TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol. 12, No. 10, Octobe
r 20
14, pp. 7082
~ 709
1
DOI: 10.115
9
1
/telkomni
ka.
v
12i8.533
4
7082
Re
cei
v
ed
De
cem
ber 1
0
, 2013; Re
vi
sed
May 18, 20
14
; Accepte
d
Ju
ne 6, 2014
Assessment of Intelligent Substation Based on an
Improved Fuzzy Sets Method
Yuancha
o
Hu
1
, Lina Yao
2
, Jiangjun Ruan*
3
, Yunzh
u
An
4
, Fan Chen
5
1,3,
4,5
School of Electrical E
ngi
neer
ing, W
u
h
a
n
Univ
ersit
y
(W
HU)
NO.8. South Road of Easter
n Lake, 43
00
72, W
uhan, Hu
bei
Provinc
e
, P.R.Chin
a
2
Z
hengzho
u U
n
iversit
y
of Li
g
h
t Industr
y
No.5. Don
g
fen
g
Roa
d
, 450
00
2, Z
hengzh
ou,
Hen
an Provi
n
c
e
, P.R.China
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: ruanj
ian
g
j
un1
968
@12
6
.com
A
b
st
r
a
ct
In the co
nstru
c
tion pr
ocess
of smart grid,
an
ass
e
ss
me
nt on pr
i
m
ary
equ
ip
me
nts i
n
telli
ge
nt
transformatio
n
is an urge
nt content in the first
batch
of intelli
gent
substation p
i
lot proj
ect. Wid
e
investi
gatio
n of intelli
ge
nt sub
s
tation pi
lot pr
ojects
w
a
s conducted ser
i
ous
l
y
in several re
gio
nal p
o
w
e
r grid
s
in Ch
ina. An
mathe
m
atic
al
mode
l w
a
s estab
lishe
d to
asses
s
the substatio
n
pri
m
ary e
q
u
i
pments inte
lli
g
ent
transformatio
n
firstly; then, the Va
g
ue s
e
ts mu
lti-ob
jectiv
e
decisi
on th
eory
w
a
s appl
ied t
o
the ass
e
ss
ment
mo
de
l an
d the
consiste
ncy ins
pectio
n
an
d w
e
ight solv
in
g me
thod of Vag
ue
set theory w
a
s improve
d
in th
i
s
pap
er. F
i
nal
ly,
a practic
a
l
exa
m
p
l
e w
a
s g
i
ve
n to sh
ow
the ration
ality a
nd
accuracy
of th
e ab
ove
i
m
pro
v
e
d
meth
od. T
h
e
meth
od ca
n
provid
e pra
c
tical gu
id
anc
e to assess
primary e
q
u
i
p
m
e
n
ts intel
l
i
gen
t
transformation.
Ke
y
w
ords
:
sm
art grid, substation inte
ll
ige
n
t transfor
m
ati
on, va
gue
sets, fuz
z
y
set,
com
p
r
e
hensiv
e
consiste
ncy ins
pectio
n
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
Smart gri
d
is
the mediu
m
-and-l
ong
-term devel
opm
e
n
t dire
ction o
f
powe
r
gri
d
i
n
Chi
na.
The
key for i
t
s evaluatio
n
is p
e
rfo
r
man
c
e ta
rget, p
e
r
forma
n
ce
ch
ara
c
teri
stics,
key te
chnol
o
g
y
and fu
nctio
n
implem
entat
ion [1
-4]. Ma
ny re
sea
r
ch
institution
s
a
nd e
n
terp
ri
se
s in
China
a
r
e
actively prom
oting the co
n
s
tru
c
tion of smart gr
id a
n
d
have alrea
d
y made so
me
achi
evement
s. In
2009,
a sta
g
ed g
oal
wa
s pro
p
o
s
ed
b
y
Chin
a Stat
e Gri
d
fo
r constructin
g
strong
sm
art
grid:
makin
g
maj
o
r brea
kth
r
ou
g
h
and
wide
sp
read
use in t
e
rm
s of key tech
nolo
g
y an
d equi
pment
s. In
this ca
se, as the acqui
sition sou
r
ce an
d comma
nd
executio
n uni
t for basic o
peratin
g data
of
power g
r
id, d
i
gitalize
d
stat
ions a
nd sm
art stat
ion
s
h
a
s be
com
e
the key di
re
ction for prese
n
t
sma
r
t grid
constructio
n
to pro
g
ressiv
ely ac
compli
sh the t
r
an
sformation
an
d upg
rad
e
from
traditional
su
bstation. A
ccordin
g to
the
200
9-2020
sma
r
t g
r
id
d
e
velopme
n
t
plan
s, the
smart
transformation of substat
i
on will enter a comp
rehensive
const
r
uction period since 2012,
achi
eving an
intelligent rate of 30-50
percent
and
10 percent
for new
sub
s
tation
s and
old
sub
s
tation
s resp
ectively. State Grid h
a
s c
ond
ucte
d a lot of rese
arch
sin
c
e 2010 a
n
d
the
transfo
rmatio
n
and co
nst
r
uction proj
ect
s
for sm
art
substatio
n
initi
a
ted succe
ssively. The first
pilot p
r
oje
c
t
constructio
n
h
a
s
almo
st b
e
en fini
she
d
[5
-6] a
nd th
e
seco
nd
pilot p
r
oject
ha
s m
a
d
e
great a
c
hi
eve
m
ent as
well.
In this ca
se,
it has
be
co
m
e
the top p
r
io
rity to assessment for p
r
e
s
en
t
intelligent level of substati
on. Acco
rdin
g to
the High Voltage Equipment Intell
igent Techni
cal
Guideli
ne, attention
s
sh
oul
d be firstly pa
id to the
smart transformati
on level in different area
s. In
orde
r to a
c
hi
eve the sma
r
t transfo
rmati
on of su
bstat
i
on it sho
u
ld
be mad
e
cle
a
r the ne
ed f
o
r
sma
r
t transf
o
rmatio
n or
upgrade a
n
d
the standa
rd
of intelligence. To coo
perate
with the
se
con
dary pil
o
t proje
c
t, State Grid Co
m
pany will
arra
nge so
me experts
spe
c
iall
y to investigate
and evaluate
the pilot pro
j
ect in a wid
e
rang
e.
