
to build eval
u
a
tion matrix
b
y
Membe
r
ship
Functio
n
s
or Delphi m
e
th
od, and the
matrix sho
u
ld
be com
pou
nded
with in
dex
weig
ht whi
c
h
can
be
obtai
ned by
An
alytic Hie
r
a
r
chy
Process
(AHP). Finally th
e re
sult
can
be
gaine
d after
step
wise
cal
c
ulatio
n. Fu
zzy evalu
a
tion
mod
e
l lay
s
empha
si
s o
n
obje
c
tivity and
sci
entificity. In ord
e
r to re
alize
thi
s
pu
rpose, confid
e
n
ce level
w
ill
be introd
uce
d
into this m
odel
and the re
sult
’s relia
bility can be gu
ara
n
t
eed via confi
den
ce an
alysi
s
.
3.2. Model Application Pr
ocedur
e
The Figu
re 2
belo
w
sh
ows
the evaluatio
n pro
c
e
ssi
ng
pro
c
ed
ure of the model:
B
Figure 2. Model Evaluatio
n Procedu
re
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Im
proved Fu
zzy E
v
aluatio
n
Model Anal
ysis
of Nucl
ear Power Plant
… (Do
n
g
x
iao
Niu)
4507
From th
e fig
u
re
above, t
he first thing
to do i
s
to
establi
s
h th
e
importa
nce
hiera
r
ch
y
matrix P and
the perfo
rma
n
ce
evaluatio
n hierarchy
m
a
trix Q. Supp
ose th
e num
b
e
r of expe
rts
is
k, an
d tho
s
e
experts a
r
e
required to
ev
aluate th
e
im
portan
c
e
of e
v
ery indi
cato
r by voting.
Count
the votes and
the matrix P
can b
e
obtain
ed by
norm
a
li
zation p
r
o
c
e
s
sing. Fu
rther,
index weight
A
and confide
n
c
e C
can b
e
cal
c
ulate
d
from the
matrix P via
data proce
ssing. In addition,
according
to
experts’
eval
uation o
n
nu
clea
r o
per
ational
safety i
ndicators, th
e matrix Q
a
nd R
also
can b
e
obtain
ed. F
i
nally, the e
v
aluation
re
sult B ca
n b
e
gain
ed th
rough
multipl
y
ing
prelimi
nary e
v
aluation mat
r
ix
B
by synthetic co
nfiden
ce
i
b
.
3.2.1. Importance Hie
r
arc
h
y
Matrix
The im
po
rta
n
ce
level
of ea
ch
index
is not i
dent
ical. T
he i
n
d
i
cator
whi
c
h
is mo
re
important al
so has great
er
credi
bility. Therefore, t
he
confiden
ce
of each
i
n
dicator can be
determi
ned b
y
establi
s
hing
importa
nce h
i
era
r
chy ma
tri
x
to incre
a
se
the obje
c
tivity and cre
d
ibility
of the evaluation re
sults.
Acco
rdi
ng to
the indi
cato
r sy
stem in
se
ction
1,
th
e
index set can be reg
a
r
ded
a
s
{}
12
n
,,
,
Uu
u
u
=
. Suppose t
he impo
rtan
ce hie
r
a
r
chy is divided
into N level
s
, and the
importa
nce degre
e
de
cre
a
se
s from l
e
vel 1 to
le
vel N. The experts
sho
u
l
d evaluate the
importa
nce le
vel of every indicators in the set
an
d make a
sign at
the approp
ria
t
e level in Table
1 belo
w
.
Table 1. Impo
rtance Hie
r
a
r
chy Evaluatio
n by No.k exp
e
rt
Indicator
Importance level (t=1,2,…,k)
(level decreasing )
t
1
t
2
…
t
N
u
1
u
2
…
u
n
As reveal
ed
above, the vo
tes by every
expert ar
e sh
own in th
e table. Cou
n
ting
all votes
of all expe
rts,
the stati
s
tic t
able
can
be
o
b
tained.
After pro
c
e
s
sing t
he data
in th
e statisti
c tab
l
e,
the impo
rtan
ce hie
r
archy m
a
trix P in the
Figur
e 2
can
be dete
r
min
e
d
, and the
n
th
e wei
ght vect
or
A and confid
e
n
ce ve
ctor C
also
can b
e
o
b
tained.
