Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
9
, No
.
6
,
Decem
ber
201
9,
pp. 4
625~
4636
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v9
i
6
.
pp4625
-
46
36
4625
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
An an
alyt
ical m
ulti
cr
it
eria mod
el b
ased on
graph
theor
y for
re
li
ability enh
ancement in
distribu
tion ele
ctrical n
etworks
A.
Le
k
bich
1
, A.
Be
lf
qih
2
, T.
Ouder
hm
an
3
,
J
.
Bou
kher
ouaa
4
, F.
El
mari
ami
5
1
,2,4,5
Depa
rtment
of
Elec
tr
ical
N
etw
orks a
nd
Static
Converters,
Na
t
iona
l
Schoo
l
of
El
e
ct
ri
ci
t
y
and
Mec
hanics,
ENSEM
-
Hass
an
II
Univ
ersity
of
Casabl
an
ca,
Mor
occ
o
3
LS3M L
abor
a
to
r
y
,
FP
K,
Hass
an 1 Unive
rsit
y
,
M
oroc
co
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
hist
or
y:
Re
cei
ved
Oct
24
, 201
8
Re
vised
A
pr
2
7
, 2
01
9
Accepte
d
J
un
2
6
, 201
9
The
e
lectr
i
cal
di
stribut
ion
n
et
wo
rk
is
a
cri
t
ic
a
l
a
nd
complex
s
y
s
t
em
in
te
rm
s
of
saf
ety
and
r
el
i
abi
lit
y
,
b
ecau
se
i
t
is
compos
ed
of
d
iffe
r
ent
components
(sw
it
che
s,
r
ec
los
ers,
e
tc
.
).
The
i
m
prove
m
ent
of
i
ts
rel
i
abi
l
ity
is
t
her
efo
re
one
of
th
e
m
ost
important
t
asks
th
rough
th
e
good
m
ana
gement
o
f
remote
-
cont
rolled
sw
it
c
hes
and
re
cl
ose
rs
in
thi
s
n
et
wo
rk.
Thi
s
p
ape
r
pre
sents
an
ana
l
y
t
ical
m
odel
base
d
on
gr
aph
the
or
y
to
eva
lu
ate
SA
IDI
and
SA
IFI
indi
c
es
base
d
on
the
n
et
work
ar
chitect
ure
and
th
e
lo
c
at
ion
of
r
emote
-
cont
rol
le
d
rec
losers
and
sw
it
ch
es.
The
se
ind
ic
a
tors
hav
e
b
een
used
to
form
alize
a
m
ult
i
-
obje
c
ti
ve
m
at
he
m
at
ic
a
l
m
odel
t
hat
r
espe
ct
s
the
rea
l
oper
at
ion
c
onstrai
nts
of
equi
pm
ent
s i
n
s
m
art
grid.
The applie
d
m
ode
l,
in this
article, was
eva
lu
at
ed
on
an
I
EE
E
13
bus
net
work
using
th
e
TOPS
IS
m
et
hod
to
de
te
rm
ine
the
opti
m
a
l
loc
a
ti
on
of
the
s
witc
hes
and
re
clos
ers
and
to
improve
th
e
over
al
l
rel
i
abi
l
ity
of
the
d
istri
but
ion net
work.
Ke
yw
or
d
s
:
Distrib
ution p
ower
syst
em
ENS
Gr
a
ph the
or
y
Mult
i
-
obj
ect
iv
e optim
iz
at
i
on
Re
cl
os
ers
Rem
ote co
nt
rol
le
d
switc
hes
Re
li
abili
t
y
SA
I
D
I
SA
I
FI
TOP
S
IS
Copyright
©
201
9
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
An
as
s Le
kbic
h,
Dep
a
rtm
ent o
f El
ect
rical
N
et
works a
nd Stat
ic
Conver
te
rs,
Nati
on
al
Scho
ol of Ele
ct
rici
ty
an
d M
echa
ni
cs (
E
NS
EM
-
H
assan II
Unive
r
sit
y of
Casa
blanca,
M
orocc
o),
Road
El Ja
did
a
, K
m
7
, B
P: 8
118,
Oasis Casa
blanca,
Mo
ro
c
co
.
Em
a
il
:
anass.
le
kb
ic
h@gm
ai
l.c
om
1.
INTROD
U
CTION
Power
dist
rib
ut
ion
netw
orks
need
to
be
m
or
e
reli
able
a
nd
intel
li
gen
t
du
e
to
cu
stom
er
requirem
ents
and
the
c
om
plexity
of
t
opologies.
It’s
becom
e
ver
y
com
plex
a
nd
operate
withi
n
norm
al
ope
rati
ng
li
m
i
ts
[
1]
.
Wh
ic
h
re
quire
s
the
usi
ng
th
e
autom
at
ic
dev
ic
es
f
or
prot
ect
ion
,
m
on
it
ori
ng
an
d
m
et
ri
ng
t
hat
use
de
dicat
e
com
m
un
ic
at
ion
li
nk
[
2
,
4]
.
T
he
recl
os
ers
an
d
rem
ote
con
t
r
olled
s
witc
hes
hav
e
ta
ke
n
a
cr
ucial
par
t by
ne
twork
m
anag
ers
to
m
on
it
or,
detect
a
nd
a
uto
m
at
ic
a
l
ly
cl
e
ar
el
ect
rical
fau
lt
s
an
d
r
econfig
ur
e
distrib
ution
net
work
s
t
o
reduce t
he
f
aul
t nu
m
ber
a
nd
outa
ge
ti
m
e.
The
cu
rrent
s
it
uation
ha
s
encou
rag
e
d
el
e
ct
rical
util
ities
to
f
oc
us
the
ir
effor
ts
on
i
m
pr
ovin
g
the
reli
abili
ty
of
t
he
el
ect
rici
ty
distribu
ti
on
netw
ork
[
5]
,
w
hich
requires
s
ign
ific
a
nt
inv
e
st
m
ents
in
prot
ect
ion
and
s
witc
hing
dev
ic
es
s
uc
h
a
s
r
ecl
os
e
rs,
f
use
s
a
nd
disco
nn
ect
or
s
[
6
]
.
T
he
tree
arc
hitec
tu
re
of
t
he
el
ect
r
ic
al
distrib
ution
net
work
a
nd
the
hi
gh
nu
m
ber
of
el
ect
rical
fau
lt
s
pu
t
the
util
it
ies
op
e
rato
r
in
pe
rm
anen
t
diff
ic
ulti
es.