Ho
wever, comp
rehe
nsive ev
aluation of smart
transformation level is relat
i
vely deficient
.
Since
Ca
u a
n
d
Bueh
re
r a
d
v
ance
d
the V
ague
Set
Th
e
o
ry in
199
3, it ha
s a
c
hieve
d
ra
pid
development
as
a pr
omotion for
m
of fuzzy s
e
ts
[8
-9]. Tr
aditional multi-
objec
tive dec
i
s
i
on-
m
ak
ing
theorie
sa
re l
a
cking i
n
un
certai
n info
rmation de
scription [10-12]
, such a
s
A
nalytic Hi
era
r
chy
Process, entropy weig
ht method,
grey i
n
cid
e
n
c
e the
o
ry and Te
ch
nique for
Ord
e
r Prefe
r
en
ce
by
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Asse
ssm
ent of Intelligent Substation B
a
se
d on
an I
m
prove
d
Fuzzy Set
s
Meth
od (Yua
ncha
o Hu)
7083
Similarity to Ideal Solution
.Compa
re
d with fu
zzy sets, Vague Set Theo
ry can
mathemati
c
al
ly
rep
r
e
s
ent
an
d process m
o
re a
bun
dant
uncertain
info
rmation. T
h
e
r
efore, Va
gue
Set Theo
ry can
be
widely
ap
plied i
n
field
s
as fuzzy
con
t
roller de
sig
n
, MOMSF
D
M,
artificial
intell
igen
ce
and
son
on. Beside
s, the po
wer
se
ctor is
ra
rely in
volved in this theory.
Due to th
e in
volvement of
much un
ce
rtain
informatio
n in smart transfo
rmatio
n, Vague
Set Theory
was a
pplied to
the com
p
reh
ensive eval
u
a
tion for
sma
r
t tran
sform
a
tion of su
bstat
i
on
and th
e
con
s
i
s
ten
c
y che
c
k
and
wei
ght
solving of Va
g
ue Set T
heo
ry wa
s imp
r
ov
ed in
this pa
p
e
r
.
And the tran
sform
a
tion le
vel of seve
ral
pilot pr
oje
c
ts wa
s a
s
se
ssed with
the a
bove evalu
a
tion
model. La
stly, a pra
c
tical
example
wa
s given to
sta
t
e the cal
c
ul
ation
process an
d practi
cal
appli
c
ability of this method
in detail.
2. Cons
truc
tion of Smart
Trans
f
orma
tion Ev
aluation Model for
Substa
tions
Smart statio
n
is comp
osed
of pro
c
e
s
s l
e
vel (eq
u
ipm
ent level),
co
mpartme
n
t le
vel and
station level.
The integ
r
ate
d
intelligent e
quipme
n
t
is u
s
ed to fulfill f
unctio
n
s
su
ch as
coll
ectio
n
,
measurement
, control, protecti
on, cal
c
ulatio
n, monitor and re
al-time onlin
e analysi
s
a
nd
deci
s
io
n, whi
c
h
assu
re th
e
coll
abo
rative
intera
ct
ive o
peratio
n of th
e po
we
r g
r
id.
Intelligentialized
sub
s
tation i
s
mainly featured
by digital mea
s
urem
ent, control netwo
rk,
state visuali
z
atio
n,
function
al int
egratio
n, an
d
inform
ation i
n
tera
ction.
It basi
c
ally req
u
ire
s
total di
gital
informati
on,
comm
uni
cati
on network p
l
atform and i
n
formatio
n s
harin
g sta
n
d
a
rdi
z
ation. In
accordan
ce
with
techn
o
logi
cal
feature
s
a
n
d
req
u
ire
m
ent
s, the di
gital
measurement
and info
rmati
on inte
ra
ction
of
importa
nt electri
c
acce
sso
r
y is the key to fulfill
intelligentialized
sub
s
tation. T
herefo
r
e, in the
con
s
tru
c
tion
of the first
and
se
co
nd
sma
r
t
sub
s
tation, the i
n
tegrate
d
le
vel of intelli
gent
comp
one
nts
and a
d
van
c
e
d
appli
c
atio
n
of monito
ring
informatio
n a
r
e the
con
c
en
trated
refle
c
tion
of smart tra
n
sformation lev
e
l of pilot sub
s
tation proje
c
ts.
(a) T
r
an
sfo
r
m
e
r
intelligent mo
nitoring of
oil and ga
s
(b) Potential t
r
an
sform
e
r
intelligent co
mpone
nts
(c) GIS partia
l
discharge
monitori
ng de
vices
(d) swit
chin
g
intelligent
assembly cab
i
net
Figure 1. The
Intelligent Transfo
rmatio
n Arra
n
gem
ent of Substation
Primary Equi
pments
Acco
rdi
ng to
the a
c
compl
i
shme
nt of th
e tr
an
sformat
i
on an
d con
s
truction, Stat
e Gri
d
Comp
any su
ccessively arrange
d som
e
experts fo
r
Hi
-pot test an
d equipm
ent to investigate a
n
d
evaluate the f
i
rst an
d se
co
nd pilot proje
c
ts of
eq
uipm
ent intelligent
ializatio
n. Layered
evaluati
on
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 10, Octobe
r 2014: 708
2
– 7091
7084
deci
s
io
n pri
n
ciple
wa
s uti
lized from th
e whol
e
to the pa
rt: sele
cting
conve
r
ting station f
o
r
evaluation an
d perfo
rming
an integral evaluation fo
r e
quipme
n
t su
ch as tran
sfo
r
mer, GIS (HG
I
S)
combi
nation
unit, brea
ker
and lightnin
g
arreste
r
[13-1
4
]. And acco
rding to the re
quire
ment
s for
intelligent co
mpone
nts of
each
ele
c
tri
c
acce
ssory
in
Intelligent
Su
bstation Te
ch
nical Guid
elin
e
and
Intelligent Substation Desi
gn Specifications
, it refine
s the monitori
ng category of e
a
ch
equipm
ent su
ch a
s
the oil tempe
r
ature, bra
cki
sh
wate
r and pa
rtial d
i
scharge, and
so on.