3.2.2. Matrix
Opera
t
ional
Safety
Performance Ev
aluation Ma
tri
x
Different indi
cators h
a
ve different influen
ce
in op
eratio
nal safety performan
ce of
nucl
e
a
r
power pla
n
t
,
requi
rin
g
expert gro
up to evaluate the
impac
t of ea
ch indi
cato
r. Therefore, th
e
comm
ent set
sho
u
ld be the
first to be est
ablished, viz,
{}
V
E
xc
e
l
l
e
nt
G
o
od
,
,
P
o
or
=
,
.
Table 2. Ope
r
ational Safety Performa
nce
Evaluation
Indicator
Oper
ational Perf
ormance Hiera
r
ch
y
Excellent Good
…
Poor
u
1
u
2
…
u
n
Firstly, the e
x
perts
sho
u
l
d
make a si
gn in the Ta
ble 2 und
er
the app
rop
r
ia
te level
according to
the op
eratio
na
l safety impa
ct hiera
r
chy. Then all th
e vo
tes will
be a
g
greg
ated into
a
statistic tabl
e so that to es
t
ablish the evaluation mat
r
i
x
R
after data
normali
zatio
n
pro
c
e
ssi
ng.
3.2.3. Building of Calc
ulating Mode
Above p
r
ovid
es th
e
data
need
ed fo
r t
he eval
uatio
n
,
the follo
win
g
will
be
the
sp
ecifi
c
step
s to esta
blish the mo
d
e
l:
(1)
Determinatio
n of Confide
n
c
e Vect
or
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4505 – 4
511
4508
The T
able
1
can
provide t
he d
a
ta n
eed
ed, from
which the i
m
po
rta
n
ce
hie
r
a
r
chy matrix
12
,,
,
ii
i
m
pp
p
P
kk
k
=
æö
ç÷
èø
can b
e
obt
ained by n
o
rmalizatio
n pr
oce
s
sing. Assumin
g
the l
e
vel of
matrix take
s N, and
the
evaluation i
n
terval
[0, 1]
can
be
divided into
N
portion, th
en
the
importa
nce hi
era
r
chy evalu
a
tion interval
is
[]
(
)
1
,,
1
,
2
,
ii
Ni
Ni
ab
i
N
NN
--
+
==
éù
êú
ëû
, the interv
al
,
jj
ab
éù
ëû
can b
e
define
d
usin
g the following fo
rmul
a:
[]
1
,,
N
ij
jj
ii
i
p
ab
a
b
k
=
éù
ëû
å
(1)
In the form
ula above,
ij
p
is the vote
of indicato
r
j
u
on the interv
al
[]
,
ii
ab
. Let
()
1
2
jj
j
ab
z
=+
, then the weight vecto
r
(
)
12
,,
,
j
A
aa
a
=
can be d
e
termined by n
o
rmalizatio
n
pro
c
e
ssi
ng o
f
j
z
. If
[]
,
j
ii
ab
z
Î
, the confiden
ce of
indicato
r
j
u
can
be dete
r
mine
d a
s
(
)
1,
2
,
,
ij
j
p
cj
n
k
==
, and thus to obtain confid
ence vector
(
)
12
,,
,
j
Cc
c
c
=
.
(2)
Confid
en
ce Analysi
s
If the eval
uation m
a
tri
x
11
1
1
j
ii
j
rr
R
rr
=
éù
êú
êú
êú
êú
ëû
is kno
w
n,
the con
f
idence
ve
ctor
(
)
12
,,
,
j
Cc
c
c
=
sh
ould
be
synthe
sized
with the
mat
r
ix
R, a
nd t
he
confid
en
ce synth
e
tic
formula i
s
as
followe
d:
{}
(
)
12
1
1
ma
x
,
,
,
1
,
1
,
2
,
,
n
ii
i
i
n
i
i
j
i
j
im
n
be
q
q
q
e
q
=
=+
-
=
å
(2)
And,
{
}
(
)
(
)
mi
n
,
1
2
,
1
,
2
,
,
n
j
i
j
j
ji
j
j
ji
cr
c
r
j
qe
e
=+
-
+
=
(3)
In the fo
rmul
a (2)
and
(3),
the valu
e of
i
e
is
{}
12
mi
n
,
,
,
ii
i
n
i
eq
q
q
=
and the
valu
e of
j
e
is
{
}
mi
n
,
j
jj
i
cr
e
=
. The preli
m
inary evalu
a
tion matrix
BA
R
=
is kno
w
n,
and then th
e final
evaluation
re
sult B can
be
determi
ned
by
synthe
sizi
ng
i
b
with
B
. Lastl
y, the operational
safety
perfo
rman
ce
evaluation
hi
era
r
chy can
be d
e
termi
n
e
d
a
c
cordi
ng t
o
the
maxim
u
m me
mbe
r
ship
degree p
r
in
ciple.