The
opti
m
a
l
locat
ion
of
reclo
sers,
switc
hes
and
protect
ion
dev
ic
es
in
distr
ibu
ti
on
netw
or
ks
is
one
of
t
he
m
os
t
eff
ect
ive
strat
e
gies
for
i
ncr
ea
sing
reli
abili
ty
an
d
reli
evi
ng
the
op
e
rato
r
[
7
].
H
ow
e
ve
r,
th
e
ch
oice
of
l
oc
at
ion
rem
ai
ns
a
c
omplex
ta
sk
requi
rin
g
non
-
li
near
ha
r
d
NP
m
ulti
-
ob
j
ect
ive
pro
bl
e
m
s
us
in
g
reli
abili
ty
ind
ic
es
base
d
abstract m
at
hem
at
ic
al
f
or
m
ul
a.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
2
01
9
:
4625
-
4636
4626
Ba
sic
al
ly
,
two
appr
oach
es
ca
n
be
use
d
to
e
valuate
the
reli
abili
ty
ind
ic
es
of
t
he
distrib
ution
syst
em
:
analy
ti
cal
fo
r
m
ula
ti
on
an
d
Mon
te
Ca
rl
o
s
i
m
ulati
on
(MCS)
[
8
]
.
Mo
re
i
m
po
rtantl
y,
an
al
yt
ic
al
te
chn
iqu
e
s
are
gen
e
rall
y
us
e
d
to
e
valuate
t
he
aver
a
ge
val
ues
of
reli
abili
ty
ind
ic
at
ors
[
9
,
10]
an
d
cha
racte
rize
the
di
ff
ic
ul
ty
of
form
ulati
ng
in
dicat
or
s
gi
ven
the
com
plexity
of
t
he
arc
hi
te
ct
ur
e
of
t
he
el
ect
rici
ty
net
works.
M
on
te
Ca
rlo
si
m
ulati
on
is
a
ble
t
o
pro
vid
e
i
nfor
m
at
ion
on
ind
e
x
va
riabil
it
y
an
d
is
flexi
ble
e
nough
to
ac
com
m
od
at
e
co
m
plex
op
e
rati
ng c
ondi
ti
on
s.
B
ut this
m
et
ho
d re
qu
i
re
s a
ver
y i
m
po
rt
ant cal
culat
io
n t
i
m
e p
er m
achi
ne [
11
].
In
r
ecentl
y
publis
hed
wor
k,
m
et
ho
dolo
gi
es
hav
e
bee
n
pr
opos
e
d
to
i
m
pr
ove
t
he
r
el
ia
b
il
it
y
o
f
the
distrib
utio
n
po
wer
syst
e
m
by
red
uci
ng
reli
abili
ty
ind
ic
es,
The
wor
k
[
12
]
ha
s
ad
op
te
d
a
Mo
nte
Ca
rlo
si
m
ulati
on
ap
proac
h
t
o
asses
s
reli
abili
ty
and
cost
i
n
a
powe
r
syst
em
us
ing
a
pa
rall
el
com
pu
ti
ng
e
nvir
onm
ent,
howe
ver,
an
d
t
o
deal
with
t
he
pro
blem
of
ev
al
uation
a
nd
re
li
abili
ty
op
tim
i
zat
ion
in
a
go
od
c
om
pu
ta
ti
on
tim
e,
researc
hers
ha
ve
opte
d
f
or
t
he
an
al
yt
ic
al
m
et
ho
d
base
d
on
bin
a
ry
val
ue
s.
I
n
[
13
]
,
th
e
auth
or
m
inim
iz
es
3
obj
ect
ives
S
AIDI
,
S
AIFI
an
d
cost
us
in
g
a
n
a
naly
ti
cal
m
od
el
howe
ver
the
e
xact
f
orm
ula
of
the
i
nd
ic
at
ors
is
not
addresse
d.
I
n
[
1
4
]
a
ne
w
nonl
inear
bin
a
ry
pro
gr
am
m
ing
m
od
el
ai
m
ing
at
m
in
i
m
iz
ing
th
e
reli
abili
ty
in
dice
s
SA
I
D
I
a
nd
SA
I
FI
of
a
distri
bu
ti
on
fe
e
der
this
m
od
el
al
lows
t
o
e
valu
at
e
with
pr
eci
sion
the
ind
ic
at
or
s
,
des
pite
this
f
orm
ula
is
com
ple
x
but
is
ve
ry
accurate
a
nd
easy
to
be
us
e
d
f
or
the
optim
al
locat
ion
of
reclo
ser
a
nd
switc
hes.
In
[
1
5
]
,
th
e
auth
ors
present
a
m
ulti
-
ob
j
ect
ive
op
t
i
m
iz
ation
s
to
ob
ta
i
n
the
best
reli
abi
li
ty
of
t
he
distr
ibu
ti
on
net
wor
k
w
hile
m
ini
m
iz
ing
the
co
sts
of
the
syst
em
.
I
n
[
1
6
]
,
t
he
a
uthor
s
pro
po
se
a
n
ana
ly
ti
cal
m
od
el
c
apab
le
to
repre
sent
t
he
interac
ti
on
s
bet
ween
t
he
protect
io
n
a
nd
s
witc
hing
de
vices
instal
le
d
on
the
el
ect
rici
ty
dist
rib
ution
netw
ork
,
w
hich
m
akes
it
po
ssible
t
o
opti
m
iz
e
the
two
S
AID
I
an
d
SA
I
FI
ind
ic
at
ors.
T
he
se
w
orks
s
how
s
that
the
reli
ab
il
ity
dep
e
nd
s
on
3
pa
ram
et
ers,
an
acc
ur
at
e
a
nd
sim
ple
evalu
at
ion
,
the
opti
m
iz
at
ion
m
od
el
of
the
evaluate
d
pa
ra
m
et
ers
and
fi
na
ll
y
the
equ
i
pm
e
nt
us
e
d
t
o
op
ti
m
iz
e
these
ind
i
cat
or
s.
In
this
pa
per
t
he
pro
blem
is
so
lve
d
i
n
it
s
in
te
gr
it
y
by
us
i
ng
a
new
f
or
m
ulati
on
of
reli
ab
il
ity
ind
ic
es
base
d
on
the
grap
hs
the
i
nd
ic
es
are
exa
ct
ly
assess
i
n
relat
ion
s
hi
p
with
t
he
locat
io
n
of
t
he
rem
ote
-
con
t
ro
ll
ed
swict
es
a
nd
t
he
recl
os
er
.
The
f
or
m
ula
cl
earl
y
desc
ribes
th
e
opti
m
iz
a
ti
on
pr
o
blem
of
re
li
abili
ty
by
stu
dyin
g
the
be
hav
i
or
of
equ
i
pm
ent
in
t
he
sm
art
gr
id,
t
he
eval
uation
of
these
in
dicat
ors
is
autom
at
ic
si
m
ple
to
m
achine
-
cod
e
d
usi
ng
the
force
of
the
ad
j
ace
nt
m
at
rix,
w
hi
ch
al
lo
ws
us
to
ge
ne
rat
a
set
of
sce
nar
i
os
.
