T
E
L
KO
M
N
I
KA
T
e
lec
om
m
u
n
icat
ion
,
Com
p
u
t
i
n
g,
E
lec
t
r
on
ics
an
d
Cont
r
ol
Vol.
18
,
No.
4
,
Augus
t
2020
,
pp.
2235
~
224
4
I
S
S
N:
1693
-
6930,
a
c
c
r
e
dit
e
d
F
ir
s
t
G
r
a
de
by
Ke
me
nr
is
tekdikti
,
De
c
r
e
e
No:
21/E
/KP
T
/2018
DO
I
:
10.
12928/
T
E
L
KO
M
NI
KA
.
v18i4.
12957
2235
Jou
r
n
al
h
omepage
:
ht
tp:
//
jour
nal.
uad
.
ac
.
id/
index
.
php/T
E
L
K
O
M
N
I
K
A
GWO
-
b
ase
d
e
st
imat
io
n
of
i
n
p
u
t
-
o
u
t
p
u
t
p
ar
a
m
e
t
e
r
s
of
t
h
e
r
m
al
p
ow
e
r
p
la
n
t
s
Os
e
a
Z
e
b
u
a
1
,
I
M
ad
e
Gi
n
ar
s
a
2
,
I
M
ad
e
Ar
i
Nr
a
r
t
h
a
3
1
D
e
p
art
me
n
t
o
f
E
l
ec
t
ri
ca
l
E
n
g
i
n
eer
i
n
g
,
U
n
i
v
er
s
i
t
y
o
f
L
a
mp
u
n
g
,
In
d
o
n
es
i
a
2,
3
D
e
p
art
me
n
t
o
f
E
l
ec
t
ri
ca
l
E
n
g
i
n
eer
i
n
g
,
U
n
i
v
er
s
i
t
y
o
f
Mat
aram
,
In
d
o
n
e
s
i
a
Ar
t
icle
I
n
f
o
AB
S
T
RA
CT
A
r
ti
c
le
h
is
tor
y
:
R
e
c
e
ived
Apr
18
,
2019
R
e
vis
e
d
Apr
11
,
2020
Ac
c
e
pted
M
a
y
1
,
2020
T
h
e
fu
e
l
co
s
t
cu
r
v
e
o
f
t
h
ermal
g
e
n
erat
o
rs
w
as
v
er
y
i
mp
o
r
t
an
t
i
n
t
h
e
cal
c
u
l
a
t
i
o
n
o
f
eco
n
o
mi
c
d
i
s
p
at
c
h
an
d
o
p
t
i
ma
l
p
o
w
er
fl
o
w
.
T
em
p
er
at
u
re
an
d
ag
i
n
g
c
o
u
l
d
mak
e
ch
an
g
es
t
o
f
u
el
c
o
s
t
cu
rv
e
s
o
c
u
rv
e
e
s
t
i
mat
i
o
n
n
ee
d
t
o
b
e
d
o
n
e
p
eri
o
d
i
cal
l
y
.
T
h
e
accu
racy
o
f
t
h
e
c
u
rv
e
p
aram
et
er
s
es
t
i
ma
t
i
o
n
s
t
ro
n
g
l
y
affect
ed
t
h
e
cal
cu
l
at
i
o
n
o
f
t
h
e
d
i
s
p
a
t
ch
.
T
h
i
s
p
ap
er
ai
m
s
t
o
es
t
i
ma
t
e
t
h
e
fu
e
l
co
s
t
cu
r
v
e
p
aram
e
t
ers
b
y
u
s
i
n
g
t
h
e
g
rey
w
o
l
f
o
p
t
i
m
i
zer
met
h
o
d
.
T
h
e
p
ro
b
l
em
o
f
cu
rv
e
p
aramet
er
es
t
i
ma
t
i
o
n
w
a
s
mad
e
as
an
o
p
t
i
m
i
za
t
i
o
n
p
ro
b
l
em.
T
h
e
o
b
j
ec
t
i
v
e
fu
n
ct
i
o
n
t
o
b
e
mi
n
i
m
i
zed
w
a
s
t
h
e
t
o
t
a
l
n
u
mb
er
o
f
ab
s
o
l
u
t
e
erro
r
o
r
t
h
e
d
i
fferen
ce
b
e
t
w
ee
n
t
h
e
act
u
a
l
v
al
u
e
an
d
t
h
e
es
t
i
m
at
ed
v
a
l
u
e
o
f
t
h
e
fu
el
co
s
t
fu
n
ct
i
o
n
.
T
h
e
es
t
i
mat
e
d
v
al
u
es
o
f
p
aramet
er
t
h
at
p
ro
d
u
ce
t
h
e
s
ma
l
l
e
s
t
t
o
t
a
l
ab
s
o
l
u
t
e
erro
r
w
er
e
t
h
e
v
al
u
es
o
f
f
i
n
a
l
s
o
l
u
t
i
o
n
.
T
h
e
s
i
mu
l
at
i
o
n
res
u
l
t
s
s
h
o
w
e
d
t
h
a
t
p
arame
t
er
es
t
i
mat
i
o
n
u
s
i
n
g
g
ray
w
o
l
f
o
p
t
i
mi
zer
me
t
h
o
d
fu
r
t
h
er
m
i
n
i
mi
ze
d
t
h
e
v
al
u
e
o
f
o
b
j
ect
i
v
e
f
u
n
c
t
i
o
n
.
B
y
u
s
i
n
g
t
h
ree
mo
d
el
s
o
f
fu
e
l
co
s
t
c
u
rv
e
a
n
d
g
i
v
en
t
es
t
d
a
t
a,
p
arame
t
er
es
t
i
ma
t
i
o
n
u
s
i
n
g
g
re
y
w
o
l
f
o
p
t
i
mi
zer
me
t
h
o
d
p
ro
d
u
ce
d
t
h
e
b
et
t
er
es
t
i
ma
t
i
o
n
res
u
l
t
s
t
h
a
n
t
h
o
s
e
es
t
i
m
at
i
o
n
res
u
l
t
s
o
b
t
ai
n
ed
u
s
i
n
g
l
eas
t
s
q
u
are
err
o
r,
p
art
i
cl
e
s
w
arm
o
p
t
i
m
i
zat
i
o
n
,
g
en
et
i
c
al
g
o
r
i
t
h
m,
art
i
f
i
ci
a
l
b
ee
co
l
o
n
y
an
d
cu
c
k
o
o
s
earc
h
met
h
o
d
s
.
K
e
y
w
o
r
d
s
:
F
ue
l
c
os
t
c
ur
ve
Gr
e
y
wolf
op
ti
mi
z
e
r
I
nput
-
output
pa
r
a
mete
r
s
P
a
r
a
mete
r
e
s
ti
mation
Th
i
s
i
s
a
n
o
p
en
a
c
ces
s
a
r
t
i
c
l
e
u
n
d
e
r
t
h
e
CC
B
Y
-
SA
l
i
ce
n
s
e
.
C
or
r
e
s
pon
din
g
A
u
th
or
:
Os
e
a
Z
e
bua
,
Unive
r
s
it
y
of
L
a
mpung,
P
r
of
.
S
umantr
i
B
r
ojonegor
o
S
t
.
No
.
1
B
a
nda
r
L
a
mp
ung
35145,
I
ndone
s
ia.
E
mail:
os
e
a
.
z
e
bua
@e
ng.
unil
a
.
a
c
.
id
1.
I
NT
RODU
C
T
I
ON
T
he
pla
nning
a
nd
ope
r
a
ti
on
of
the
powe
r
s
ys
tem
r
e
quir
e
s
a
n
e
c
onomi
c
dis
pa
tch
r
e
view
.
One
im
po
r
tant
f
a
c
tor
in
s
olvi
ng
e
c
onomi
c
dis
pa
tch
p
r
oblems
is
th
e
f
ue
l
c
os
t
c
u
r
ve
o
f
ther
mal
ge
ne
r
a
tor
s
.
T
he
f
ue
l
c
os
t
c
ur
ve
f
unc
ti
on
o
r
the
he
a
t
c
ha
r
a
c
ter
is
ti
c
c
u
r
ve
e
xp
r
e
s
s
e
s
t
he
input
-
output
r
e
lations
hip
o
f
a
ther
mal
g
e
ne
r
a
tor
.
T
his
f
ue
l
c
os
t
f
unc
ti
on
is
in
f
luenc
e
d
by
the
tempe
r
a
tur
e
a
nd
a
ging
o
f
the
ge
ne
r
a
to
r
un
it
s
a
nd
a
f
f
e
c
ts
t
he
s
ha
pe
of
f
ue
l
c
os
t
c
ur
ve
,
s
o
the
e
s
ti
mating
the
f
ue
l
c
os
t
c
ur
ve
ne
e
ds
to
be
e
va
luate
d
pe
r
iod
ica
l
ly
[
1
]
.