Practi
cal re
searche
s
sho
w
ed that differen
c
e
s
am
o
ng sm
art tra
n
sformation
of different
voltage level
s
are a
bout th
e refo
rm difficulty and pr
og
ram
s
of different manufa
c
t
u
re
rs. In
spite
of
the installm
ent of intelligent components, differ
ent equipments of
the
sam
e
converting
station
coul
d all be
evaluated i
n
terms
of mea
s
ureme
n
t, co
ntrol an
d co
mmuni
cation
according to
the
techni
cal
gui
de [15]. An
e
v
aluation m
o
del for smar
t
tran
sform
a
tion level
wa
s pro
p
o
s
ed
in
this
pape
r to tran
sform a
nd ev
aluate the m
a
in high
-p
re
ssure equi
pm
ents such as transfo
rme
r
, GIS
combi
nation
unit, mutual indu
ctor, switch, lightning arreste
r
an
d ele
c
tri
c
ca
ble
T
able 1.
Inte
llectualiza
t
ion
Assessm
ents of Substa
tion Equ
i
pmen
t Ob
jects
First indicator
Second indicator
First indicator
Second indicator
Measurement
function
Digitized sample
s
Protection
PDIF, PTR
C
and
other standa
rd
model
MMXU, MMXN standard
model
Net
w
ork contr
o
l
G
M
RP multicast
protocol d
y
na
mically
allocated
Digitized sample
s
D
y
namic, stead
y-state
comprehensive
collection of data
Place installation
Power
Q
uality
Monitoring
G
M
RP multicast
protocol
d
y
namicall
y
allocated
Control
functions
CSWI, CILO stan
dard
model
Metering
function
MMTR standa
rd
model
Net
w
ork contr
o
l
Digitized sample
s
G
M
RP multicast
protocol d
y
na
mically
allocated
G
M
RP multicast
protocol
d
y
namicall
y
allocated
Emergenc
y
Ope
r
ation
Communication
function
Virtual LAN
(VLAN)
Contempo
raneo
us
function w
i
th th
e
same
period the voltag
e
selector
Pr
ior
i
ty
tr
ansmit
Condition
Monitoring
Monitoring data
digitization
IEC61588-
precision netw
o
r
k
time
Standardization of
the
measurement re
sults
G
M
RP multicast
protocol
d
y
namicall
y
allocated
Abnormal alarm
state
Wireless handheld
communication
devices
3. Promoting Vague Set T
h
eor
y
as a M
u
lti-objec
tiv
e
Decision
-making The
or
y
The big
g
e
s
t
prom
otion o
f
Vague
sets as fr
om f
u
zzy set
s
is that the de
gree
of
membe
r
ship,
non
-mem
be
rshi
p a
nd
h
e
sitation
de
g
r
ee
s
(un
c
e
r
tainty) were
con
s
id
ere
d
more
flexible and ri
che
r
with Vag
ue set
s
thant
hat with tradit
i
onal fuzzy se
ts.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Asse
ssm
ent of Intelligent Substation B
a
se
d on
an I
m
prove
d
Fuzzy Set
s
Meth
od (Yua
ncha
o Hu)
7085
3.1. Vague Value Judgme
n
t Matrix
Acco
rdi
ng to
the definition of Vague sets, the dom
ain
X
(any el
ement of the
domain
r
e
pr
es
e
n
t
ed
b
y
x
) on th
e Vague
set
V
use
real m
e
mbershi
p
fun
c
tion
t
v
and f
a
ke m
e
mbe
r
ship
function
f
v
c
h
arac
teriz
a
tion:
t
v
:
X
,
→
[0,1]
f
v
:
X
→
[0,1
]
(1)
Assu
ming
th
at
t
v
(
x
) an
d
f
v
(
x
) respe
c
tively re
pre
s
e
n
ts the
mem
bership
de
gree lo
we
r
boun
d derive
d
from the evidence of for and ag
ain
s
t
x
,and
t
v
(
x
)+
f
v
(
x
)
≤
1, the membe
r
ship
of
element
x
i
n
Vague set
V
will be defined by the
subi
nterval [
t
v
(
x
)
,
1-
f
v
(
x
)]
on [0,
1]. Ac
cording to
this definition
,
the nece
ssi
ty of supporti
ng
x
∈
x
is
c
h
ar
ac
te
r
i
z
ed w
i
th
t
V
(
x
); the possi
bility o
f
sup
portin
g
x
∈
x
(1 -
f
v
(
x
))
is ch
aract
e
rized with
(1-
f
v
(
x
)
)
; And
(1 -
t
v
(
x
) -
f
v
(
x
)) d
e
scri
be
mathemati
c
al
ly the uncerta
inty of
x
.
Setting
X
= {
x
1
,
x
2
,... ,
x
n
} as a
set of
attributes, the
0.1
-
0.9
scale m
e
thod [16] i
s
u
s
ed fo
r
pairwise
com
pari
s
on
of e
a
c
h
attribute,t
hus con
s
tituting the
jud
g
m
ent matrix
V=
[
v
ij
]
m
×
n
based on
Vague val
u
e
s
, where
v
ij
=[
t
ij
,
1
-f
ij
] for Vague
value
s
,
t
ij
and
f
ij
re
spectively rep
r
ese
n
ts
de
cisi
on
make
rs’ preferen
ce
deg
ree of
x
i
a
nd
x
j ,
(
1-
t
ij
-f
ij
)
rep
r
e
s
ent
s d
e
ci
sion ma
kers’
un
certai
nty.
Acco
rdi
ng to the nature of
the judgi
ng
matrix stru
ctu
r
e, it can get
t
ij
∈
[0,1],
f
ij
∈
[0
,1] and
t
ij
+ f
ij
≤
1,
diago
nal ele
m
ents
t
ii
=f
jj
=0
.5,non di
ago
n
a
l elem
ents
meet the
com
p
lementa
r
y :
t
ij
=f
ji .
In the
actual
asse
ssm
ent,
different
exp
e
rts’
judg
me
nt on
a
parti
cula
r attribut
e
can often be subj
ective
and
variou
s, so y
ou ca
n take the mea
n
valu
e of different
deci
s
io
n ma
kers’
asse
ssm
ent value a
s
the
Vague value
s
of the judgment matrix.
3.2.