These
scena
rio
s ar
e
subse
qu
e
ntly
treat
ed by
TOP
S
IS.
2.
RESEA
R
CH MET
HO
D
2.1.
Mathem
ati
cal
mo
d
el
ing
of d
istri
but
i
on
net
w
orks
Distrib
ution
ne
tworks
are
ge
ner
al
ly
de
sig
ne
d
as
netw
ork
s
of
c
om
plex
t
rees
with
l
oops,
in
no
rm
a
l
op
e
rati
on,
t
he
el
ect
rical
netw
orks
operate
i
n
rad
ia
l
a
rch
it
ec
ture.
w
he
n
a
fa
ult
occ
ur
e
the
swiche
s
an
d
re
cl
os
er
change
the
c
on
fig
ur
at
io
n
of
el
ect
rical
networ
ks
in
orde
r
to
s
upply
cust
om
e
rs
w
hile
res
pec
ti
ng
the
rad
ia
li
ty
of
the
netw
ork
.
F
ro
m
the
forego
ing
c
onsiderati
on
s
,
it
is
possible
to
re
presen
t
an
el
ect
rical
distrib
ution
ne
twor
k
us
in
g
a
tree str
uctu
re,
as
s
hown in Fi
g
ur
e
1.
Fig
ure
1.
T
he
e
le
ct
rical
d
ist
rib
ution net
wor
k pr
ese
ntati
on
The
s
ource
of
el
ect
rical
energ
y
is
the
zo
ne
1
of
t
he
gr
a
ph,
t
he
el
ect
rical
connecti
on
bet
wee
n
the
al
l
areas
of
t
he
el
ect
rical
netw
ork
is
pr
ese
nte
d
by
a
rro
ws
re
presenti
ng
t
he
di
recti
on
of
the
powe
r
fl
ow
w
hi
ch
al
lo
ws
us
to
def
i
ne
the do
wnstrea
m
zo
nes
of a
gi
ven sect
io
n.
Ma
them
a
ti
cal
m
od
el
of
el
ect
r
ic
al
distrib
utio
n
netw
orks
Let
,
G
V
E
be
a g
ra
ph,
w
her
e
V
is
a
se
t
of
N vertic
es,
a
nd is
a set
of
M ed
ges
b
et
we
en ve
rtic
es.
T
he
adjace
nt m
at
rix
is
the
m
at
rix
re
pr
e
sentat
io
n o
f
the
gr
a
ph r
e
presen
ti
ng
the
to
polo
gy of t
he
el
ect
r
ic
al
n
et
wor
k.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
An a
na
ly
ti
cal
mu
lt
ic
rit
eri
a
m
od
el
base
d on
gr
ap
h
t
heo
ry
…
(
Ana
ss
Lek
bi
ch
)
4627
The
tra
nsp
os
ed
m
a
trix of
A
is n
oted by:
(
)
(
2
)
,
tt
Aa
ij
(2)
The
tra
ns
it
ive
cl
os
ure
of G,
de
no
te
d by
**
(
)
(
3
)
,1
1
Aa
i
j
i
n
jn
(3)
is def
i
ned as
fol
lows
:
*
1
,
a
ij
if the
re
is a pat
h betwe
en verte
x
i a
nd v
e
rtex j
or i =
j
; 0
o
t
herwise.
The
G
raph
G
*
t
ran
sit
ive
cl
os
ure
of
G
a
nd
t
he
m
at
rix
A*
pr
ese
ntati
on
of
t
he
G
*
a
re
s
how
n
in
Fig
ure
2.
1
1
1
1
1
1
1
1
0
1
1
1
1
1
1
1
0
0
1
1
1
1
1
1
0
0
0
1
1
1
1
1
*
(
4
)
0
0
0
0
1
0
0
0
0
0
0
0
0
1
1
1
0
0
0
0
0
0
1
0
0
0
0
0
0
0
0
1
A
(4)
Figure
2
.
F
ai
lu
re
prop
a
gatio
n
t
A
*
the
tra
nsp
os
ed
m
at
rix
of
the
a
dj
ace
nt
m
at
rix
A*
of
a
gra
ph.
It
dete
rm
ines
the
el
em
ents
j
of
th
e
pr
e
decess
or ele
ct
rical
n
et
w
ork
o
f
the se
ct
ion
s
i d
e
no
te
d
P
. F
i
g 2 shows t
he gra
ph of
fail
ure pro
pagat
ion
wh
ic
h
gro
up
s
the
upst
ream
zo
nes of
the
fau
lt
z
on
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
2
01
9
:
4625
-
4636
4628
2.2.
Formul
at
i
on
of the pr
ob
le
m
To
represe
nt
t
he
lo
cat
ion
of
rem
ote
con
t
r
olled
s
witc
hes
an
d
reclose
rs
in
t
he
rad
ia
l
distrib
utio
n
netw
ork,
the
di
screte
va
riable
s
we
re
us
e
d
to
ind
ic
at
e
the
ins
ta
ll
at
ion
of
the
reclose
r
or
RC
S
in
a
sect
io
n
i
of
the
powe
r gr
id
and
are defi
ned as
fo
ll
ows
[
9]
:
0
(
5
)
1
i
f
a
r
e
c
l
o
s
e
r
i
n
i
X
i
i
f
n
o
t
(5)
0
(
6
)
1
i
f
a
R
C
S
i
n
i
Y
i
i
f
n
o
t
(6)
The
l
ocati
on m
at
rix
is
def
i
ned as foll
ows:
,
1
,
1
11
(
)
(
.
)
(7
)
t
x
i
j
i
n
i
j
i
i
n
j
n
j
n
P
p
a
x
(7)
,
1
,
1
11
(
)
(
.
)
(
8
)
t
y
i
j
i
n
i
j
i
i
n
j
n
j
n
P
p
a
y
(8)
The
tra
ns
it
ive
cl
os
ure
of
x
P
is not
ed
by
*
x
P
an
d
th
e
transiti
ve
cl
os
ure
of
y
P
is n
oted
by
*
y
P
.
The
t
wo
m
at
ri
ces
*
x
P
and
*
y
P
resp
e
ct
ively
express
the
path
s
c
onnecti
ng
the
are
a
of
t
he
el
ect
ri
ca
l
netw
ork
to
the
upstream
zon
e
s
thr
ough
the
r
ecl
os
er
a
nd
the
rem
ote
con
tr
ol
le
d
switc
h,
w
hich
will
al
low
us
t
o
def
i
ne
the
protect
ed
area
in
t
he
el
ect
rical
networks.