An
a
c
c
ur
a
te
e
s
ti
mation
of
ther
mal
uni
t
input
-
output
c
ur
ve
c
oe
f
f
icie
nts
is
im
por
tant
f
o
r
s
olvi
ng
e
c
onomi
c
dis
pa
tch
or
opti
mal
powe
r
f
low
pr
oblems
.
T
he
a
c
c
ur
a
c
y
of
the
e
s
ti
mate
d
c
oe
f
f
icie
nts
a
f
f
e
c
ts
the
f
inal
a
c
c
ur
a
c
y
of
the
dis
pa
tch
pr
oc
e
s
s
.
F
ue
l
c
o
s
t
f
unc
t
ions
c
a
n
be
r
e
pr
e
s
e
nted
by
mathe
matica
l
models
.
S
e
ve
r
a
l
mathe
matica
l
models
ha
ve
be
e
n
made
,
but
in
ge
ne
r
a
l,
ther
e
a
r
e
two
main
models
f
or
r
e
pr
e
s
e
nti
ng
f
ue
l
c
os
ts
f
unc
ti
on,
i
.
e
.
s
moot
h
model
a
nd
non
-
s
moot
h
model
.
S
e
ve
r
a
l
methods
ha
ve
be
e
n
p
r
opos
e
d
a
nd
im
ple
mente
d
to
s
olve
e
s
ti
mation
pr
oblems
in
powe
r
s
ys
tems
including
e
s
ti
mation
of
f
ue
l
c
os
t
c
ur
ve
of
ther
mal
ge
ne
r
a
tor
.
S
ome
of
thes
e
tec
hniques
a
r
e
ba
s
e
d
on
s
tatic
e
s
ti
mation
a
nd
dyna
mi
c
e
s
ti
mation
tec
hnique.
S
e
ve
r
a
l
s
tatic
e
s
ti
mation
tec
hniqu
e
s
,
s
uc
h
a
s
lea
s
t
s
qua
r
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
4
,
Augus
t
2020
:
2235
-
224
4
2236
e
r
r
or
(
L
S
E
)
,
Ga
us
s
-
Ne
wton,
B
a
r
d
a
lgor
it
h
m,
M
a
r
qua
r
dt
a
lgor
it
hm
da
n
P
owe
ll
r
e
g
r
e
s
s
ion
[
2
]
,
li
ne
a
r
r
e
gr
e
s
s
ion
[
3]
a
nd
l
inea
r
s
e
que
nti
a
l
r
e
gr
e
s
s
ion
te
c
hnique
[
4]
,
lea
s
t
a
bs
olut
e
va
lue
[
5]
,
a
nd
Gr
a
m
-
S
c
hmi
dt
or
thonor
maliza
ti
on
[
6
]
ha
ve
be
e
n
p
r
opos
e
d
a
nd
im
pleme
nted
in
e
s
ti
mating
the
f
ue
l
c
os
t
c
ur
ve
pa
r
a
mete
r
s
.
M
os
t
of
thes
e
e
s
ti
mation
tec
hniques
c
a
n
im
pr
ov
e
c
ompu
tational
e
f
f
icie
nc
y
a
nd
nu
mer
ica
l
s
tabili
ty,
but
the
r
e
s
ult
ing
e
r
r
or
s
a
r
e
s
ti
ll
lar
ge
a
nd
r
e
duc
e
the
a
c
c
ur
a
c
y
of
th
e
e
s
ti
mation
p
r
oc
e
ss.
Ka
lm
a
n
f
il
ter
is
one
of
the
dyna
mi
c
e
s
ti
mation
tec
hniques
whic
h
ha
ve
the
a
dva
ntage
of
be
ing
a
ble
to
upda
te
the
f
ue
l
c
os
t
c
ur
ve
pa
r
a
mete
r
e
s
ti
mation
us
ing
ne
w
mea
s
ur
e
ment
da
ta.
T
he
dis
a
dva
ntage
of
thi
s
tec
hnique,
a
s
we
ll
a
s
other
dyna
m
ic
f
il
ter
s,
is
tha
t
it
r
e
quir
e
s
lar
ge
da
ta
to
a
c
hieve
a
be
tt
e
r
s
olu
ti
on
[
7
-
9
].
M
e
ta
-
h
e
ur
is
ti
c
opti
mi
z
a
ti
on
methods
ha
ve
be
c
ome
popula
r
to
s
olve
many
op
ti
mi
z
a
ti
on
pr
ob
lems
in
many
f
ields
of
s
tudy.
E
volut
ionar
y
a
lgor
it
hm
-
ba
s
e
d
meta
he
ur
is
ti
c
methods
s
uc
h
a
s
a
r
ti
f
icia
l
ne
ur
a
l
ne
twor
ks
(
AN
N)
,
ge
ne
ti
c
a
lgor
it
hm
(
GA
)
a
nd
Di
f
f
e
r
e
nti
a
l
e
volut
ion
(
DE
)
c
a
n
s
olve
op
ti
mi
z
a
ti
on
p
r
oblems
with
non
mat
he
matica
l
model
f
unc
ti
on
a
nd
many
non
-
s
moot
h
opti
mi
z
a
ti
on
pr
oblems
with
non
-
c
onve
x
a
nd
dis
c
onti
nue
s
f
unc
ti
on.
One
of
the
d
r
a
wba
c
ks
us
ing
AN
N
-
ba
s
e
d
methods
is
the
huge
a
mount
of
da
ta
r
e
qui
r
e
d
f
o
r
ne
twor
k
tr
a
ini
ng,
whic
h
may
not
be
a
va
il
a
ble
in
s
ome
c
a
s
e
s
[
1
0]
.
T
he
GA
method
ha
s
be
e
n
us
e
d
to
e
s
ti
mate
pa
r
a
mete
r
s
of
a
s
moot
h
a
nd
non
-
s
moot
h
f
ue
l
c
os
t
c
ur
ve
bu
t
the
r
e
s
ult
ing
e
s
ti
mation
e
r
r
or
s
ti
ll
l
a
r
ge
[
11
]
.
T
he
mo
r
e
a
c
c
ur
a
te
r
e
s
ult
s
of
pa
r
a
mete
r
e
s
ti
mation
with
s
moot
h
a
n
d
non
-
s
moot
h
f
ue
l
c
os
t
c
ur
ve
s
ha
ve
be
e
n
p
r
op
os
e
d
a
nd
im
pleme
nted
us
ing
the
DE
method
[
12]
a
nd
im
pr
o
ve
d
DE
method
[
13]
.
M
e
tahe
ur
is
ti
c
methods
ba
s
e
d
on
s
wa
r
m
int
e
ll
ige
nc
e
s
uc
h
a
s
pa
r
t
icle
s
wa
r
m
opti
mi
z
a
ti
on
(
P
S
O)
,
a
r
ti
f
icia
l
be
e
c
olony
(
AB
C
)
,
a
nd
c
uc
koo
s
e
a
r
c
h
(
C
S
)
a
r
e
mo
r
e
r
obus
t
a
nd
e
a
s
e
s
of
us
e
a
ls
o
c
a
n
s
olve
opti
mi
z
a
ti
on
pr
ob
lem
with
many
types
of
objec
ti
v
e
f
unc
ti
on
with
s
mall
da
ta.
All
of
thes
e
methods
h
a
d
be
e
n
a
lr
e
a
dy
us
e
d
to
s
uc
c
e
s
f
ull
y
s
olve
many
opt
im
iz
a
ti
on
pr
oblems
in
powe
r
s
ys
tems
[
14
-
16]
.
I
n
e
s
ti
mating
the
pa
r
a
mete
r
s
of
the
f
ue
l
c
os
t
c
ur
ve
,
the
AB
C
met
hod
[
17]
is
mor
e
a
c
c
ur
a
te
than
the
P
S
O
method
[
18
,
19
]
a
nd
the
C
S
method
[
20]
,
with
a
s
maller
e
s
ti
mation
e
r
r
or
.
Gr
e
y
wolf
o
pti
mi
z
e
r
(
GW
O)
is
one
of
meta
he
ur
is
ti
c
opti
mi
z
a
ti
on
methods
ba
s
e
d
on
the
pr
e
y
hunti
ng
m
e
c
ha
nis
m
of
a
gr
oup
of
g
r
e
y
w
olf
.
T
he
va
r
ious
opti
mi
z
a
ti
on
pr
oblems
in
powe
r
s
ys
tems
ha
ve
be
e
n
s
olved
by
the
GW
O
method
a
nd
pr
ovided
be
tt
e
r
r
e
s
ult
s
than
thos
e
r
e
s
ult
s
ob
taine
d
us
ing
s
ome
other
opti
mi
z
a
ti
on
methods
[
21
-
23]
.