4. Example Analy
s
is
A nuclear power plant has two million-ki
l
o
watt Pressuri
zed Wa
ter Reactor (PWR)
nucl
ear
po
we
r unit. The in
stalled
cap
a
city of unit one is 98
4,000
KW, and the
unit two also is
984,00
0KW.
This p
o
wer pl
ant has
devel
oped a
n
indi
cator man
age
ment syste
m
, and e
s
tabli
s
hed
a WA
NO
Be
nchm
arkin
g
Comp
re
hen
si
ve Que
r
y Pl
atform, throu
gh a
nalyzi
n
g
the o
p
e
r
atio
nal
safety performance indica
tor to im
prov
e its
safety perform
a
nce.
The articl
e above
will eval
uate
the SPIs of th
is nu
clea
r po
wer pl
ant.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Im
proved Fu
zzy E
v
aluatio
n
Model Anal
ysis
of Nucl
ear Power Plant
… (Do
n
g
x
iao
Niu)
4509
In this example, the indicator set is
{}
A1
-
B
1
2
3
,,
Uu
u
u
=
.
Suppo
se the
importan
c
e
level
N=5, and th
e comm
ent set is
{}
1
234
5
,,
,
,
V
v
vv
vv
=
(v
1
: Especi
a
lly Important, v
2
: Imp
o
rtant, v
3
:
Relatively Important, v
4
: Less
Important, v
5
: Not Imp
o
rtant). If th
e
num
ber of
e
x
perts is 9, t
hen
the importa
nce hiera
r
chy e
v
aluation stati
s
tic table i
s
a
s
Table 3:
Table 3. Impo
rtance Hie
r
a
r
chy Evaluatio
n of A
1
-B Indic
a
tors
A
1
-B
Importance Level
ζ
j
Weight
A
Confidence
C
1 2
3
4
5
[0.8,1]
[0.6,0.8]
[0.4,0.6]
[0.2,0.4]
[0,0.2]
u
1
4
4
1
0
0
0.7667
0.3366
0.4444
u
2
4
3
2
0
0
0.7444
0.3268
0.3333
u
3
5
2
2
0
0
0.7667
0.3366
0.2222
Acco
rdi
ng to the formula
(1),
j
z
can be
obtained, a
nd the weig
h
t
vector A and the
confid
en
ce vector
C can
be dete
r
mine
d by norm
a
lization processing. The resu
lts are
sho
w
n
in
Table 3.
Operational
safety perform
ance co
mme
nt set
{}
1
234
,,
,
Vv
v
v
v
=
(
v
1
: E
x
c
e
llent, v
2
: Good
,
v
3
: Fair, v
4
: Poor
)
i
s
kno
w
n, then the p
e
rform
a
n
c
e h
i
era
r
chy evaluation stati
s
tics
are
sh
own
in
Table 4.
Table 4. Perf
orma
nce Hie
r
archy Evaluation of A
1
-B Indic
a
tors
A
1
-B
Evaluation Level
Excellent Good
Fair
Poor
u
1
5
2 2 0
u
2
6
1 2 0
u
3
6
3 0 0
Evaluation m
a
trix can be o
b
tained by no
rmali
z
ation p
r
oce
s
sing
:
A1
-B
0
.
556
0.222
0.222
0
0.667
0.
11
1
0
.222
0
0.667
0.333
0
0
R
=
éù
êú
êú
êú
êú
ëû
Preliminary matrix
A1
-
B
B
can be o
b
tained u
s
in
g
the formula
BA
R
=
:
(
)
A1
-
B
0.
6296,
0
.
223
1
,
0.
1473
,
0
B
=
。
Confid
en
ce synthetic vecto
r
A1
-
B
b
can b
e
obtai
ned by usi
ng
the formula (2) and
(3):
(
)
A1
-
B
0.