T
hey
ar
e
us
e
d
to
pr
eci
s
e
ly
locat
e
areas
protect
ed
by
cut
-
off
dev
ic
es
in
t
he r
adial
n
et
w
ork
.
Fig
ure
3
sho
w
s
in
re
d
the
f
ai
lure
pro
pa
ga
ti
on
bet
ween
sect
ion
s
i
an
d
j
cau
sed
by
a
fau
l
t
in
i.
T
he
c
urren
t
fa
ults
pro
pag
at
es
in
the
upstrea
m
zon
es
an
d
ge
ner
at
es
a
n
el
ect
rical
fau
lt
s
in
the
zo
nes
do
w
ns
tream
of all
the z
on
e
s
aff
ect
e
d by th
e fa
ults.
Fig
ure
3.
The
fa
ult rate i
n el
ect
rical
n
et
w
orks
The
ij
desc
ribes
t
he fault
r
at
e
in j ca
us
e
d by fa
ul
t i
n
i.
The
fa
ult
rate
is
lim
i
te
d
by
th
e
locat
ion
of
t
he
reclose
r,
how
ever
t
he
us
e
of
a
rem
ote
con
tr
olled
switc
h
m
ini
m
iz
es
the
ou
ta
ge
ti
m
e
by
the
fast
a
nd
a
uto
m
at
ic
isolation
of
the
sect
i
on
s
af
fected
by
the
fa
ult.
Fig
ur
e
4
sh
ows
the
Fail
ur
e
propa
gatio
n
in
elec
tric
al
n
et
w
orks wit
h and wit
ho
ut r
e
cl
os
er
.
Fig
ure
4.
Fail
ure
prop
a
gatio
n i
n
el
ect
rical
n
e
tworks
with
and
without
reclo
ser
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
An a
na
ly
ti
cal
mu
lt
ic
rit
eri
a
m
od
el
base
d on
gr
ap
h
t
heo
ry
…
(
Ana
ss
Lek
bi
ch
)
4629
If
we
us
e
a
rec
loser
in
t
he
pat
h
bet
ween
t
he
fau
lt
locat
io
n
a
nd
powe
r
sou
r
ce,
the
de
fau
lt
pro
pag
at
io
n
m
us
t
be
lim
it
ed
to the
downstream
are
a of the
r
ecl
os
e
r.
In or
der to
det
ect
all
sect
ion
s
of the
power
gri
d
a
ff
ect
ed
by t
he fa
ult,
we de
fine
the
m
at
rix
as foll
ows:
**
.
(
,
)
(
9
)
xx
P
A
C
i
j
(9)
If
we
us
e
a
r
e
m
ote
co
ntro
ll
ed
s
witc
h
i
n
t
he
path
bet
we
en
th
e
f
a
ult
lo
cat
ion
a
nd
power
sou
rce,
the f
a
ult p
ropa
gation m
us
t be
lim
it
ed
to the
downstream
are
a of the
r
em
ote contr
olled s
witc
hes.
**
(
,
)
(
1
0
)
yy
P
A
C
i
j
(10)
If
(
,
)
0
C
i
j
X
,
it
m
eans
that
the
j
is
no
t
protect
ed
against
f
ault
with
recl
os
er
/swit
ch
f
ro
m
the
fau
lt
i
n
i.
If
(
,
)
0
C
i
j
X
it
m
eans th
at
a
r
ecl
os
er/s
witc
h
is
instal
le
d i
n t
he
p
at
h betwe
en
i
and
j.
E
val
uatio
n
of
re
pair
ti
m
e
and
fa
ult
rate
i
n
rad
ia
l
feed
e
r
netw
orks
.
If
*
1
,
a
ij
,
t
his
im
plies
that
the
loa
d
j
is
dow
ns
tream
of
t
he
sect
io
n
i
of
t
he
el
ect
rical
network,
eac
h
f
ault
in
i
ge
ner
a
te
s
a
po
w
er
outa
ge
in
j
a
nd
the
rest
or
e
d
el
e
ct
rical
p
owe
r
i
s r
el
at
ed
t
o
t
he repair
of t
he
se
ct
ion
i.
In this case
, the
f
a
ult rate a
nd the
rep
ai
r
ti
m
e
are
i
den
ti
cal
in
al
l t
he se
ct
io
ns
j
.
(
1
1
)
i
j
i
(11)
(
1
2
)
rr
i
j
i
(12)
If
*
0
,
a
ij
,
it
m
eans
that
the
loa
d
j
is
dow
nst
ream
of
the
sect
ion
i
of
the
el
ect
rical
network,
the
instal
la
ti
on
of
t
he
reclo
se
ra
ns
Re
m
ote
co
ntr
olled
s
witc
hes
will
m
ake
possible
to
isolat
e
t
he
fa
ulty
sect
ion
i.
and
de
crease th
e
ou
t
age tim
e.
.
(
,
)
(
1
3
)
C
i
j
i
j
i
x
(13)
.
(
,
)
.
(
,
)
(
1
4
)
,
r
r
C
i
j
r
C
i
j
i
j
i
x
i
s
o
i
y
(14)
2.3.
For
mul
at
i
on
of rel
iabil
it
y
in
dice
s
To
e
valuate
th
e
reli
abili
ty
of
el
ect
rici
ty
gr
id
s,
m
any
ind
ic
e
s
are
us
e
d
by
ut
il
i
ti
es,
accor
di
ng
to
IEE
E
Std
13
66
-
20
12
[
17
].
P
ow
e
r
ut
il
ities
fr
eq
ue
ntly
us
e
SAID
I
a
nd
S
AI
F
I
to
asse
ss
the
reli
abili
ty
of
the
powe
r
gr
i
d.
The
(SAI
D
I)
giv
es
the
a
ve
rag
e
inter
r
up
ti
on
ti
m
e
per
consum
er
f
or
one
ye
ar
a
nd
t
he
(SAI
FI)
gi
ves
the av
e
ra
ge of
the syst
em
ind
ex of
the
fr
e
qu
ency o
f
i
nterru
ptions
[
18]
.
..
11
(
1
5
)
1
nn
rN
i
j
i
j
i
ij
S
A
I
D
I
n
N
i
i
(15)
.
11
(
1
6
)
1
nn
N
i
j
i
ij
S
A
I
F
I
n
N
i
i
(16)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
2
01
9
:
4625
-
4636
4630
These
i
nd
ic
es
de
pend
on
the
a
bili
ty
of
the
protect
io
n
de
vices
to
e
lim
inate
fau
lt
s
as
quic
kly
as
possible
an
d
to
rec
onfi
gur
e
the
netw
ork
t
opology
an
d
st
ron
gly
relat
ed
on
t
he
locat
io
n
of
the
reclose
r
s
and
the
rem
ote
co
ntr
olled
s
witc
he
s.