Fuzz
y
Approximatio
n to Judgme
n
t Matrix of
Vague Value
s
Vague
set
s
gene
rally o
n
l
y
con
s
ide
r
s t
he for an
d a
gain
s
t memb
ership i
n
the
multi-
obje
c
tive eva
l
uation, la
cki
ng of
un
ce
rtain info
rmati
on. In fa
ct, a
s
a
kind
of
deci
s
io
n ma
kers’
attitude, un
certain i
n
form
ation
shoul
d
be fully d
e
tailed mi
ning.
In additio
n
, Vague
valu
e a
s
judgme
n
t ma
trix element
s increa
se
s t
he computati
onal
compl
e
xity. Conside
r
ing th
ese
two
asp
e
ct
s, tran
sformi
ng Va
g
ue into fu
zzy sets is fe
asi
b
le to de
al
wi
th the Vagu
e
value jud
g
m
ent
matrix.
Setting
Vagu
e
set as
V
={[
t
v
(
x
),1
-
f
v
(
x
)]
x
∈
X
}, fuzz
y
s
e
t as
F
={[
x,
F
(
x
)]
x
∈
X
}, mappin
g
R
:
V
→
F
meet,
v
vv
tx
Fx
tx
+
f
x
(2)
Thus F i
s
the fuzzy ap
proxi
m
ation to Vague set V.
In the same
way, prom
otion to the fuzzy
approximati
on of Vague
value judgm
ent matrix,
it can get the fuzzy ap
proxi
m
ati
on to Vague value ju
d
g
ment matrix
Q=
[
q
ij
]
n
×
n
,
wh
ere:
ij
ij
ij
ij
t
q
tf
,
i
,
j=1,2,
…,
n
(3)
3.3. Consis
tent Check a
nd Modifica
tion of Fuzz
y
Judgmen
t M
a
trixe
s
Literatu
re (16
)
a
nd (17
)
co
nsid
ere
d
the
trans
itivity of
judgme
n
t o
r
d
e
r i
n
the
con
s
iste
ncy
check of fuzzy judgment m
a
trix, avoiding the
appea
rance of
sequent
ial logic contradi
ction such
as
x
i
≺
x
j
≺
x
k
≺
x
iin
in the at
tributes
ci
rcu
l
ation ch
ain,
but this m
e
thod ign
o
re
d the deviati
on
accepta
b
ility [16, 17].Tho
ugh literature
(18
)
set a
deviation thresh
old, it co
nsid
ere
r
little of
transitivity of judgment order. Actually, fuzzy j
udge
ment matrix sho
u
ld sati
sfy both consi
s
tent
and l
ogi
cal transitivity at the
same
tim
e
. The
r
efo
r
e,
a comp
re
he
nsive
i
n
spe
c
tion stand
ard
and
corre
c
tion m
e
thod
wa
s p
r
opo
se
d in
pre
s
ent
stud
y to make t
he con
s
iste
n
c
y ch
eck m
o
re
rea
s
on
able.
(1) Verification of compatibility indicators
Fuzz
y judgement matr
ix
Q
meet
s the
compl
e
me
nta
r
y co
ndition,
∀
k
∈
(1
,
n
)
,
q
ij
=0.
5
+
q
ik
-
q
jk
so mat
r
ix Q i
s
completely cons
i
s
te
nt matrix. Fo
r vari
ou
s d
e
cision
informat
ion, compl
e
te
ly
con
s
i
s
tent m
a
trix is ha
rd
to exi
s
t in
practi
cal evaluation. Thu
s
, corre
c
tion
is
ne
eded
for
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 10, Octobe
r 2014: 708
2
– 7091
7086
approximatio
n. Gen
e
rally, co
nsi
s
ten
c
y
index
C
I
i
s
used
to m
easure
the
deviation
de
gree
(gen
erally take less than 0.
1).
Q
*
=
(
q
ij
)
n×n
is
tak
e
n as
the c
h
arac
teris
t
ic
matrix of
Q,
whe
r
e:
nn
ij
ik
jk
k1
k1
q=
0
.
5
+
q
q
n
(4)
Compatibility indicators
C
I
are
expresse
d by fu
zzy ju
dgment
matri
x
and it
s
ch
aracteri
sti
c
matrix Q * de
viations a
s
:
nn
*2
*
2
Ii
j
i
j
i1j
1
CQ
Q
n
q
q
n
(5)
In general practice, whe
n
C
I
is
les
s
than 0.15,
Q
can
be rega
rd
ed
as con
s
iste
nt matrix.
(2) Ve
rificatio
n
of transitivity index
Rea
c
h
able
matrix T in grap
h theo
ry is cited
to
verify the logic
trans
i
tivity of fuz
z
y
judgme
n
t ma
trix
Q
[19]. Define matrix
Q
T
as a
c
com
pany re
acha
ble matrix of
fuzzy ju
dgem
ent
matrix
Q
, wh
en
q
ij
> 0.5,
Q
T
take
s "tru
e", and expressed a
s
"1"
.
Instead,
Q
T
to "false", and
expre
s
sed
as "0”. Su
ccessively solving
n-o
r
de
r
of th
e adj
oint rea
c
ha
ble m
a
trix
, get reachab
le
matrix
T
whi
c
h can ju
dge t
he co
nsi
s
ten
cy of
Q.
T
=
Q
T
||
Q
T
2
||
⋯
||
Q
T
n
(6)
Whe
r
e "| |" mean
s "or" op
e
r
ation in Bo
olean
op
eration
.
If all the main diago
nal el
ements
of re
achable
matrix T
a
r
e "zero"
valu
e,
Q
i
s
co
nsi
dere
d
to
me
et the te
st of
logi
c
con
s
i
s
tent
transitivity.
(3)
Comp
re
h
ensive in
sp
ection and corre
c
tion
Whe
n
Q
satisfies both compatibility and transitivity
at the
same time, its
consi
s
tency is
accepta
b
le; o
t
herwi
se, th
e
colum
n
s who
s
e
deviation
i
s
too
large
ne
ed to b
e
co
rrected.