T
he
main
objec
ti
ve
of
thi
s
pa
pe
r
is
int
r
oduc
ing
a
ne
w
method
ba
s
e
d
on
gr
e
y
wolf
opti
mi
z
e
r
f
o
r
e
s
ti
mating
input
-
output
pa
r
a
mete
r
s
of
ther
mal
ge
ne
r
a
tor
unit
.
GW
O
is
r
e
latively
ne
w
method
ba
s
e
d
on
s
wa
r
m
i
ntelli
ge
nc
e
a
nd
a
lr
e
a
dy
ha
ve
be
tt
e
r
f
inal
s
olut
ion
c
ompar
e
d
to
P
S
O.
I
n
thi
s
pa
pe
r
,
e
s
ti
mation
of
inp
ut
-
output
pa
r
a
me
ter
s
of
f
ue
l
c
os
t
c
ur
ve
is
f
or
mul
a
ted
a
s
a
n
opti
mi
z
a
ti
on
pr
oblem.
T
he
main
goa
l
o
f
thi
s
wo
r
ks
is
to
mi
nim
ize
tot
a
l
a
bs
olut
e
e
r
r
or
of
e
s
ti
mat
e
d
f
ue
l
c
os
t
f
unc
ti
on
.
GW
O
is
us
e
d
to
f
ind
the
pa
r
a
mete
r
s
o
f
f
ue
l
c
os
t
c
ur
ve
a
nd
dif
f
e
r
e
nt
s
tudy
c
a
s
e
s
a
r
e
pr
e
s
e
nted
to
va
l
idate
the
pr
opos
e
d
a
ppr
oa
c
h.
T
his
pa
pe
r
is
or
ga
nize
d
a
s
f
oll
ows
:
s
e
c
ti
on
2
is
ge
n
e
r
a
l
ove
r
view
of
gr
e
y
wol
f
opti
m
ize
r
.
S
e
c
ti
on
3
is
r
e
s
e
a
r
c
h
method,
whic
h
c
ons
is
t
of
modeling
the
f
u
e
l
c
os
t
c
ur
ve
a
nd
e
s
ti
mating
input
-
output
pa
r
a
mete
r
of
f
ue
l
c
os
t
c
ur
ve
.
S
e
c
ti
on
4
is
r
e
s
ult
s
a
nd
a
na
lys
is
,
whic
h
c
ons
is
t
of
s
im
ulation
r
e
s
ult
s
of
e
s
ti
mating
pa
r
a
met
e
r
s
us
ing
GW
O
f
or
e
a
c
h
c
a
s
e
with
th
r
e
e
ther
m
a
l
ge
ne
r
a
tor
s
with
dif
f
e
r
e
nt
f
ue
l
types
.
2.
GRE
Y
WOL
F
OP
T
I
M
I
Z
E
R
(
GWO
)
Gr
e
y
-
W
olf
Optim
ize
r
(
GW
O)
is
a
r
e
latively
ne
w
meta
he
ur
is
ti
c
a
lgor
it
hm
that
f
ir
s
t
int
r
oduc
e
d
by
S
.
M
ir
jalil
i
e
t
a
l
.
[
24]
.
GW
O
mi
mi
c
s
the
lea
de
r
s
hip
h
ier
a
r
c
hy
a
nd
hunti
ng
mec
ha
nis
m
of
gr
e
y
wolve
s
i
n
na
tur
e
.
Us
ing
the
hier
a
r
c
hy
of
wolve
s
,
GW
O
im
pleme
nts
thr
e
e
main
s
teps
of
hunti
ng
,
i
.
e
.
s
e
a
r
c
hing,
e
nc
ir
c
li
ng
a
nd
a
tt
a
c
king
pr
e
y.
T
he
r
e
a
r
e
f
our
types
of
wol
f
s
,
i
.
e
.
a
lpha,
be
ta,
de
lt
a
a
nd
omega
f
or
s
im
ulating
the
hier
a
r
c
hy
of
lea
de
r
s
hip.
T
his
h
ier
a
r
c
hy
inf
luenc
e
s
the
f
inal
s
olut
i
on
in
hunti
ng
pr
e
y
a
nd
in
thi
s
a
lg
or
it
hm,
a
lpha
is
c
o
ns
ider
e
d
to
be
a
be
s
t
s
olut
ion
,
f
o
ll
owe
d
by
be
ta
,
de
lt
a
a
nd
o
mega
.
E
nc
ir
c
li
ng
p
r
e
y
pr
oc
e
s
s
c
a
n
be
de
s
c
r
ibed
in
e
qua
ti
on
a
s
f
oll
ows
:
⃗
⃗
=
|
⋅
(
)
−
(
)
|
(
1)
(
+
1
)
=
(
)
−
⋅
⃗
⃗
(
2)
whe
r
e
t
is
c
ur
r
e
nt
it
e
r
a
ti
on,
is
pos
it
ion
ve
c
tor
o
f
gr
e
y
wolf
,
is
pos
it
ion
ve
c
tor
of
pr
e
y
a
nd
a
nd
a
r
e
c
oe
f
f
icie
nts
ve
c
tor
that
c
a
lcula
ted
by
f
oll
owing
e
q
ua
ti
ons
:
=
2
⋅
1
−
(
3)
=
2
⋅
2
(
4)
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
GW
O
-
bas
e
d
e
s
ti
mation
of
input
-
output
par
ame
ter
s
of
ther
mal
pow
e
r
plant
(
Os
e
a
Z
e
bua
)
2237
whe
r
e
1
a
nd
2
a
r
e
r
a
ndom
ve
c
tor
s
be
twe
e
n
0
a
nd
1
a
nd
is
s
e
t
de
c
r
e
a
s
e
li
ne
a
r
ly
f
r
om
2
to
0
du
r
ing
it
e
r
a
ti
on
pr
oc
e
s
s
.
D
ur
ing
hunti
ng
pr
oc
e
s
s
,
thr
e
e
be
s
t
s
olut
ions
obtaine
d
s
o
f
a
r
a
r
e
s
a
ve
d
a
nd
the
other
s
e
a
r
c
h
a
ge
nts
(
including
omega
)
upda
te
their
pos
it
ions
a
c
c
or
ding
to
pos
it
ion
of
the
be
s
t
s
e
a
r
c
h
a
ge
nts
.
T
he
s
c
or
e
a
nd
pos
it
ion
of
th
r
e
e
s
e
a
r
c
h
a
ge
nts
(
i.
e
.
a
lpha,
be
ta,
a
nd
de
lt
a
)
is
upda
ted
us
ing
i
n
(
5
-
7)
,
r
e
s
pe
c
ti
ve
ly:
⃗
⃗
=
|
1
⋅
−
|
(
5)
⃗
⃗
=
|
2
⋅
−
|
(
6)
⃗
⃗
=
|
3
⋅
−
|
(
7)
T
he
pos
it
ion
ve
c
tor
of
p
r
e
y
with
r
e
s
pe
c
t
to
a
lpha,
be
ta
a
nd
de
lt
a
wolve
s
is
c
a
lcula
ted
u
s
ing
in
(
8
-
10
)
,
r
e
s
pe
c
ti
ve
ly.
T
he
be
s
t
pos
it
ion
o
f
pr
e
y
in
the
ne
xt
it
e
r
a
ti
on
is
c
a
lcula
ted
by
taking
a
ve
r
a
ge
va
lue
s
of
pr
e
y
pos
it
ion
with
r
e
s
pe
c
t
to
a
lpha,
be
ta
a
nd
de
lt
a
wolv
e
s
a
s
wr
it
ten
in
(
11
)
.
1
=
−
1
⋅
(
8)
2
=
−
2
⋅
(
9)
3
=
−
3
⋅
(
10)
(
+
1
)
=
⃗
1
+
⃗
2
+
⃗
3
3
(
11)
T
he
a
bil
it
y
of
s
e
a
r
c
hing
a
nd
a
tt
a
c
king
pr
e
y
of
g
r
e
y
wolf
s
r
e
pr
e
s
e
nt
the
a
bil
it
y
of
e
xplor
a
ti
on
a
nd
e
xpl
oit
a
ti
on
of
thi
s
a
lgor
i
thm
.
T
he
s
e
a
ll
a
r
e
identif
ied
by
va
lues
o
f
A
,
whe
r
e
A<
1
is
a
tt
a
c
king
a
nd
A>
1
is
s
e
a
r
c
hing.
3.
RE
S
E
AR
CH
M
E
T
HO
D
3.
1.
M
od
e
li
n
g
of
t
h
e
f
u
e
l
c
os
t
c
u
r
ve
T
he
f
ue
l
c
os
t
c
u
r
ve
of
ther
mal
ge
ne
r
a
to
r
c
a
n
be
e
xpr
e
s
s
e
d
a
s
a
n
input
-
output
r
e
lations
hip,
whic
h
is
be
twe
e
n
the
tot
a
l
c
os
t
pe
r
hour
or
the
tot
a
l
a
moun
t
of
e
ne
r
gy
us
e
d
pe
r
hou
r
a
nd
output
of
a
c
ti
ve
powe
r
.