We
re
place
the
fa
ult
rates
(
13)
an
d
re
pa
ir
ti
m
es
(1
4)
by
their
f
or
m
ulas
f
or
getti
ng
t
he fo
r
m
ula
ti
on
s
of
S
AIDI an
d SA
I
FI
.
..
11
*
1
1
(
1
7)
(
.
(
,
)
.
(
,
)
.
(
1
(
,
)
)
)
.
11
*
0
1
nn
rN
i
i
i
ij
i
f
a
ij
n
N
i
i
S
A
I
D
I
nn
C
i
j
r
C
i
j
r
C
i
j
N
i
x
i
y
i
s
o
y
i
ij
i
f
a
ij
n
N
i
i
(17)
.
11
*
1
1
(
.
(
,
)
)
.
11
*
0
1
(
1
8
)
nn
N
ii
ij
i
f
a
ij
n
N
i
i
S
A
I
F
I
nn
C
i
j
N
i
x
i
ij
i
f
a
ij
n
N
i
i
(18)
2.4.
Formul
at
i
on
of c
os
t
f
u
ncti
on
All
decisi
on
m
ade
by
the
el
e
ct
rical
netw
ork
en
gin
eer
s
to
m
ini
m
iz
e
the
r
el
ia
bili
ty
ind
ic
es
re
qu
ire
d
inv
est
m
ents
to
ei
ther
the
e
qu
ipm
ents
le
vel
or
t
he
an
nual
m
ai
ntenan
ce
of
the
e
qu
i
pm
e
nts
[
19
]
.T
he
optim
a
l
so
luti
on
that
i
m
pr
ov
es
the
re
li
abili
ty
of
the
el
ect
rical
distribu
ti
on
syst
em
thr
ough
a
set
of
decisi
ons
i
nvolv
i
ng
the
us
e
of
pr
otect
ion
s
an
d
s
w
it
ches
dev
ic
es
by
a
n
acce
pta
bl
e
pri
ce
rem
ai
n
s
on
e
of
the
c
ha
ll
eng
es.
T
he
no
ti
on
of
c
os
t
is
pa
ra
m
ou
nt
for
in
ve
st
m
ents
giv
e
n
the
com
plexit
y
of
the
pro
po
sed
s
olu
ti
ons
a
nd
t
he
li
m
it
ati
on
of
resou
rces
for m
os
t energ
y
distribu
t
or
s
ope
r
at
ing
in
a c
ompeti
ti
ve
m
ark
et
[
20
]
.
The
c
os
t
of
the
so
l
ution
is
t
he
su
m
of
the
fi
xe
d
c
os
ts
as
so
ci
at
ed
with
ca
pital
inv
est
m
ent
on
s
witc
hes
and pr
otect
ion
dev
ic
es
an
d
t
he
co
st
of inter
r
up
ti
ons
[
21
]
.
T
his c
os
t i
s
desc
ribe
d by the
fo
l
lowing e
quat
io
n:
*
(
1
9
)
C
o
s
t
F
C
C
t
E
N
S
(19)
ENS
is
the t
otal no
t
distrib
ute
d
e
nergy
descr
i
bed as
fo
ll
ows:
.
.
(
2
0
)
11
nn
E
N
S
L
r
i
i
j
i
j
ij
(20)
Ct
is the
cost
of e
nergy,
in
thi
s p
a
pe
r
Ct
=
0,0
79 e
uro/k
w
h.
F
C:
the c
os
t
of
purc
hasin
g eq
ui
pm
ent (
switc
he
s and
reclose
rs) in e
uro
inclu
ding m
ai
ntenance
.
2.5.
Constr
aints
To
ch
oose
the
locat
ion
of
th
e
s
wiches
a
ns
recl
os
ers
i
n
el
ect
ri
cal
ne
t
works
,
f
un
ct
io
nal
co
ns
t
raints
m
us
t
be
ta
ke
n
int
o
a
ccount
to
li
m
i
t
the
searc
h
s
pa
ce
and
f
ocu
s
on
the
feasible
s
cenari
os
;
the
m
axi
m
u
m
nu
m
ber
of
reclose
rs
in
li
ne
,
the
co
ordi
na
ti
on
bet
ween
e
qu
i
pm
ents
and
the
locat
io
n
of
the
switc
hes
re
la
ti
ve
to
the
re
cl
os
er.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
An a
na
ly
ti
cal
mu
lt
ic
rit
eri
a
m
od
el
base
d on
gr
ap
h
t
heo
ry
…
(
Ana
ss
Lek
bi
ch
)
4631
On
e
of
the
m
os
t
im
po
rtant
const
raints
of
the
obj
ect
ive
f
u
nctio
n
is
t
he
coor
din
at
io
n
of
protect
i
on
dev
ic
es
(r
ecl
ose
rs).
The
nu
m
ber
of
de
vic
es
that
can
be
instal
le
d
in
series
is
lim
it
e
d
f
or
the
pur
pose
of
m
ai
ntaining
c
oor
din
at
io
n
[
22
].
I
n
this
arti
cl
e
the
m
axi
m
u
m
num
ber
of
re
cl
os
ers
in
se
ri
es
is
lim
it
ed
t
o
3
to
ex
pl
oit t
he
m
axim
u
m
f
un
ct
io
nalit
y of
t
he
e
quipm
ent in a
n op
ti
m
al
co
ordinati
on level
[1
5
].
2.6.
Multicri
teri
a deci
sion an
alysi
s ba
se
d
on
t
he TOPS
IS m
eth
od
To
de
te
rm
ine
t
he
num
ber
an
d
locat
ion
of
s
witc
hes
dev
ic
e
s
in
a
distrib
ut
ion
netw
ork,
we
will
nee
d
a
m
ult
i
-
crit
eria
decisi
on
a
naly
sis
m
et
ho
d;
TOP
S
IS
is
on
e
of
the
m
os
t
widely
us
e
d
m
ul
ti
-
crit
eria
decisi
on
analy
sis
m
et
ho
ds
[23],
[
24]
.
T
his
m
et
ho
d
ra
nk
s
t
he
a
lt
ern
at
ives
acc
ordin
g
to
the
p
r
ox
im
ity
be
tween
the
al
te
r
native
s
an
d
the
ideal
al
te
rn
at
ives
,
wh
ic
h
a
re
obta
ined
by
cal
cul
at
ing
t
he
dista
nce
of
th
e
al
te
rn
at
ives
from
the p
osi
ti
ve
ideal
an
d
t
he
ideal n
egati
ve
alt
ern
at
ives
.