Cal
c
ul
ate
the sum of de
viation value in each ro
w:
n
*
ii
j
i
j
j1
h q
q
,
i
=
1,2,
…,
n
(7)
Select the ele
m
ents of maximum deviation in
h
-th ro
w and take a k-th co
rre
ction,
where
α
(0
≤
α
≤
1
)
is th
e prop
ortio
n
of original m
a
trix informatio
n,
q
hj
(
k
+1)
=
(1
-
α
)
q
hj
(
k
)
+
α
q
hj
*
(
k
)
,
j=1,2,…,n
(8)
Similarly, the element of the maximum d
e
viation in l-th colum
n
is correcte
d:
q
il
(
k
+1)
=(1
-
α
)
q
il
(
k
)
+
α
q
il
*
(
k
)
,
i=1,
2,…,n
(9)
Q
matrix has
a consi
s
tency
check after
each
correction, until it sati
sfies the
compatibility
and tran
sitivity at the same
time.
3.4. Solv
e Evaluation Ind
ex Weigh
t
s
The
Q
m
a
trix which m
eets a
co
n
s
iste
nc
y che
ck have
a layering of
weig
hts
solutio
n
.Ch
a
racteri
s
tic ve
ct
or meth
od i
s
use
d
to calcu
l
ate the maxi
mum charact
e
risti
c
value
λ
ma
x
of
Q
matrix
and the ch
aracteri
stic ve
ctor of char
acteristic valu
e. Characte
rist
ic vector
wa
s
norm
a
lized toget
n
inde
x weight ve
ctors of k-th
layer as
w
j
(k)
=(
w
1
(k)
,
w
2
(k)
,…
w
n
(k)
)
,
w
her
e
j
=
1,2,
…,
n
. Th
en the
j
-th we
ight evaluatio
n index for the combi
natio
n of total target is cal
c
ulate
d
.
w
j
=
w
j
(
k
)
·
w
j
(
k
-1)
·…·
w
j
(1)
(10)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Asse
ssm
ent of Intelligent Substation B
a
se
d on
an I
m
prove
d
Fuzzy Set
s
Meth
od (Yua
ncha
o Hu)
7087
3.5. Solv
e th
e Compre
he
nsiv
e
Ev
alua
tion Re
sults
For a
n
evalu
a
tion with
m
obje
c
ts a
nd
n
indexes,
con
s
tru
c
ted a
we
ighted de
ci
sio
n
matrix
Y=
(
y
ij
)
m
×
n
,
where
y
ij
dete
r
mi
ned
by expe
rt
s a
s
se
ssed v
a
lue
or
a
c
tual
pa
ramete
r v
a
lue
s
of va
rio
u
s
indicators of
different a
ssessm
ent obj
ects multiplie
d
by
the
in
d
e
x weig
hts
(Pay attention
to
effective type paramete
r
s
and
co
st
type pa
ramete
rs' normali
zatio
n
process). Each evaluat
ion
index contra
st betwee
n
ob
jects
ca
n be
dire
ctly rep
r
ese
n
ted by
Weig
hted d
e
c
isi
on mat
r
ix. In
pra
c
tice, t
a
ke
Level
1 a
s
se
ssment in
dica
tors
z
j
a
s
t
h
e
colum
n
v
e
ct
o
r
of
de
cisio
n
mat
r
ix
Z=
(
z
ij
)
m
×
n
,
whe
r
e:
n
ij
i
j
j
i1
zy
w
(11)
Acco
rdi
ng to
the wei
ghted
deci
s
io
n mat
r
ix,
get ea
ch
appraisal obj
ect'
s final ev
aluatio
n
value
η
i
n
ii
j
j1
z
,
i
=
1,2,
…,
m
(12)
3.6. Flo
w
Ch
art of Vag
u
e
Sets Ev
aluation Method
Asse
ssing
th
e a
c
tual
obje
c
t with
Vagu
e
set evalu
a
tion meth
od
h
a
s
bee
n
sho
w
n f
r
om
se
ction 3.1 to
sectio
n 3.5, and the spe
c
i
f
ic flow ch
art
sho
w
n in Fig
u
re 2.
Figure 2. The
whole Flo
w
Cha
r
t of Vague Sets Asse
ssment Meth
od
4. Analy
s
is o
n
Practical E
xample
4.1. Practical
Example
Main p
r
ima
r
y equip
m
ent
of the pilot
p
r
oje
c
t of the
initial co
mple
ted sub
s
tatio
n
s
wa
s
taken a
s
the
apprai
sal ta
rget. It mainly includi
ng 1
10kV Intellig
ent Substatio
n
transfo
rmat
ion
inclu
d
ing tra
n
sformers i
n
three
staton
s an
d
GIS
HV combin
e
d
ele
c
tri
c
al
equipm
ent in
two
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 10, Octobe
r 2014: 708
2
– 7091
7088
station
s
fo
r
evaluation
o
b
ject. T
he in
telligen
ce
ref
o
rm l
e
vel of
each p
r
ima
r
y equi
pment
wa
s
asse
ssed a
c
cording
to E
v
aluation
system in
Tab
l
e. Firstly, u
s
e Va
gue
sets a
s
sessm
ent
method
s to solve the weig
ht vector for
each le
vel indicato
r of the first and se
cond indi
cato
rs.
Take
the
seconda
ry indi
ca
tor a
s
an
exa
m
ple, Va
g
ue value
j
udgm
ent
matrix wa
s con
s
tituted by
the mean val
ue of the se
cond indi
cato
r from different
experts.
0.
50
,
0
.50
0
.
3
5
,
0.
50
0.
41
,
0
.
6
8
0
.
3
8
,
0.
73
0
.
4
8
,
0
.7
3
0
.
5
9
,
0
.
5
0
,
0
.
6
5
0
.
50,
0.
50
0.
53,
0.
68
0.
32
,
0
.59
0
.
3
2
,
0.
47
0.
50
,
0
.
5
0
0.
27
,
0
.62
0
.
2
7
,
0.
52
0.
32
,
0
.
6
2
0.
27
,
0
.52
0
.
2
2
,
0.
52
0.
32
,
0
.
5
2
0.
33
,
0
.41
0
.1
2
,
0.
32
0
.
5
5
,
0.6
5
V=
0.
6
7
0.
48
,
0
.
7
3
0
.4
8,
0
.
7
8
0.
68
,
0
.
8
8
0.
38
,
0
.
6
8
0
.4
8,
0
.
6
8
0.
35
,
0
.
4
5
0.
50
,
0
.
5
0
0
.