39
93
,
0
.
2
3
9
1
,
0.
20
92,
0.
152
4
b
=
Final com
p
re
hen
sive evaluation re
sult
A1
-B
B
can be determ
i
ned by multiplying prelimi
nary
res
u
lt
A1
-
B
B
by con
f
idence synth
e
tic vecto
r
A1
-B
b
:
(
)
A1
-
B
0.
2514,
0.
053
4,
0.
0
308
,
0
B
=
The wei
ght, confiden
ce lev
e
l and synth
e
t
ic conf
id
en
ce
of other indicators a
r
e
sho
w
ed in
Table 5:
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 6, June 20
14: 4505 – 4
511
4510
Table 5. Rel
a
tive Calcul
ation of O-A-B Indicators
O-A Ind
e
x
Weight
Confidence
A-B Index
Weight
Conf
idence
S
y
nthetic Confid
ence and Preliminar
y
Vecto
r
A
1
0.2006
0.5556
B
1
0.3366
0.4444
β
0.3993
0.2391
0.2092
0.1524
B
2
0.3268
0.3333
B
0.6296
0.2231
0.1473
0.0000
B
3
0.3366
0.2222
A
2
0.2119
0.6667
B
4
0.3333
0.3333
β
0.3942
0.2632
0.1750
0.1676
B
5
0.3239
0.4444
B
0.6680
0.2960
0.0360
0.0000
B
6
0.3427
0.5556
A
3
0.1977
0.1111
B
7
0.3581
0.7778
β
0.3751
0.2473
0.2003
0.1773
B
8
0.2930
0.2222
B
0.6033
0.2960
0.1100
0.0000
B
9
0.3488
0.6667
A
4
0.2006
0.3333
B
10
0.3589
0.6667
β
0.3916
0.2496
0.1893
0.1695
B
11
0.3110
0.2222
B
0.6325
0.2932
0.0744
0.0000
B
12
0.3301
0.5556
A
5
0.1893
0.3333
B
13
0.3237
0.3333
β
0.4031
0.2549
0.1824
0.1597
B
14
0.3140
0.2222
B
0.6321
0.2928
0.0751
0.0000
B
15
0.3623
0.6667
Multiplying preliminary result
B
by confid
e
n
ce
β
in th
e
Table
5 can
get the eval
u
a
tion
result matrix of indicato
r O
-
A after norm
a
lizatio
n pro
c
essing:
0.
74
92
0.
15
90
0
.
0
918
0
0
.
7
577
0.
22
41
0
.
0
181
0
0.
70
38
0
.
2
277
0
.
0
685
0
0.
73
95
0.
21
85
0
.
0
420
0
0.
74
26
0.
21
75
0
.
0
399
0
OA
R
-
=
éù
êú
êú
êú
êú
êú
êú
êú
êú
ëû
Then
the pre
liminary eval
uation re
sult
O-
A
B
can
be
dete
r
mined
by
synthesi
z
in
g
weight
A
O-A
and evaluation matrix
O-
A
R
:
(
)
O-A
0.7389,
0
.
2094,
0.
051
8
,
0
B
=
If the model is the ba
sic fu
zzy evalu
a
tio
n
model, the
vector
O-
A
B
will be the final result.
Acco
rdi
ng to
the
Maxim
u
m Mem
bership Degree
pri
n
ciple, the val
ue 0.73
89 i
s
in the maxim
u
m
asse
ssm
ent level, and the
n
the nucl
ear
power p
r
oje
c
t
evaluation re
sult is supe
rio
r
. However, the
model u
s
ed
here i
s
not th
e basi
c
mo
de
l, but t
he advanced mo
del
introdu
cing
confiden
ce lev
e
l.
So,
the confi
den
ce synth
e
tic
ve
ctors sho
u
ld
be
o
b
tained
u
s
in
g the fo
rmul
a (2
)
and
(3
) a
s
follows
:
(
)
O-
A
0.
326
8
,
0.
2
4
1
2
,
0
.
214
9,
0
.
21
26
b
=
The confid
en
ce
ve
ctor
O-
A
b
sh
ows that th
e
confid
en
ce le
ve
l of excelle
nt is
0.326
8,
so
the evaluatio
n re
sult
sho
u
l
d
be
obtain
e
d
by synth
e
si
zing
b
with pre
liminary re
sul
t
B
, not usin
g
B
dire
ctly.