I
n
thi
s
s
tudy,
the
f
ue
l
c
os
t
c
ur
ve
is
c
ons
ider
e
d
to
be
a
s
moot
h
c
ur
ve
model.
T
he
f
ue
l
c
os
t
c
ur
ve
f
o
r
the
ther
mal
ge
ne
r
a
tor
unit
n
a
s
a
f
unc
ti
on
of
output
a
c
ti
ve
powe
r
c
a
n
be
modele
d
by
a
polynom
ial
f
unc
ti
on
whic
h
e
xpr
e
s
s
e
d
in
the
f
ol
lowing
f
o
r
m:
(
)
=
0
+
∑
+
=
1
,
2
,
.
.
.
,
=
1
(
12)
whe
r
e
F
n
is
the
f
ue
l
c
os
t
f
u
nc
ti
on
of
n
th
ge
ne
r
a
tor
,
P
n
is
a
c
ti
ve
powe
r
ge
ne
r
a
ted
by
the
n
th
ther
mal
g
e
ne
r
a
tor
,
a
on
a
nd
a
mn
a
r
e
the
n
th
ge
ne
r
a
tor
c
ur
ve
c
oe
f
f
icie
nts
,
r
n
is
e
r
r
or
a
s
s
oc
iate
d
with
the
n
th
e
qua
ti
on,
N
is
nu
mber
of
ther
mal
ge
ne
r
a
tor
s
,
a
nd
L
is
e
qua
ti
on
or
de
r
.
I
n
thi
s
s
tud
y,
ther
e
a
r
e
thr
e
e
models
f
or
r
e
pr
e
s
e
nti
n
g
f
ue
l
c
os
t
f
unc
ti
on:
M
ode
l
1.
F
i
r
s
t
or
de
r
polynom
ial
model
or
li
ne
a
r
m
ode
l.
I
n
thi
s
c
a
s
e
,
(
12
)
will
be
in
th
e
f
or
m:
(
)
=
0
+
1
+
(
13)
M
ode
l
2.
S
e
c
ond
o
r
de
r
polynom
ial
model
or
qua
d
r
a
ti
c
model.
I
n
thi
s
c
a
s
e
,
in
(
12)
will
be
in
the
f
o
r
m:
(
)
=
0
+
1
+
2
2
+
(
14)
M
ode
l
3.
T
h
ir
d
o
r
de
r
polyno
mi
a
l
model
or
c
ubic
model:
I
n
thi
s
c
a
s
e
,
in
(
12)
will
be
in
the
f
o
r
m:
(
)
=
0
+
1
+
2
2
+
3
3
+
(
15)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
4
,
Augus
t
2020
:
2235
-
224
4
2238
All
thr
e
e
-
models
a
r
e
c
ons
ider
e
d
a
s
a
dis
c
r
e
te
s
ys
tem
a
nd
in
s
tate
s
pa
c
e
f
or
m
c
a
n
be
wr
it
ten
a
s
:
=
(
,
)
+
(
16)
whe
r
e
Z
n
is
a
f
ue
l
c
os
t
ve
c
tor
f
or
the
n
th
ge
ne
r
a
tor
,
X
n
a
r
e
pa
r
a
mete
r
ve
c
tor
to
be
e
s
ti
mate
d
(
a
0
,
a
1
,
a
2
,
a
3
)
f
or
n
th
ge
ne
r
a
tor
,
R
n
is
e
r
r
or
ve
c
tor
a
s
s
oc
iate
d
with
Z
n
.
T
he
n
,
a
s
s
oc
iate
d
e
r
r
or
with
e
a
c
h
mea
s
ur
e
ment
c
a
n
be
c
a
lcula
ted
a
s
:
=
(
)
−
(
)
(
17)
T
he
pr
oblem
is
f
or
mul
a
ted
a
s
to
f
ind
a
n
e
s
ti
mate
f
or
pa
r
a
mete
r
ve
c
to
r
X
that
mi
nim
ize
e
r
r
or
ve
c
tor
R
.
3.
2.
E
s
t
im
at
io
n
of
f
u
e
l
c
os
t
c
u
r
ve
p
ar
a
m
e
t
e
r
u
s
i
n
g
gr
e
y
wolf
op
t
i
m
ize
r
(
GWO
)
E
s
ti
mation
o
f
inpu
t
-
output
pa
r
a
mete
r
s
of
f
ue
l
c
os
t
c
ur
ve
us
ing
GW
O
is
pe
r
f
or
med
a
s
a
n
opti
mi
z
a
ti
on
pr
oblem.
T
he
ob
j
e
c
ti
ve
f
unc
ti
on
to
be
mi
nim
ize
d
is
s
um
of
a
bs
olut
e
e
r
r
or
be
twe
e
n
a
c
tual
c
os
t
a
nd
e
s
ti
mate
d
c
os
t.
T
he
objec
ti
ve
f
unc
ti
on
o
f
n
th
ge
ne
r
a
tor
is
the
s
um
of
a
bs
olut
e
e
r
r
o
r
o
f
(
17
)
a
nd
c
a
n
be
wr
it
ten
a
s
:
,
=
∑
|
,
(
)
−
,
(
)
|
=
1
(
18)
whe
r
e
k
is
ve
c
tor
of
inp
ut
da
ta
whic
h
c
ons
is
ts
of
e
ne
r
gy
us
e
d
in
GJ
/h
a
nd
it
s
c
or
r
e
s
ponding
a
c
ti
ve
powe
r
ou
tput
in
M
W
,
a
nd
M
is
numbe
r
of
tot
a
l
da
ta.
Numbe
r
of
s
e
a
r
c
h
dim
e
ns
ion
de
pe
nds
on
c
ur
ve
m
ode
l.
F
or
li
ne
a
r
model
,
the
s
e
a
r
c
h
dim
e
ns
ion
is
2,
f
or
qua
dr
a
ti
c
model,
the
s
e
a
r
c
h
d
im
e
ns
ion
is
3
a
nd
f
or
c
ubic
model,
the
s
e
a
r
c
h
dim
e
ns
ion
is
4
.
P
os
it
io
n
of
e
a
c
h
s
e
a
r
c
h
a
ge
nt
is
e
va
luate
d
e
a
c
h
it
e
r
a
ti
on
to
f
ind
the
va
lue
of
objec
ti
ve
f
unc
ti
on
a
nd
the
e
s
ti
mate
d
va
lu
e
of
f
ue
l
c
os
t.
T
hr
e
e
be
s
t
va
lues
o
f
f
it
ne
s
s
a
r
e
s
a
ve
d
a
s
s
c
or
e
va
lue,
i
.
e
.
a
lpha
s
c
or
e
,
be
ta
s
c
or
e
a
nd
de
lt
a
s
c
or
e
.
P
os
it
ion
of
e
a
c
h
s
e
a
r
c
h
a
ge
nt
is
then
up
da
ted
in
ne
xt
it
e
r
a
ti
on.
T
he
s
e
pr
oc
e
dur
e
s
a
r
e
pe
r
f
or
med
un
ti
l
the
m
a
xim
um
it
e
r
a
ti
on
is
r
e
a
c
he
d.
T
he
be
s
t
s
olut
ion
a
nd
be
s
t
p
os
it
ion
obtaine
d
a
t
maximum
i
ter
a
ti
on
is
c
ons
ider
e
d
to
be
the
f
inal
s
olut
ion
.
T
he
a
lgor
it
hm
f
o
r
f
indi
ng
e
s
ti
mate
d
va
lues
o
f
f
ue
l
c
os
t
c
ur
ve
pa
r
a
mete
r
s
us
ing
GW
O
is
e
xplaine
d
s
tep
by
s
tep
a
s
f
oll
ows
:
−
I
nit
ialize
the
number
of
e
a
c
h
s
e
a
r
c
h
a
ge
nt
,
the
ma
xim
um
number
o
f
i
ter
a
ti
ons
,
a
nd
the
uppe
r
a
nd
lo
we
r
l
im
it
of
the
s
e
a
r
c
h
f
o
r
pa
r
a
mete
r
s
.
S
c
o
r
e
s
a
nd
ini
ti
a
l
pos
it
ions
of
e
a
c
h
s
e
a
r
c
h
a
ge
nt,
a
lpha,
be
ta
a
nd
de
lt
a
a
r
e
s
e
t
to
inf
ini
ty
f
or
thi
s
mi
nim
iza
ti
on
pr
oblem
.
−
S
e
t
the
number
of
s
e
a
r
c
h
dim
e
ns
ions
a
c
c
or
ding
to
the
c
os
t
c
ur
ve
model
a
nd
the
ini
t
ial
it
e
r
a
ti
on
.