5
0
,
0.
67
0.6
5
,
0.
47
0.
33
,
0
.
5
0
0
.
5
0
,
0.
50
0.
54
,
0
.
7
6
0
.
5
3
,
0
.3
5
0
.
2
4
,
0.
46
0.
50
,
0
.
5
0
Usi
ng Eq
uati
on (2),
se
eki
ng fu
zzy
app
roximation
m
a
trix
Q
of V
a
gue valu
e ju
dgment
matrix, According to Equation
(4) and
(5), verify
the compatibility indicato
rs of the matrix,
and
C
I
wa
s equ
aled
to 0.0458,
whi
c
h me
et the co
mpat
ibi
lity require
m
ents deviatio
n
s. Verifing t
he
transfe
r of
Q
acco
rdin
g to
Equation
(6
), not all
elem
ent
s
of the m
a
in dia
gon
al
are
ze
ro. T
h
us,
corre
c
t the most biased lin
e 6, column 6
acco
rdi
ng to
α
=
0
.4. After t
w
o times
of correc
tion, it g
o
t
the final fuzzy
judgment ma
trix:
0.50
0
.
41
0.56
0
.
58
0.64
0
.
65
0.58
0
.
50
0.62
0.64
0.
69
0.78
0.44
0
.
38
0.50
0.54
0.
60
0.52
Q=
0.42
0
.
36
0.4
6
0.50
0.
60
0.57
0.36
0
.
31
0.40
0
.
40
0.50
0.60
0.35
0
.
22
0.48
0.43
0.4
0
0.50
After c
o
rrec
tion, trans
i
tivity i
ndicato
rs
meet the re
q
u
irem
ents,
C
I
wa
s equ
aled
to 0.0228. T
hus,
comp
atibility bias
ha
s al
so
been
co
rrect
ed. Succe
ssi
vely solving t
he two i
ndi
ca
tors Va
gue v
a
lue
judgme
n
t mat
r
ix and
cond
u
c
ting fu
zzy a
pproxim
at
ion
and
co
nsi
s
te
ncy of j
udgm
ents. Th
e
re
sults
are sho
w
n in
Table 2
Table 2. Vag
ue Sets Matri
x
of Seconda
ry
Indexes an
d Con
s
i
s
ten
cy Check Results
First indicator
Consistent
checks
Successi
ve cor
r
e
ction matr
ix
Q
1st
2nd
3rd
Measure
indicators
C
I
0.1125
0.0781
0.0566
-
T
N N
Y
-
Control
indicators
C
I
0.0696
0.0475
-
-
T
N Y
-
-
Monitor
indicators
C
I
0.1111
0.0815
0.0652
0.0489
T
N N
N
Y
Protection
indicators
C
I
0.1437
0.0991
-
-
T
N Y
-
-
Estimate
indicators
C
I
0.0531
-
-
-
T
Y
-
-
-
Communicate
indicators
C
I
0.0926
0.0653
0.0495
-
T
N N
Y
-
With the
co
rrection fu
zzy judgme
n
t mat
r
ix,using
the
eigenve
c
tor
method to
o
b
tain on
e ind
e
x
weig
ht vector
w
(1)
=(0.1
866
,0.2133,0.16
6
1
,0.
1614,0.1
4
16,0.131
0). Similarly, accordin
g to fuzzy
judgme
n
t matrix of each two indicators
to solve corre
s
po
ndin
g
wei
ght vector
w
j
(2)
, acco
rding
to
Equation (10) obtained two
overall index
weight vecto
r
sho
w
n in T
a
ble 3.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Asse
ssm
ent of Intelligent Substation B
a
se
d on
an I
m
prove
d
Fuzzy Set
s
Meth
od (Yua
ncha
o Hu)
7089
Table 3. Co
m
p
reh
e
n
s
ive Weight Vect
o
r
s
Solution of Seco
nda
ry Indexes
First indicator
The t
w
o ove
r
all index
weights
w
j
Measure indicators
(0.0466, 0.
0307,
0.0267, 0.04
29,
0.0397)
Control indicator
s
(0.0510, 0.
0538,
0.0414, 0.02
92,
0.0378)
Monitor indicator
s
(0.0648, 0.
0446,
0.0567)
Protection indicators
(0.0274, 0.
0333,
0.0285, 0.02
11,
0.0206)
Estimate indicato
rs
(0.0545, 0.
0449,
0.0421)
Communicate indicators
(0.0303, 0.
0377,
0.0244, 0.01
34,
0.0251)
Select the pri
m
ary equi
pm
ent to be evaluated of
the substatio
n
,the se
con
dary in
dicato
rs
to evaluate each equi
pm
ent intelligen
t modificatio
n
rating (T
a
b
le 1). Amo
ng them, for and
again
s
t a target to com
p
leted level score
s were
t
v
(
x
) a
nd
f
v
(
x
)(Score
t
v
(
x
)
+f
v
(
x
)
≦
1), then this
index revie
w
score
s
con
s
ti
tute a Vague
value
:
[
t
v
(
x
),
1-
f
v
(
x
)], acc
o
rding to Equation (2) to tak
e
the fuzzy ap
p
r
oximation
F
v
(
x
) as the in
di
cator’
s a
c
tual
sco
re
s.
Diffe
rent de
cisi
on
-make
r
s asse
ss
the Vagu
e value of
a p
a
rticul
ar i
nde
x varies
, ta
ke the ave
r
a
ge value
of
the two f
u
zzy
approximatio
n of indicato
rs that ma
ke
up the equi
pment revie
w
sco
r
e
s
vector
F
i
(
v
). As an
example, the
revie
w
scores ve
cto
r
sol
v
ing in
t
r
an
sforme
r A, afte
r the
a
s
sessment by
exp
e
rts
according to Table 1, the review sco
r
e
s
results of tran
sform
e
r A sh
own in Ta
ble
4.
Table 4. Sco
r
e Re
sults of
Tran
sfo
r
mer
A Based on V
ague Sets
First indicator
The t
w
o indicator
s
mean score res
u
lts
v
F
Measure indicators
(0.8095, 0.
7857,
0.7021, 0.81
71,
0.6923)
Control indicator
s
(0.7701, 0.
8816,
0.7021, 0.78
57,
0.7701 )
Monitor indicator
s
(0.8252, 0.