The Speci
fic
steps of
T
OP
S
IS m
e
tho
d:
S
TE
P
1
:
is
t
o
est
ablis
h
a
de
ci
sion
m
at
rix
f
or
the
rankin
g.
T
he
struct
ur
e
of
the
m
at
rix
can
be
e
xpres
s
ed
a
s
fo
ll
ows
[24]
:
........
1
2
3
...
1
1
1
2
1
3
1
1
...
2
2
1
2
2
2
3
2
(
2
1
)
...
3
3
1
3
2
3
3
3
...
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
...
1
2
3
C
C
C
C
n
z
z
z
z
A
n
A
z
z
z
z
n
D
A
z
z
z
z
n
A
z
z
z
z
m
m
m
m
m
n
(21)
In
decisi
on
m
a
trix,
A
i
denotes
the
al
te
rn
at
ive
s
i
,
i
1,2
,...,m
.
C
j
rep
rese
nts
th
j
crit
eria,
relat
e
d
to
th
i
al
te
rn
at
ive.
z
ij
is t
he per
f
or
m
ance
rati
ng
s
fo
r
eac
h
al
te
r
native
A
i
with
res
pec
t t
o
each
crite
ri
a
C
j
.
STE
P 2:
Ca
lc
ulati
on
of the
nor
m
al
iz
ed
decisi
on m
at
rix
R usin
g n
or
m
al
izati
on
vecto
r:
(
)
(
2
2
)
1
1
Rr
i
j
i
m
jn
(22)
Wh
e
re
:
(
2
3
)
2
1
z
ij
r
ij
n
z
ij
i
(23)
STE
P
3:
Ca
lc
ulati
on
of
the
weig
hted
no
r
m
al
iz
ed
m
at
ri
x
V
by
m
ulti
plyi
ng
the
c
olum
ns
of
the
no
rm
alized
decisi
on m
at
ri
x
R
by the
ass
oc
ia
te
d
wei
gh
ts
j
w
sat
isfyi
ng
1
1
;
(
2
4
)
n
j
j
w
(24)
1
1
(
)
(
2
5
)
i
j
i
n
jm
Vv
(25)
w
he
re
:
.
(
2
6
)
i
j
i
j
j
v
r
w
(26)
Dep
e
ndin
g
on
the
purpose
and
the
decisi
on
-
m
aker
s,
dif
fer
e
nt
eval
uation
crit
eria
ha
ve
diff
e
re
nt
weig
hts, he
nce
the
need to
cal
culat
e the
weig
hted n
or
m
al
iz
e
d decisi
on m
atr
ix
[
25
].
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
2
01
9
:
4625
-
4636
4632
STE
P 4:
Deter
m
inati
on
of t
he
posit
ive ideal
so
luti
on
PI
S
and
of the
ne
gative ideal
so
l
utio
n NI
S
.
The
P
IS
is
the o
ne
with
the best
le
vel
fo
r
al
l
of
the
at
tri
bu
t
es
consi
der
e
d.
Th
e
N
IS
is
the o
ne
with
the
worst att
rib
ute
values
.
STE
P
5:
Ca
lc
ul
at
ion
of
the
E
uclidea
n
dista
nc
es
of
each
al
te
rn
at
ive
A
i
from
the
posit
ive
ideal
so
luti
on
P
IS
a
nd
from
the n
egati
ve
ideal
so
l
utio
n NI
S:
2
1
(
)
(
2
7
)
n
i
ij
j
j
d
v
v
(27)
a
nd
,
2
1
(
)
(
2
8
)
n
i
ij
j
j
d
v
v
(28)
The
relat
ive
c
losenes
s
to
th
e
ideal
s
olu
ti
on
is
cal
c
ulate
d,
an
d
the
ra
nking
of
the
al
te
rn
at
ives
i
n
desce
nd
i
ng orde
r
is
done. T
he
close
ness
c
oeffici
ent of eac
h al
te
rn
at
ive is c
al
culat
ed
as
f
ollow
s
[
26
]:
(
2
9
)
i
i
ii
d
CC
dd
(29)
The ran
king
of
alt
ern
at
ives
is
determ
ined
by
com
par
ing
i
CC
values.
3.
RESU
LT
S
A
ND AN
ALYSIS
3.1.
Implem
ent
ati
on
and s
olv
in
g
usin
g t
op
sis
In
order
t
o
te
st
the
i
m
pact
of
the
locat
io
n
of
r
ecl
os
ers
a
nd
s
witc
hes
de
vice
s
on
t
he
reli
abili
ty
ind
ic
at
ors
and
t
o
ch
oos
e
the
optim
al
so
luti
on,
we
us
ed
t
he
IE
EE
13
bus
netw
ork.
T
he
ta
ble
belo
w
li
sts
the
diff
e
re
nt
el
ect
rical
value
s
of
the
I
EEE
13
bu
s
.
For
each
branc
h
of
t
his
ne
twork
we
pro
pose
a
var
ia
ble
r
epair
tim
e
of
3
to
6
hours
as
s
how
n
in
T
a
ble
1
,
t
his
ti
m
e
is
li
nk
e
d
to
t
he
pe
rm
anen
t
fau
lt
an
d
a
fi
xed
is
o
la
ti
on
ti
m
e
of
0.08 h co
rr
e
spondin
g
t
o
the
re
cl
os
er i
nter
ve
ntion t
i
m
e to elim
inate
the tran
sie
nt
fau
lt
s.
Table
1
. L
ist
s
the
diff
e
re
nt elec
tric
al
v
al
ues
of the
I
EE
E
13
bus
Bran
ch
Li
(
KVA)
N of
Cu
sto
m
er
Fau
lt r
ate
Rep
aire
ti
m
e
Iso
le
m
en
t ti
m
e
1
0
0
0
,1
3
0
,08
2
0
0
0
,15
4
0
,08
3
150
55
0
,2
6
0
,08
4
630
80
0
,25
3
0
,08
5
220
200
0
,15
2
0
,08
6
720
47
0
,1
2
0
,08
7
445
10
0
,1
4
0
,08
8
30
4
0
,15
5
0
,08
9
340
200
0
,2
6
0
,08
10
400
10
0
,25
3
0
,08
11
630
117
0
,15
3
0
,08
12
1250
145
0
,1
2
0
,08
13
630
70
0
,1
2
0
,08
3.2.
SA
I
DI an
d SAIFI c
alcul
at
i
on
In
t
his
case
we
us
e
2
reclose
rs
in
sect
io
n
2;
7
and
a
r
em
ote
con
t
ro
ll
ed
switc
h
in
4.