−
C
a
lcula
te
the
e
s
ti
mate
d
c
os
t
va
lue,
F
e
s
t
i
m
a
t
e
d
,
f
or
e
a
c
h
s
e
a
r
c
h
a
ge
nt.
−
C
a
lcula
te
the
tot
a
l
a
bs
olut
e
e
r
r
or
f
o
r
e
a
c
h
s
e
a
r
c
h
a
ge
nt
a
c
c
or
ding
to
(
18)
.
−
I
f
the
a
bs
olut
e
e
r
r
or
is
s
maller
than
the
pr
e
vious
v
a
lue,
then
the
s
c
or
e
a
nd
pos
it
ions
of
e
a
c
h
s
e
a
r
c
h
a
ge
nt
a
r
e
s
tor
e
d.
I
f
the
a
bs
olut
e
e
r
r
o
r
is
gr
e
a
ter
than
the
pr
e
vious
va
lue,
th
e
n
the
s
c
or
e
a
nd
pos
it
ions
of
e
a
c
h
s
e
a
r
c
h
a
ge
nt
a
r
e
de
lete
d.
−
C
onti
nue
s
teps
3,
4
a
nd
5
f
o
r
the
ne
xt
s
e
a
r
c
h
a
ge
nt
unti
l
the
numbe
r
of
s
e
a
r
c
h
a
ge
nts
is
r
e
a
c
he
d.
−
Upda
te
the
pos
it
ion
of
e
a
c
h
s
e
a
r
c
h
a
ge
nt
a
c
c
or
ding
to
(
8
-
11)
.
−
C
onti
nue
s
tep
3
to
7
f
or
the
ne
xt
it
e
r
a
ti
on
.
−
I
f
the
it
e
r
a
ti
on
ha
s
r
e
a
c
he
d
the
maximum
it
e
r
a
ti
o
n,
the
p
r
oc
e
dur
e
is
s
topped.
P
r
in
t
the
r
e
s
ult
s
of
a
l
pha
s
c
or
e
s
a
nd
a
lpha
pos
it
ion
X
.
T
his
pr
oc
e
dur
e
is
r
e
pe
a
ted
f
or
other
ge
ne
r
a
tor
s
a
nd
other
f
ue
l
c
os
t
c
ur
ve
models
.
T
he
a
lgor
it
hm
d
e
s
c
r
ibed
a
bov
e
is
il
lus
tr
a
ted
by
the
f
low
c
ha
r
t
a
s
s
hown
in
F
igur
e
1.
4.
RE
S
UL
T
S
A
ND
AN
AL
YSI
S
T
he
a
lgor
i
thm
is
de
s
c
r
ibed
a
bov
e
a
nd
il
lus
tr
a
ted
b
y
the
f
lowc
ha
r
t
in
F
igur
e
1
is
im
pleme
nted
us
ing
M
AT
L
AB
.
S
im
ulation
us
ing
GW
O
is
pe
r
f
o
r
med
u
s
ing
pr
a
c
ti
c
a
l
da
ta
f
r
om
[
2
]
.
T
he
s
e
da
ta
a
r
e
us
e
d
to
e
s
ti
mate
pa
r
a
mete
r
s
of
thr
e
e
model
o
f
f
ue
l
c
os
t
c
u
r
ve
.
F
or
e
a
c
h
c
a
s
e
,
s
im
ulation
is
pe
r
f
or
me
d
f
o
r
1000
it
e
r
a
ti
ons
with
the
lowe
r
bound
a
nd
upp
e
r
bound
va
lues
of
e
a
c
h
pa
r
a
mete
r
a
r
e
s
e
t
be
twe
e
n
-
200
a
nd
200
.
Numbe
r
of
s
e
a
r
c
h
a
ge
nts
us
e
d
in
thi
s
s
im
ulation
is
20.
S
im
ulation
is
pe
r
f
or
med
by
d
if
f
e
r
e
nt
tr
ials
a
nd
50
be
s
t
tr
ials
a
r
e
s
a
ve
d
f
or
e
a
c
h
c
a
s
e
.
T
he
r
e
s
ult
s
obtaine
d
f
o
r
e
a
c
h
c
a
s
e
a
r
e
th
e
n
c
ompa
r
e
d
to
the
r
e
s
ult
s
obtaine
d
us
ing
other
me
thods
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
GW
O
-
bas
e
d
e
s
ti
mation
of
input
-
output
par
ame
ter
s
of
ther
mal
pow
e
r
plant
(
Os
e
a
Z
e
bua
)
2239
S
t
a
r
t
S
t
a
r
t
I
n
i
t
i
l
i
a
z
e
n
u
m
b
e
r
o
f
s
e
a
r
c
h
a
g
e
n
t
s
,
m
a
x
i
m
u
m
i
t
e
r
a
t
i
o
n
s
(
i
t
e
r
m
a
x
)
,
s
e
a
r
c
h
b
o
u
n
d
a
r
i
e
s
,
a
l
p
h
a
,
b
e
t
a
,
d
e
l
t
a
s
c
o
r
e
s
a
n
d
p
o
s
i
t
i
o
n
s
I
n
i
t
i
l
i
a
z
e
n
u
m
b
e
r
o
f
s
e
a
r
c
h
a
g
e
n
t
s
,
m
a
x
i
m
u
m
i
t
e
r
a
t
i
o
n
s
(
i
t
e
r
m
a
x
)
,
s
e
a
r
c
h
b
o
u
n
d
a
r
i
e
s
,
a
l
p
h
a
,
b
e
t
a
,
d
e
l
t
a
s
c
o
r
e
s
a
n
d
p
o
s
i
t
i
o
n
s
S
e
t
n
u
m
b
e
r
o
f
s
e
a
r
c
h
d
i
m
e
n
s
i
o
n
,
i
n
i
t
i
a
l
i
t
e
r
a
t
i
o
n
i
t
e
r
=
0
S
e
t
n
u
m
b
e
r
o
f
s
e
a
r
c
h
d
i
m
e
n
s
i
o
n
,
i
n
i
t
i
a
l
i
t
e
r
a
t
i
o
n
i
t
e
r
=
0
i
t
e
r
<
i
t
e
r
m
a
x
?
i
t
e
r
<
i
t
e
r
m
a
x
?
C
a
l
c
u
l
a
t
e
e
s
t
i
m
a
t
e
d
v
a
l
u
e
o
f
f
u
e
l
c
o
s
t
f
o
r
e
a
c
h
s
e
a
r
c
h
a
g
e
n
t
s
C
a
l
c
u
l
a
t
e
e
s
t
i
m
a
t
e
d
v
a
l
u
e
o
f
f
u
e
l
c
o
s
t
f
o
r
e
a
c
h
s
e
a
r
c
h
a
g
e
n
t
s
C
a
l
c
u
l
a
t
e
f
i
t
n
e
s
s
o
f
e
a
c
h
s
e
a
r
c
h
a
g
e
n
t
s
C
a
l
c
u
l
a
t
e
f
i
t
n
e
s
s
o
f
e
a
c
h
s
e
a
r
c
h
a
g
e
n
t
s
F
o
r
i
=
1
:
n
u
m
b
e
r
o
f
s
e
a
r
c
h
a
g
e
n
t
s
F
o
r
i
=
1
:
n
u
m
b
e
r
o
f
s
e
a
r
c
h
a
g
e
n
t
s
U
p
d
a
t
e
a
l
p
h
a
,
b
e
t
a
a
n
d
d
e
l
t
a
s
c
o
r
e
s
U
p
d
a
t
e
a
l
p
h
a
,
b
e
t
a
a
n
d
d
e
l
t
a
s
c
o
r
e
s
F
i
t
n
e
s
s
s
m
a
l
l
e
r
?
F
i
t
n
e
s
s
s
m
a
l
l
e
r
?
D
i
s
c
a
r
d
r
e
s
u
l
t
D
i
s
c
a
r
d
r
e
s
u
l
t
F
o
r
j
=
1
:
n
u
m
b
e
r
o
f
s
e
a
r
c
h
a
g
e
n
t
s
F
o
r
j
=
1
:
n
u
m
b
e
r
o
f
s
e
a
r
c
h
a
g
e
n
t
s
U
p
d
a
t
e
p
o
s
i
t
i
o
n
s
o
f
s
e
a
r
c
h
a
g
e
n
t
s
i
n
c
l
u
d
i
n
g
o
m
e
g
a
U
p
d
a
t
e
p
o
s
i
t
i
o
n
s
o
f
s
e
a
r
c
h
a
g
e
n
t
s
i
n
c
l
u
d
i
n
g
o
m
e
g
a
C
a
l
c
u
l
a
t
e
t
h
e
b
e
s
t
a
l
p
h
a
s
c
o
r
e
a
n
d
p
o
s
i
t
i
o
n
s
f
o
r
c
u
r
r
e
n
t
i
t
e
r
a
t
i
o
n
C
a
l
c
u
l
a
t
e
t
h
e
b
e
s
t
a
l
p
h
a
s
c
o
r
e
a
n
d
p
o
s
i
t
i
o
n
s
f
o
r
c
u
r
r
e
n
t
i
t
e
r
a
t
i
o
n
E
n
d
E
n
d
Y
e
s
N
o
Y
e
s
N
o
P
r
i
n
t
t
h
e
b
e
s
t
a
l
p
h
a
s
c
o
r
e
a
n
d
p
o
s
i
t
i
o
n
s
P
r
i
n
t
t
h
e
b
e
s
t
a
l
p
h
a
s
c
o
r
e
a
n
d
p
o
s
i
t
i
o
n
s
F
igur
e
1.