()
O-A
0.
2415,
0.
0505,
0.
0114
,
0
B
=
In accord
an
ce with
Maxi
m
u
m Mem
bershi
p
Degree
princi
ple, it
can be
see
n
that th
e
maximum va
lue 0.2
415
i
s
in
an
excellent level.
Then, th
e o
peratio
nal
sa
fety perform
ance
evaluation re
sult of the nu
clea
r po
we
r p
l
ant is su
peri
o
r.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Im
proved Fu
zzy E
v
aluatio
n
Model Anal
ysis
of Nucl
ear Power Plant
… (Do
n
g
x
iao
Niu)
4511
5. Conclusio
n
Thro
ugh the
establi
s
hm
ent and
cal
c
ulatio
n of evaluation
model, the
followin
g
con
c
lu
sio
n
s
can be drawn:
(1) Th
e index
weight
s of the indicator A
1
,
A
2
,
A
3
,
A
4
and A
5
are a
v
erage, rangi
ng from
the value 0.1
8
to 0.22, whi
c
h mea
n
s tha
t
the
indicators ha
s an average influen
ce
on operation
a
l
safety pe
rformance. Th
e
weig
ht of
Com
bustion Rel
i
ability
i
s
rela
tively large
r
t
han
othe
r fo
ur
indicators, so
its importan
c
e comp
ared to other
four indicators
is
more outs
tanding.
(2) F
r
om the evaluation m
a
trix
O-A
R
, it can be seen that th
e maximum value of the re
sult
vector of
Combustion Rel
i
ability
indi
cator i
s
0.7492.
Accordi
ng t
o
the
M
a
x
i
mu
m M
e
mb
e
r
sh
i
p
Deg
r
ee
pri
n
ciple, the eva
l
uation of
Com
bustion Reliability
i
s
e
x
cellent. Sim
ilarly, it can
be
observed that
the other fou
r
indicators’ e
v
aluation re
sults are all ex
celle
nt.
(3) From the
preli
m
ina
r
y
evaluation v
e
ctor
B
, it ca
n
be
dire
ctly
observed
tha
t
the
prelimi
nary
result i
s
ex
ce
llent. Ho
wev
e
r, from
the
confid
en
ce
analysi
s
it
can be
seen
the
confid
en
ce le
vel of excellent is 0.3268, the conf
ide
n
ce level of good is 0.241
2, the confiden
ce
level of Fair is 0.219
4 and
the confid
ence level of
Poor is 0.212
6. After confide
n
ce synthe
sizin
g
,
the final resul
t
of operation
a
l safety perf
o
rma
n
ce eval
uation is
sup
e
rio
r
.
The evaluatio
n result in the article can h
e
lp nu
clea
r o
perato
r
s and
regul
atory aut
horitie
s
kno
w
cl
early
about the current ope
rat
i
onal situ
at
ion of our nu
clear po
we
r p
l
ant, so that to
prom
ote the
improveme
n
t in both
nucl
ear po
wer e
n
terpri
se man
agem
ent mod
e
and
techn
o
logi
cal
level, to further enh
an
ce th
e operational
safety of nucl
ear po
we
r pla
n
t.
Referen
ces
[1]
Liu Yu, W
u
W
angzh
o
n
g
. Nucle
a
r Po
w
e
r
Plant O
perati
ona
l Safet
y
P
e
rformanc
e Indicator S
y
ste
m
Journ
a
l of T
s
in
ghu
a Univ
ersit
y
(Science a
n
d
T
e
chnol
ogy)
, 200
6; 46(3):4
2
1
-42
4
[2]
NEI. Regulatory
Assessment
Performanc
e
Indicat
o
r Gui
deli
ne.
W
a
s
h
i
ngton
DC:N
u
c
l
ear E
ner
g
y
Institute,200
1
[3]
Z
hu Jiz
hou, S
han
Jia
n
q
i
an
g. Nucl
ear
Po
w
e
r Pl
ant Saf
e
ty .Be
iji
ng
:
C
h
i
na E
l
ectric P
o
w
e
r Pres
s
,
201
0.72-
73
Evaluation Warning : The document was created with Spire.PDF for Python.