T
he
F
i
gure
5
sho
ws
the
fa
ult
rate
m
at
rix
a
fter
al
l
oc
at
ion
of
2
reclo
sers
a
nd
rem
ote
a
c
ontrolle
d
s
witc
h.
A
fter
th
e
cal
culat
ion
of
fa
ult
rate
m
at
rix
an
d
the
ti
m
e
ou
ta
ge
m
at
rix
us
i
ng
the
(13
)
a
nd
(
14)
f
orm
ula,
we
sh
ow
n
in
F
ig
ur
e
5
t
hat
t
he
fa
ult
rate
m
at
rix
dep
e
nd
e o
n
t
he
locat
ion
of
re
nclenc
heurs.
T
he ou
t
age ti
m
e
m
a
trix r
ij t
he
de
pe
nde on r
em
ote co
nt
ro
ll
ed
switc
h place
m
ent.
Using t
he fo
rm
ula (
17)
a
nd (1
8)
t
he
S
A
IDI a
nd S
AI
F
I a
re
giv
e
n belo
w:
SA
I
D
I
=
6,1
726 h/y
ear
SA
I
FI
=
1,5
724 fault/
ye
ar
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
An a
na
ly
ti
cal
mu
lt
ic
rit
eri
a
m
od
el
base
d on
gr
ap
h
t
heo
ry
…
(
Ana
ss
Lek
bi
ch
)
4633
Figure
5.
The
fa
ult rate m
at
rix
3.3.
So
lvin
g the
op
timi
sa
ti
on
probl
em
usin
g
T
OPSIS
In
this
stu
dy,
a
m
ulti
-
ob
j
ect
i
ve
ap
proac
h
to
i
m
pr
ov
e
the
rel
ia
bili
ty
of
the
powe
r
gr
i
ds
w
as
propose
d
and
pr
ese
nted
in
Fi
gure
6
an
d
it
was
eval
ua
te
d
by
a
su
cce
ssion
of
sim
ul
at
ion
s
of
a
pr
ogram
i
m
ple
m
e
nted
on
MATLAB
.
Ba
sed
on
data
o
f
t
he
Table
2
t
he
Table 3
s
ummari
zes
al
l
res
ults of
the
pro
po
s
ed
m
et
hod
so
l
ve
d
by
TOP
S
IS.
It
groups
to
gethe
r
15
ra
ndom
ly
sel
ect
ed
scen
arios.
T
hese
s
cenari
os
res
pe
ct
the
optim
izat
ion
const
raints
of
t
he
pr
ob
le
m
.
Th
e
weig
hts
of
th
e
obj
ect
ives
ar
e
consi
de
red
i
den
ti
cal
an
d
e
qu
al
t
o
0.3
3
a
nd
t
his
m
akes it p
os
si
ble to m
ake
a
de
ci
sion
base
d
on the
im
po
rtan
ce o
f
the
ob
j
ect
ives.
Figure
6.
Flo
w
char
t
of the
so
l
ution
Table
2
.
T
he
a
nnual c
os
t
of br
eakin
g dev
ic
es
inclus
ding t
he m
ai
ntenan
ce
Dev
ice
Co
st(eu
ro/Year)
Reclo
ser
6000
Re
m
o
te con
trolled
switch
2500
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
9
, N
o.
6
,
Dece
m
ber
2
01
9
:
4625
-
4636
4634
Table
3
. R
es
ults o
f propose
d
m
et
ho
d sol
ved
by T
OP
S
IS
Locatio
n
of
recloser
Locatio
n
of
RCS
Cci
S1
1
5
;8
;1
1
0
,23
8
7
0
8
3
1
2
3
1
2
S2
1;
5
7
;1
1
0
,30
5
8
6
4
7
9
0
6
5
0
S3
3
; 8
5
0
,36
7
7
9
2
7
4
4
1
8
4
S4
1
; 6; 7
11
0
,36
2
6
3
6
0
7
9
0
7
9
S5
1
;8
; 11
6
;7
;
0
,38
4
2
0
2
7
8
8
7
4
1
S6
1
7
;8
0
,22
4
2
1
8
4
1
6
5
1
6
S7
1
;8
;1
1
6
;1
3
0
,38
2
0
4
4
3
3
1
5
6
0
S8
1
;8
;1
1
7
;1
3
0
,38
9
7
7
4
1
9
5
9
4
5
S9
1
;1
1
7
;8
0
,27
1
7
7
3
0
5
2
2
7
0
S1
0
6
;9
;1
3
7
;1
1
;1
2
0
,47
6
6
1
2
0
1
5
6
3
7
S1
1
2
;5
8
;1
1
0
,35
9
3
8
8
8
2
6
1
3
1
S1
2
1
;7
;8
;1
1
4
0
,49
2
7
4
0
1
9
9
9
4
8
S1
3
4
;7
;8
;1
1
5
;6
;1
3
0
,67
1
4
3
7
1
2
6
9
5
3
S1
4
3
--
0
,35
8
4
1
7
2
3
6
5
1
9
S1
5
3
4
;7
;8
;1
1
0
,41
4
2
5
2
9
9
2
1
3
8
The
best
sce
na
rio
acco
r
ding
to
the
T
OPSIS
m
et
ho
d
is
sce
nar
i
o
nu
m
ber
6
see
t
he
T
able
4
.
The
a
dvanta
ge
of
our
m
et
hod
is
t
he
flexi
bi
li
t
y
in
te
rm
s
of
decisi
on
m
akin
g
as
well
as
the
sim
plici
ty
of
i
m
ple
m
entat
io
n
on
m
achine
,
we
can
m
anip
ulate
severa
l
obj
ect
iv
es
an
d
c
hoos
e
dif
fer
e
nt
weig
ht
.
T
he
optim
al
locat
ion
of
recl
os
ers
an
d
swit
ches
c
orres
pondin
g
to
the
bes
t
obj
ect
ive
f
unct
ion
s
value
ar
e
ch
os
e
n
as
t
he
final
so
luti
on i
n
I
EE
E 13
bus the
best
so
luti
on a
re
a r
ecl
os
er
in 1
and 2 swit
c
hes i
n
7 an
d
8.
Table
4
.
T
he
best
scena
rio
ac
cordin
g
t
o
the
TOP
S
IS
S6
recloser in 1
swicth
es in
7;8
SAID
I=5,4
2
2
6
SAIFI
=2
Co
st=1
2
7
3
9
,8 eu
ro
/Year
4.