F
low
c
ha
r
t
o
f
the
GW
O
a
lgor
it
hm
4.
1.
Cas
e
s
t
u
d
y
1
I
n
thi
s
c
a
s
e
,
li
ne
a
r
model
of
f
ue
l
c
os
t
f
unc
ti
on
de
s
c
r
ibed
in
(
13)
is
us
e
d
f
or
e
s
ti
mating
two
pa
r
a
mete
r
c
oe
f
f
icie
nts
(
a
0
a
nd
a
1
)
o
f
the
r
mal
ge
ne
r
a
tor
c
os
t
c
ur
ve
.
T
he
e
s
ti
mation
r
e
s
ult
s
us
ing
GW
O
a
r
e
c
ompar
e
d
to
the
r
e
s
ult
s
obtaine
d
us
ing
the
L
S
E
,
P
S
O,
AB
C
a
nd
C
S
methods
.
T
he
e
s
ti
mate
d
c
oe
f
f
icie
nt
va
lue
,
the
e
s
ti
mate
d
ge
ne
r
a
tor
c
os
t
f
unc
ti
on
va
lues
a
nd
the
e
s
ti
mate
d
e
r
r
or
va
lues
us
ing
the
GW
O
a
nd
the
e
s
ti
mation
r
e
s
ult
s
us
ing
L
S
E
,
P
S
O,
AB
C
a
nd
C
S
a
r
e
s
hown
in
T
a
ble
1
,
T
a
ble
2
a
nd
T
a
ble
3
,
r
e
s
pe
c
ti
ve
ly.
As
s
e
e
n
f
r
om
T
a
ble
3,
e
s
ti
mation
of
f
ue
l
c
os
t
c
ur
ve
pa
r
a
mete
r
us
ing
GW
O
c
a
n
mo
r
e
mi
nim
ize
tot
a
l
a
bs
olut
e
e
r
r
o
r
va
lues
c
ompar
e
d
to
thos
e
r
e
s
u
lt
s
obtaine
d
us
ing
f
our
other
methods
.
T
he
GW
O
method
a
c
hieve
s
c
onve
r
ge
nc
e
to
the
mi
nim
um
va
lue
of
the
objec
ti
ve
f
unc
ti
on
f
or
mo
r
e
than
20
0
it
e
r
a
ti
ons
in
the
c
a
s
e
of
ge
ne
r
a
tor
unit
1,
a
s
s
hown
in
F
igur
e
2.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
4
,
Augus
t
2020
:
2235
-
224
4
2240
T
a
ble
1.
E
s
ti
mate
d
pa
r
a
mete
r
s
f
or
c
a
s
e
s
tudy
1
(
li
n
e
a
r
model
)
U
ni
t
C
oe
f
f
ic
ie
nt
s
M
e
th
ods
L
S
E
PSO
A
B
C
CS
G
W
O
1 (
C
oa
l)
a
0
63.236
63.236
45.2120
43.566
45.2008
a
1
10.170
10.190
10.5600
10.597
10.5600
2 (
O
il
)
a
0
66.160
66.001
47.6520
62.559
47.6006
a
1
10.631
10.570
11.0310
10.655
11.0300
3 (
G
a
s
)
a
0
66.700
66.002
48.3990
62.889
48.4004
a
1
10.830
10.780
11.2210
10.860
11.2200
T
a
ble
2.
E
s
ti
mate
d
f
ue
l
c
os
t
f
unc
ti
on
f
o
r
c
a
s
e
s
tudy
1
(
l
inea
r
model
)
U
ni
t
P
(
M
W
)
F
a
c
tu
a
l
(
G
J
/h
)
F
e
s
ti
m
a
ted
(
G
J
/h
)
L
S
E
PSO
A
B
C
CS
G
W
O
1 (
c
oa
l)
10
176.62
164.936
161.
905
150.812
149.532
150.800
20
256.40
266.636
263.803
256.412
255.498
256.400
30
361.50
368.338
365.702
362.012
361.464
361.999
40
467.60
470.036
467.600
467.612
467.430
467.599
50
579.50
571.736
569.498
573.212
573.396
573.199
2 (
oi
l)
10
184.75
1
72.470
171.701
157.962
169.109
157.900
20
268.20
278.780
277.400
268.272
275.659
268.200
30
377.70
385.090
383.100
378.582
382.209
378.500
40
488.80
491.400
488.800
488.892
488.759
488.800
50
606.00
597.710
594.499
599.202
595.309
599.101
3 (
ga
s
)
10
187.20
175.000
173.802
160.609
171.498
160.600
20
272.80
283.300
281.601
272.819
280.097
272.800
30
384.30
391.600
389.401
385.029
388.696
385.000
40
497.20
499.900
497.200
497.239
497.295
497.200
50
616.50
608.200
604.999
609.499
605.894
609.40
0
T
a
ble
3.
E
s
ti
mate
d
e
r
r
or
f
or
c
a
s
e
s
tudy
1
(
li
ne
a
r
m
ode
l)
U
ni
t
P
(
M
W
)
F
a
c
tu
a
l
(
G
J
/h
)
E
r
r
or
=
|
−
|
L
S
E
PSO
A
B
C
CS
G
W
O
1 c
oa
l)
10
176.62
11.684
14.715
25.808
27.088
25.820
20
256.40
10.236
7.403
0.012
0.902
0.000
30
361.50
6.83
6
4.202
0.512
0.036
0.500
40
467.60
2.436
0.000
0.012
0.170
0.001
50
579.50
7.764
10.002
6.288
6.104
6.301
e
r
r
or
38.956
36.322
32.632
34.301
32.622
2 (
oi
l)
10
184.75
12.280
13.049
26.788
15.641
26.850
20
268.20
10.580
9.200
0.072
7.459
0.000
3
0
377.70
7.390
5.400
0.882
4.509
0.800
40
488.80
2.600
0.000
0.092
0.041
0.000
50
606.00
8.290
11.501
6.798
10.691
6.900
e
r
r
or
41.140
39.151
34.632
38.341
34.550
3 (
ga
s
)
10
187.20
12.200
13.398
26.591
15.702
26.600
20
272.80
10.500
8.801
0.019
7.297
0.000
30
384.30
7.300
5.101
0.729
4.396
0.700
40
497.20
2.700
0.000
0.039
0.095
0.000
50
616.50
8.300
11.501
7.051
10.606
7.100
e
r
r
or
41.000
38.801
34.429
38.096
34.400
F
igur
e
2.
C
onve
r
ge
nc
e
c
ha
r
a
c
ter
is
ti
c
f
or
c
a
s
e
s
tudy
1
(
l
inea
r
mo
de
l)
of
ge
ne
r
a
tor
un
it
1
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
GW
O
-
bas
e
d
e
s
ti
mation
of
input
-
output
par
ame
ter
s
of
ther
mal
pow
e
r
plant
(
Os
e
a
Z
e
bua
)
2241
4.
2.
Cas
e
s
t
u
d
y
2
I
n
thi
s
c
a
s
e
,
th
r
e
e
pa
r
a
mete
r
s
c
oe
f
f
icie
nts
(
a
0
,
a
1
a
n
d
a
2
)
of
f
ue
l
c
os
t
f
unc
ti
on
with
qua
dr
a
ti
c
model
a
s
de
s
c
r
ibed
in
(
14)
a
r
e
e
s
ti
mate
d.
T
he
s
a
me
ther
ma
l
powe
r
plants
da
ta
in
c
a
s
e
s
tudy
1
a
r
e
us
e
d
in
t
his
c
a
s
e
.
T
he
r
e
s
ult
s
obtaine
d
us
ing
GW
O
a
r
e
c
ompar
e
d
to
the
r
e
s
ult
s
obtaine
d
us
ing
L
S
E
,
P
S
O,
AB
C
,
C
S
,
GA
a
nd
DE
methods
.
T
he
e
s
ti
mate
d
pa
r
a
mete
r
c
oe
f
f
icie
nts
us
ing
GW
O
a
nd
the
other
methods
a
r
e
s
hown
in
T
a
ble
4.