7627,
0.8182)
Protection indicators
(0.8861, 0.
8736,
0.9239, 0.63
64,
0.8929)
Estimate indicato
rs
(0.8667, 0.
9135,
0.8590)
Communicate indicators
(0.7609, 0.
8462,
0.9184, 0.89
47,
0.5769 )
Sequentially cal
c
ulate sco
r
e
result
s
of other
p
r
ima
r
y equipm
ent a
nd co
nstitud
e
deci
s
ion
matrix
Z
accordin
g to formula (11
)
. Th
e weightin
g value
s
of each
intelligent pri
m
ary equip
m
ent
evaluation in
dexare
sho
w
n in Table 5.
Table 5. Weighted Value
s
of Evaluation
Index of Intelligent Prima
r
y Equipment
Inde
x
O
b
jects
Measure
Control
Monitor
Prot
ection Estimate
Communicate
Transform
er
A
0.1431
0.1678
0.1339
0.1115
0.1244
0.1038
Transform
er
B
0.1397
0.1665
0.1316
0.1141
0.1238
0.1133
Transform
er
C
0.1307
0.1561
0.1217
0.1086
0.1167
0.1086
GIS equipment
A
0.1337
0.1586
0.1251
0.1098
0.1158
0.1076
GIS equipment
B
0.1135
0.1386
0.1055
0.0981
0.1107
0.0995
By the formula (1
2) the
prima
r
y eq
uipment
intel
ligent tran
sfo
r
med
com
p
rehen
sive
evaluation
re
sults a
r
e
sh
o
w
n i
n
T
able
6
.
At the
same
time, Ta
ble
6 sho
w
s the
cal
c
ulate
d
val
ues
usin
g AHP [10] and TO
PSIS method [7]. Becaus
e differe
nt asse
ssm
ent method
s use
d
in
cal
c
ulatio
n steps are different, the final asse
ssm
ent value wi
ll be differe
nt, but different
as
se
ssm
ent
’s
sequ
en
ce re
sult
s a
r
e
con
s
ist
e
nt
.
Table 6. Asse
ssment Re
sul
t
of Intelligent Primary Equi
pment
Assessment
O
b
jects
Vague sets Assessment Act
AHP Assessmen
t Act
TOPSIS Assessment Act
η
i
Sequence
η
i
Sequence
η
i
Sequence
Transform
er
A
0.7845
2
0.6707
2
0.7061
2
Transform
er
B
0.7890
1
0.6830
1
0.7180
1
Transform
er
C
0.7424
4
0.6276
4
0.6682
4
GIS equipment
A
0.7506
3
0.6682
3
0.6830
3
GIS equipment
B
0.6659
5
0.5993
5
0.5993
5
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 10, Octobe
r 2014: 708
2
– 7091
7090
4.2. Analy
s
is
on Asse
ss
ment Re
sults
Asse
ssme
nt results
sho
w
that, as the main po
wer equipme
n
t in the sub
s
ta
tion, the
intelligent tra
n
sformation o
f
transformer
is pri
o
r
to the
other p
r
ima
r
y equipme
n
ts i
n
the intellige
n
t
transfo
rmatio
n pro
c
e
s
s
.
Ne
w sma
r
t intelligent sub
s
tati
ons were co
nstru
c
ted in
accordan
ce
with
stand
ard
co
n
s
tru
c
tion,an
d
the intellige
n
ce leve
l
of the prima
r
y equipm
ents
were hig
her
than
other tran
sformation
sub
s
t
a
tions.Th
e
a
p
p
rai
s
ed
va
lue
of the GIS e
quipme
n
t A is highe
r tha
n
the
transfo
rme
r
C in the
num
erical exam
pl
e.
That's because
the sub
s
tation whi
c
h
GIS
(ele
ctri
cal
equipm
ent A) located in
is a ne
w
built intellige
n
t sub
s
tation
, and the substatio
n
wh
ich
transfo
rme
r
C locate
d in
is a transfo
rmati
on sub
s
tation.The asse
ssm
ent re
sults me
et the
pra
c
tical p
r
oj
ect accu
ratel
y
.
From th
e re
sults
of different a
s
se
ssment
metho
d
s
,Vague
sets asse
ssment
method
int
r
odu
ce
s u
n
ce
rt
aint
y
t
o
cha
r
a
c
t
e
ri
ze f
u
zzy
d
a
t
a
m
a
thematically, so it i
s
mo
re re
asona
ble
to
cha
r
a
c
teri
ze
human fa
cto
r
,the discrimi
n
a
tion is hi
ghe
r for the
asse
ssment of obj
ects
with
simi
lar
levels, and th
e method p
r
o
posed in this
pape
r is
mo
re
con
s
iste
nt wi
th the actual
proje
c
t.
5. Conclusio
n
In this pap
er,
Vague sets t
heory
wa
s used to
solve a
s
sessme
nt q
uestio
n
s of
substatio
n
prima
r
y equi
p
m
ent intellige
n
t transfo
rma
t
ion. First, a
s
se
ssm
ent mo
del wa
s b
u
ilt according to t
he
relevant
stan
dard
s
a
nd re
sea
r
ch informati
on,
and Vague sets evaluation
al
gorithm has bee
n
improve
d
, which
ma
ke
s the a
s
se
ssment meth
o
d
mo
re
rea
s
on
able
and
accu
rate.
The
asse
ssm
ent
method
wa
s
applie
d to
asse
ss the i
n
te
l
ligen
ce refo
rm
level of
po
rtion of
pri
m
a
r
y
equipm
ents i
n
the
su
bstati
on, an
d was
verified
with
a given
exam
ple. The
next
wo
rk target i
s
to
apply thi
s
m
e
thod to
ev
aluate th
e o
v
erall intelli
g
ent tra
n
sfo
r
mation level
of the
intell
igent
sub
s
tation
an
d the
regi
on
al gri
d
. The
work
co
ndu
ct
ed in thi
s
pa
per
provid
es reality b
a
si
s to
asse
ss
sub
s
t
a
tion sma
r
t transfo
rmatio
n and
devel
op
approp
riate st
anda
rd
s for the grid.
Referen
ces
[1]
Haase P. Intelligrid:a Smart Net
w
ork of Po
w
e
r.