CONCL
US
I
O
N
I
n
this
a
rtic
le
,
we
ha
ve
pro
posed
a
n
analy
ti
cal
m
od
el
to
e
va
luate
the
t
wo
re
li
abili
ty
ind
ic
es
S
AIDI
an
d
SA
I
FI.
This
si
m
ple
m
achine
-
i
m
ple
m
ented
m
od
el
us
es
the
adj
ace
nt
m
at
ri
x
t
o
determ
ine
fau
lt
pro
pa
gation
paths
,
wh
ic
h
al
lo
ws
a
ccur
at
e
e
valuat
ion
o
f
fa
ult
rates
an
d
re
pair
ti
m
es
fo
r
eac
h
ne
twork
a
rea.
A
fter
e
valuati
on o
f
t
he
reli
abili
ty
ind
ic
es
,
a
m
ulti
-
ob
j
ect
iv
e
m
at
h
e
m
at
ic
al
m
od
el
is
propose
d
to
fin
d
the
op
ti
m
al
locat
ion
of
the sw
it
c
hes
a
nd r
e
cl
os
e
rs using the
TOPSI
S
m
e
tho
d.
Si
nce
the m
ulti
-
obj
e
ct
ive proble
m
i
s a d
i
ff
ic
ult pro
ble
m
and
ta
kes
a
lo
ng
ti
m
e
to
reso
lve,
t
he
T
OPSIS
m
et
ho
d
breaks
m
any
senar
i
os
that
res
pect
the
ex
ploi
ta
ti
on
const
raints
a
nd
s
olv
es
t
he
pro
blem
of
im
pr
ov
in
g
the
reli
abili
ty
via
t
he
posit
ioni
ng
of
reclose
rs
an
d
the locati
on
of
switc
hes
i
n
a
n effici
ent cal
c
ulati
on
ti
m
e.
REFERE
NCE
S
[1]
Daz
ahr
a
M
N,
El
m
ari
ami
F
,
B
el
fqih
A,
Boukh
ero
uaa
J.,
"S
m
art
Lo
ca
l
Ba
ckup
Protec
ti
on
For
S
m
art
Subs
ta
ti
on
,
"
Inte
rnational
Jo
urnal
of El
e
ct
ri
c
al
and
Comput
er
Engi
n
ee
ring
(
IJE
CE)
;
vol. 7, no. 5,
pp
.
2321
-
23
28,
Oct
2017.
[2]
Draga
n
M,
Sret
e
N
,
Emir
A.,
"
Designing
Auto
m
at
ic
Met
er
Re
adi
ng
S
y
st
em
U
sing
Open
Sour
ce
Hardwar
e
an
d
Software
,
"
Int
ernati
onal
Journal
of
Elec
tric
al
an
d
Computer
Eng
ine
ering
(
IJE
C
E
)
,
vol
7,
no.
6,
pp
.
3282
-
3291,
D
e
c
2017.
[3]
Le
kbi
ch
A
,
B
el
f
qih
A,
Ze
d
ak
C
,
Boukherouaa
J
,
Mari
ami
F,
"
A
sec
ure
wire
le
ss
cont
rol
of
Remo
te
Te
rm
in
al
Unit
using
th
e
In
te
r
net
of
Thi
ngs
i
n
sm
art
g
rid
,
"
6th
In
te
rnationa
l
Conf
ere
nc
e
o
n
Wireless
Net
w
orks
and
Mob
ile
Comm
unic
ati
ons (
WINCOM
)
,
2018.
[4]
Ze
dak
C
,
Le
kb
ich
A,
Be
lfqi
h
A,
Boukheroua
a
J,
Haidi
T
,
M
ari
a
m
i
F,
"
A
propos
ed
se
cur
e
r
emote
d
ata
ac
qu
isit
io
n
arc
hi
te
c
ture
of
p
hotovol
taic
s
y
ste
m
s
base
d
on
the
Inte
rn
et
of
Thi
n
gs
,"
6th
Inte
rnat
ional
Conf
ere
nc
e
on
Mul
ti
medi
a
Computing
and
Syste
ms
(
ICMCS)
,
2018
.
[5]
Venka
ta
S
B
K,
Madhusudan
V
,
G
ane
sh
V,
"P
roba
bil
ist
ic
Perfo
rm
anc
e
Inde
x
b
ase
d
Con
ti
ngen
c
y
Scre
eni
ng
for
Com
posite
Pow
er
S
y
st
em
Relia
bil
ity
Ev
al
ua
ti
on
,
"
I
nte
rnat
ional
Journal
of
Elec
t
rical
and
Comput
er
Engi
n
ee
rin
g
(
IJE
CE)
,
vol. 8,
no.
5
,
pp
.
2661
-
2670,
Oct
2018.
[6]
Kum
ar
KS
,
Ja
y
a
bar
at
h
i
T
,
"
Pow
er
s
y
st
em
re
conf
igura
ti
on
and
lo
ss
m
ini
m
iz
at
ion
for
an
d
istri
but
i
on
s
y
stems
usin
g
bac
t
eri
a
l
for
agi
n
g
opti
m
izati
on
a
l
gorit
hm
,
"
Int
J
E
le
c
tr P
ower
Ener
gy
Syst
,
vol
.
36
,
no.
1,
pp.
13
–
1
7
,
2012
.
[7]
R.
Bil
l
int
on
,
P.
W
ang,
"
Te
a
chi
n
g
distri
buti
on
s
ystem
rel
ia
b
il
i
t
y
eva
lu
at
ion
using
Monte
Carl
o
si
m
ula
ti
on,
"
I
EEE
Tr
ans.
Powe
r Sy
st.
,
vo
l.
14
,
no.
2
,
pp
.
397
–
403
,
Ma
y
1999.
[8]
Chit
um
odhu
B,
Renuka
L,
"
An
open
source
tool
for
re
li
ab
il
i
t
y
e
val
ua
ti
on
of
dist
ribut
ion
s
y
stem
using
Monte
Car
l
o
sim
ula
ti
on
,"
Ind
onesian
Journal
of
E
lectric
a
l
Engi
ne
ering
an
d
Computer
Sc
i
enc
e
(
IJE
ECS)
,
v
ol
.
14
,
no.
3
,
pp.
1065
-
1075
,
J
un
2019
.
[9]
Za
m
bon
E
,
Bo
ss
ois
DZ,
Gar
cia B
B,
Az
ere
do
EF,
"
A
novel
nonl
in
e
ar
progra
m
m
ing m
odel
for
distr
ibut
ion
prot
ec
t
io
n
opti
m
iz
ation,
"
I
EE
E
Tr
ans P
ower
Delive
ry
2009
,
vol
.
24
,
pp
.
195
1
–
8
,
2009
.
Evaluation Warning : The document was created with Spire.PDF for Python.