T
he
e
s
ti
mate
d
va
lue
of
f
ue
l
c
os
t
f
unc
ti
on
a
nd
th
e
to
tal
a
bs
olut
e
e
r
r
or
be
twe
e
n
a
c
tual
va
lue
a
nd
e
s
ti
mate
d
va
lue
of
f
ue
l
c
os
t
f
unc
ti
on
obtaine
d
f
r
om
e
a
c
h
method
a
r
e
s
hown
in
T
a
ble
5
a
nd
T
a
ble
6
,
r
e
s
pe
c
ti
ve
ly.
As
s
e
e
n
in
T
a
ble
6,
the
tot
a
l
a
bs
olut
e
e
r
r
or
s
obtaine
d
us
ing
GW
O
a
r
e
s
maller
than
thos
e
r
e
s
ult
s
obtaine
d
us
ing
the
P
S
O,
L
S
E
,
GA
,
AB
C
a
nd
C
S
methods
,
but
s
ti
ll
s
li
ghtl
y
lar
ge
r
than
the
tot
a
l
a
bs
olut
e
e
r
r
or
s
obtaine
d
us
ing
the
DE
method.
I
t
is
c
lea
r
that
the
GW
O
method
pr
oduc
e
s
a
be
tt
e
r
s
olut
ion
than
the
s
olut
ion
obtaine
d
us
ing
t
he
L
S
E
,
P
S
O,
GA
,
AB
C
a
nd
C
S
methods
,
a
lt
hough
it
is
s
ti
ll
les
s
a
c
c
ur
a
te
than
the
s
olut
ion
obtaine
d
us
ing
the
DE
method.
I
n
th
i
s
c
a
s
e
,
f
o
r
ge
ne
r
a
tor
unit
1,
GW
O
method
r
e
quir
e
s
mo
r
e
than
900
it
e
r
a
ti
ons
to
a
c
hieve
c
onve
r
ge
nc
e
to
the
be
s
t
mi
nim
um
va
lue
of
the
ob
j
e
c
ti
ve
f
unc
ti
on
a
s
s
hown
in
F
igur
e
3
.
T
a
ble
4.
E
s
ti
mate
d
pa
r
a
mete
r
s
f
or
c
a
s
e
s
tudy
2
(
qu
a
dr
a
ti
c
model
)
U
ni
t
C
oe
f
f
ic
ie
n
ts
M
e
th
ods
L
S
E
PSO
GA
A
B
C
CS
DE
G
W
O
1
(
C
oa
l)
a
0
95.856
96.279
100.3937
96.6046
96.540
96.6000
96.5936
a
1
7.374
7.592
6.9761
7.5874
7.575
7.5880
7
.5879
a
2
0.047
0.042
0.0533
0.0414
0.042
0.0414
0.0414
2 (
O
il
)
a
0
100.710
101.000
107.1688
101.5360
100.887
101.53125
101.5306
a
1
7.670
7.800
7.7235
7.8779
7.890
7.8800
7.8800
a
2
0.049
0.046
0.0467
0.0442
0.045
0.044188
0.0442
3
(
G
a
s
)
a
0
101.100
10
2.00
116.3854
101.8179
99.239
101.8125
101.8110
a
1
7.881
7.900
6.7342
8.0991
8.138
8.1000
8.1002
a
2
0.049
0.048
0.0667
0.0439
0.045
0.043875
0.0439
T
a
ble
5.
E
s
ti
mate
d
f
ue
l
c
os
t
f
unc
ti
on
f
o
r
c
a
s
e
s
tudy
2
(
qua
d
r
a
ti
c
mo
de
l)
U
ni
t
P
(
M
W
)
F
a
c
tu
a
l
(
G
J
/h
)
F
e
s
ti
m
a
ted
(
G
J
/h
)
L
S
E
GA
PSO
A
B
C
CS
DE
G
W
O
1 c
oa
l)
10
176.62
174.252
175.485
176.358
176.619
176.480
N
/A
176.613
20
256.40
261.968
261.236
264.765
264.913
264.800
N
/A
264.914
30
361.50
359.004
357.647
361.500
361.487
361.500
N
/A
361.497
40
467.60
465.360
464.718
466.562
466.341
466.580
N
/A
466.360
50
579.50
581.036
582.449
579.952
579.475
580.040
N
/A
579.504
2 (
oi
l)
10
184.75
182.346
184.295
183.600
184.735
184.248
N
/A
184.750
20
268.20
273.862
272.449
275.400
276.774
276.525
N
/A
276.806
30
377.70
375.258
373.089
376.400
377.653
377.718
N
/A
377.700
40
488.80
486.534
485.729
486.600
487.372
487.827
N
/A
487.431
50
606.00
607.690
610.369
606.000
605.931
606.851
N
/A
606.000
3 (
ga
s
)
10
187.20
184.824
188.648
185.780
187.799
185.145
N
/A
185.7
80
20
272.80
278.368
277.749
279.121
281.360
280.111
N
/A
279.121
30
384.30
381.732
378.441
382.022
384.301
384.137
N
/A
382.022
40
497.20
494.916
492.473
494.484
496.022
497.223
N
/A
494.484
50
616.50
617.920
619.845
616.507
616.523
619.369
N
/A
616.5
07
T
a
ble
6.
E
s
ti
mate
d
e
r
r
or
f
or
c
a
s
e
s
tudy
2
(
qua
dr
a
ti
c
model
)
U
ni
t
P
(
M
W
)
F
a
c
tu
a
l
(
G
J
/h
)
E
r
r
or
=
|
F
F
|
e
s
t
i
m
a
t
e
d
e
c
t
u
a
l
−
L
S
E
GA
PSO
A
B
C
CS
DE
G
W
O
1 c
oa
l)
10
176.62
2.368
1.135
0.262
0.001
0.140
0.000
0.067
20
256.40
5.568
4.836
8.365
8.513
8.400
8.
5200
8.514
30
361.50
2.496
3.853
0.000
0.013
0.000
0.000
0.004
40
467.60
2.240
2.882
1.038
1.259
1.020
1.240
1.240
50
579.50
1.536
2.949
0.452
0.025
0.540
0.000
0.004
e
r
r
or
14.208
15.655
10.117
9.810
10.100
9.760
9.769
2 (
oi
l)
10
184.75
2.404
0.
455
1.150
0.015
0.502
0.000
0.000
20
268.20
5.662
4.249
7.200
8.574
8.325
8.606
8.606
30
377.70
2.442
4.611
1.300
0.047
0.018
0.000
0.000
40
488.80
2.266
3.071
2.200
1.428
0.973
1.368
1.369
50
606.00
1.690
4.369
0.000
0.069
0.851
0.001
0.000
e
r
r
or
14.464
16.755
11.850
10.133
10.669
9.975
9.975
3 (
ga
s
)
10
187.20
2.376
1.448
1.420
0.599
2.055
0.000
0.000
20
272.80
5.568
4.949
6.321
8.560
7.311
8.563
8.562
30
384.30
2.568
5.859
2.278
0.001
0.163
0.000
0.001
40
497.20
2.284
4.727
2.716
1.178
0.023
1.187
1.188
50
616.50
1.420
3.345
0.007
0.023
2.869
0.000
0.000
e
r
r
or
14.216
20.328
12.741
10.361
12.421
9.750
9.751
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
4
,
Augus
t
2020
:
2235
-
224
4
2242
F
igur
e
3.
C
onve
r
ge
nc
e
c
ha
r
a
c
ter
is
ti
c
f
or
c
a
s
e
s
tudy
2
(
qua
d
r
a
ti
c
model)
of
ge
ne
r
a
tor
uni
t
1
4.
3.
Cas
e
s
t
u
d
y
3
I
n
thi
s
c
a
s
e
,
f
our
pa
r
a
mete
r
s
(a
0
,
a
1
,
a
2
a
nd
a
3
)
o
f
f
ue
l
c
os
t
f
unc
ti
on
us
ing
c
ubic
model
a
s
de
s
c
r
ibed
in
(
15)
a
r
e
e
s
ti
mate
d.
T
he
ther
mal
ge
ne
r
a
tor
da
t
a
us
e
d
in
thi
s
c
a
s
e
a
r
e
the
s
a
me
a
s
the
da
ta
us
e
d
in
c
a
s
e
s
tudy
1
a
nd
c
a
s
e
s
tudy
2.
T
he
r
e
s
ult
s
obtaine
d
us
in
g
GW
O
a
r
e
c
ompar
e
d
to
the
r
e
s
ult
s
obtaine
d
us
i
ng
L
S
E
,
P
S
O,
AB
C
,
a
nd
DE
methods
.
T
he
r
e
s
ult
s
of
e
s
ti
m
a
ted
pa
r
a
mete
r
of
f
ue
l
c
os
t
c
ur
ve
s
obtaine
d
by
us
ing
GW
O
method
a
nd
the
L
S
E
,
P
S
O,
AB
C
,
DE
methods
a
r
e
s
hown
in
T
a
ble
7
.