EPRI Journal.
200
5(F
a
ll):
17-2
5
.
[2]
Amin M, Sche
w
e
P F
.
Preve
n
t
ing bl
acko
uts: Buil
din
g
a Sma
r
ter Po
w
e
r Grid
.
Scientific America
n
. 200
8;
(8): 60-67.
[3]
YX
YU, WP LUAN.
Smart Grid
and Its Imple
m
e
n
tatio
n
s
. Proceedings of the C
SEE. 2009; 29(34): 1-8.
[4]
C XUE, CB L
I, XQ HUANG
, SW
T
A
NG,
Y LIU. A
pplic
ation
of Intelli
gent Su
bstatio
n
Informatio
n
Integratio
n.
Electric Power Auto
ma
ti
on
Eq
ui
pm
en
t.
2011; 3
1
(
7):110-
114.
[5]
Smart Grid D
e
partment
of C
h
ina
State Gri
d
Comp
an
y. Inte
l
lige
n
t T
e
chnic
a
l Co
nd
itions
of
High
Vo
ltag
e
S
w
itc
h
gear/GIS Equipm
ents. Bei Jing: Smart Grid
Departm
ent of China Stat
e Grid Company
.
2009
[6]
Smart Grid D
e
partment
of Ch
ina St
ate Grid
Comp
an
y. Intel
lige
n
t T
e
chnica
l Ana
l
ysis
Rep
o
rt on
Hi
g
h
Voltag
e Equ
i
p
m
ents[R]. Bei Jing: Smart Grid
D
epartme
n
t of Chin
a State Grid Comp
an
y. 2
009.
[7]
YC HU, JJ RU
AN, ZY DU, L RUAN, BH Y
A
NG, ZH
PU,
QJ XIE. Evalu
a
tion
of Subst
a
tion Prim
a
r
y
Equi
pment
Inte
llectu
a
liz
ation
Based on
T
O
PSIS.
Electric Pow
e
r Auto
mati
on Eq
ui
pment
. 201
2;
32(
12):
22-2
7
.
[8]
F. LI, A. LU, L.J. CAI. Multicriteri
a Decis
i
o
n
Makin
g
Bas
ed on
V
agu
e
Sets
.
J. H
u
a
z
hong
U
n
i
v
. o
f
Sci
.
& Tech.
, 2001, 29(7): 1-3
[9] LB
ZHU.
T
h
e
o
r
y and
App
lic
ation
Res
earch
on R
o
u
gh S
e
ts and
Va
gue
S
e
ts.
Che
ngd
ou
: South W
e
s
t
Jiao T
ong Univ
ersit
y
. 2
006; 1
0
-19.
[10]
YF
Z
H
AO, JF
CHEN. An
al
yti
c
hier
arch
y
pro
c
ess
a
nd
its a
pplic
atio
n i
n
p
o
w
e
r s
y
st
em. Electric P
o
w
e
r
Automatio
n
Eq
uipm
ent. 200
4; 24(9): 85-8
7
.
[11]
HZ NIE, P Lv
, Y QIAO, X
P
YAO. Com
p
reh
ensiv
e F
u
zz
y
Eval
uati
o
n
for
T
r
ansmission N
e
t
w
o
r
k
Plan
nin
g
Sche
me Based o
n
Entrop
y W
e
ig
h
t
Method.
Pow
e
r System T
e
c
hno
logy
. 2
009;
33(11): 60-
64.
[12]
W
Z
YAN, LS GU. Applic
atio
n
of the metho
d
of Entrop
y Pro
portio
n
in th
e E
ngi
neer
in
g Mar
k
.
Journa
l of
Xi
’
a
n Univ
ersit
y
of Architecture &techno
lo
gy
. 2004; 3
6
(1): 9
8
-10
0
.
[13]
Smart Grid D
epartme
n
t of
Chin
a State
Grid Com
p
a
n
y
.
Report on printi
ng
Inte
lli
g
ent
tech
nica
l
cond
itions of
transfor
m
er
s a
n
d
Hig
h Volta
g
e
Sw
itchgear
. BeiJin
g: Chi
na S
t
ate Grid Comp
an
y. 201
0.
[14] Chin
a
State
Grid Com
pan
y.
T
e
chnic
a
l sp
e
c
ificatio
ns for smartn
ess up
g
r
ade of su
bsta
tion
. BeiJi
ng:
Chin
a State Grid Com
pan
y. 2
010.
[15]
Z
H
AO Chenp
eng. T
he sele
cti
on an
d Ap
plicati
on Inves
t
igatio
n
of Primar
y
Eq
ui
pment in Smar
t
Substatio
n
. Bei
jing: N
o
rth Chi
na Electric Po
w
e
r Univ
ersit
y
.
2010.
[16]
ZS XU. Resea
r
ch on Comp
atibil
it
y
a
nd Co
n
s
is
tenc
y
of F
u
zz
y
Com
p
lem
e
ntar
y
Ju
dg
eme
n
t Matrices.
Journ
a
l of PLA
University of S
c
ienc
e an
d T
e
chno
logy
. 2
002;
3(2): 94-96.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Asse
ssm
ent of Intelligent Substation B
a
se
d on
an I
m
prove
d
Fuzzy Set
s
Meth
od (Yua
ncha
o Hu)
7091
[17]
CY Z
H
ANG, HY Z
H
OU. T
r
ansformati
on
and E
n
trop
y f
o
r Vag
ue S
e
ts. Computer
E
ngi
neer
in
g an
d
Appl
icatio
ns. 2
003; 33(
4): 20-
23.
[18]
GX
SONG, DL YANG. Methods fo
r
Ide
n
tifyi
n
g
an
d Impr
ovin
g th
e C
o
n
s
istenc
y
of F
u
zz
y
J
u
d
g
me
n
t
Matrix
.
Syste
m
s Engin
eeri
ng.
200
3; 21(1): 11
0-11
6.
[19]
SH
X
I
AO, ZP FAN, MG WANG. Study
on
Co
nsiste
nc
y of F
u
zz
y Ju
d
geme
n
t Matri
x
.
Journ
a
l of
Systems En
gin
eeri
n
g
. 20
01; 1
6
(2): 142-
14
5.
Evaluation Warning : The document was created with Spire.PDF for Python.