T
a
ble
7
.
E
s
ti
mate
d
pa
r
a
mete
r
s
f
or
c
a
s
e
s
tudy
3
(
c
u
bic
model)
U
ni
t
C
oe
f
f
ic
ie
nt
s
M
e
th
ods
L
S
E
PSO
A
B
C
DE
G
W
O
1 (
C
oa
l)
a
0
123.180
120.241
124.5362
127.0667
127.3003
a
1
3.535
3.939.
3.4859
3.1187
3.0794
a
2
0.193
0.184
0.1872
0.1999
0.2021
a
3
-
0.002
-
0.002
-
0.0015
-
0.0016
-
0.0017
2 (
O
il
)
a
0
128.640
130.278
129.2351
132.5000
132.7809
a
1
3.746
3.542
3.4859
3.3325
3.2672
a
2
0.199
0.200
0.1872
0.2059
0.2094
a
3
-
0.002
-
0.002
-
0.0015
-
0.00166
-
0.0017
3 (
G
a
s
)
a
0
128.400
128.376
126.0143
132.3333
131.0319
a
1
4.046
4.146
3.8044
3.6250
3.8076
a
2
0.195
0.188
0.1896
0.2024
0.1962
a
3
-
0.002
-
0.002
-
0.0015
-
0.0016
-
0.0016
T
he
e
s
ti
mation
r
e
s
ult
s
of
f
ue
l
c
os
t
f
unc
ti
ons
,
a
bs
olut
e
e
r
r
or
s
a
nd
tot
a
l
a
bs
olut
e
e
r
r
o
r
s
e
it
he
r
us
ing
the
GW
O
method
or
us
ing
the
L
S
E
,
P
S
O,
AB
C
a
nd
D
E
methods
a
r
e
s
ho
wn
in
T
a
ble
8
a
nd
T
a
ble
9,
r
e
s
p
e
c
ti
ve
ly.
As
s
e
e
n
f
r
om
T
a
ble
9,
e
s
ti
mating
pa
r
a
mete
r
us
ing
the
GW
O
c
a
n
pr
oduc
e
tot
a
l
a
bs
olut
e
e
r
r
or
s
s
m
a
ll
e
r
than
thos
e
obtaine
d
us
ing
the
L
S
E
,
P
S
O,
a
nd
AB
C
met
hods
.
B
ut
the
tot
a
l
number
o
f
a
bs
olut
e
e
r
r
or
s
obtai
ne
d
us
ing
GW
O
method
is
s
ti
ll
g
r
e
a
ter
than
the
r
e
s
ult
s
obtain
e
d
us
ing
the
DE
method.
T
a
ble
8.
E
s
ti
mat
e
d
f
ue
l
c
os
t
f
unc
ti
on
f
o
r
c
a
s
e
s
tudy
3
(
c
ubic
model)
U
ni
t
P
(
M
W
)
F
a
c
tu
a
l
(
G
J
/h
)
F
e
s
ti
m
a
ted
(
G
J
/h
)
L
S
E
PSO
A
B
C
DE
G
W
O
1
(
c
oa
l)
10
176.62
174.227
176.806
176.615
N
/A
176.648
20
256.40
258.274
260.557
257.134
N
/A
256.478
30
361.50
359.721
361.951
357.093
N
/A
356.854
40
467.60
470.968
471.446
467.492
N
/A
467.840
50
579.50
582.415
579.500
579.331
N
/A
579.500
2 (
oi
l)
10
184.75
184.301
184.076
184.739
N
/A
184.68
6
20
268.20
269.562
268.200
269.163
N
/A
268.218
30
377.70
374.223
373.010
373.507
N
/A
373.119
40
488.80
488.084
488.863
488.771
N
/A
489.129
50
606.00
600.945
606.119
605.955
N
/A
605.991
3 (
ga
s
)
10
187.20
186.804
187.101
187.188
N
/A
187.166
20
27
2.80
274.688
274.326
274.632
N
/A
273.162
30
384.30
382.452
381.000
380.561
N
/A
379.638
40
497.20
500.496
498.074
497.170
N
/A
497.211
50
616.50
619.220
616.500
616.659
N
/A
616.500
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
GW
O
-
bas
e
d
e
s
ti
mation
of
input
-
output
par
ame
ter
s
of
ther
mal
pow
e
r
plant
(
Os
e
a
Z
e
bua
)
2243
T
a
ble
9.
E
s
ti
mate
d
e
r
r
or
f
or
c
a
s
e
s
tudy
3
(
c
ubic
m
ode
l)
U
ni
t
P
(
M
W
)
F
a
c
tu
a
l
(
G
J
/h
)
E
r
r
or
=
|
−
|
L
S
E
PSO
A
B
C
DE
G
W
O
1 c
oa
l)
10
176.62
0.393
0.186
0.0048
0.000
0.028
20
256.40
1.874
4.157
0.7342
0.000
0.078
30
361.50
1.779
0.451
4.4068
4.854
4.646
40
467.60
3.368
3.846
0.1078
0.002
0.240
50
579
.50
2.915
0.000
0.1688
0.004
0.000
e
r
r
or
10.329
8.641
5.422
4.860
4.992
2 (
oi
l)
10
184.75
0.449
0.674
0.0109
0.000
0.064
20
268.20
1.362
0.000
0.9631
0.000
0.018
30
377.70
3.477
4.690
4.1929
4.825
4.581
40
488.80
0.716
0.063
0.0289
0.000
0.329
50
606.00
5.005
0.119
0.0449
0.000
0.010
e
r
r
or
11.059
5.547
5.421
4.825
5.002
3 (
ga
s
)
10
187.20
0.396
0.099
0.0167
0.000
0.034
20
272.80
1.888
1.526
1.8323
0.000
0.362
30
384.30
1.848
3.300
3.7387
4.917
4.662
40
497.20
3.296
0.874
0.0297
0.000
0.011
50
616.50
2.720
0.000
0.159
0.000
0.000
e
r
r
or
10.148
5.799
5.777
4.917
5.069
T
he
c
onve
r
ge
nc
e
c
ha
r
a
c
ter
is
ti
c
of
s
im
ulation
f
or
g
e
ne
r
a
tor
unit
1
s
hows
that
GW
O
method
is
a
ble
to
a
c
hieve
opti
mal
f
it
ne
s
s
va
lues
in
mo
r
e
than
500
i
ter
a
ti
ons
a
s
s
hown
in
F
igu
r
e
4.
T
he
tot
a
l
number
a
bs
olut
e
e
r
r
or
s
f
or
thr
e
e
-
unit
ther
mal
ge
ne
r
a
tor
s
obtaine
d
with
thi
s
model
a
r
e
much
lowe
r
than
thos
e
obtaine
d
in
c
a
s
e
s
tudy
1
a
nd
2
.
T
h
is
mea
ns
that
the
thi
r
d
or
de
r
o
r
c
ubic
model
is
mor
e
s
uit
a
ble
f
or
r
e
p
r
e
s
e
nti
ng
f
ue
l
c
o
s
t
c
ur
ve
of
ther
mal
ge
ne
r
a
to
r
[
25]
.
F
r
om
the
r
e
s
ult
s
,
the
GW
O
-
ba
s
e
d
method
is
a
ble
to
mi
nim
ize
the
tot
a
l
number
of
a
bs
olut
e
e
r
r
or
s
be
tt
e
r
than
the
L
S
E
,
P
S
O,
AB
C
a
nd
C
S
methods
s
o
that
the
e
s
ti
mate
d
va
lue
of
the
f
ue
l
c
os
t
f
unc
ti
on
is
c
los
e
r
to
the
a
c
tual
v
a
lue
of
f
ue
l
c
os
t
f
unc
ti
on.
Although
the
tot
a
l
number
o
f
a
bs
olut
e
e
r
r
or
s
obtaine
d
is
s
ti
ll
gr
e
a
ter
than
that
va
lue
obtaine
d
us
ing
the
D
E
method
,
the
GW
O
method
c
a
n
be
the
one
o
f
the
be
s
t
opti
on
tool
s
f
o
r
e
s
ti
mating
the
pa
r
a
mete
r
of
f
ue
l
c
os
t
c
ur
ve
of
the
r
m
a
l
ge
ne
r
a
ti
ng
unit
s
.
T
he
GW
O
method
take
s
a
bout
1
.
5
s
e
c
onds
to
c
onve
r
ge
with
the
c
ur
r
e
nt
s
im
ulation
pa
r
a
mete
r
s
.
F
igur
e
4.
C
onve
r
ge
nc
e
c
ha
r
a
c
ter
is
ti
c
f
or
c
a
s
e
s
tudy
3
(
c
ubic
model)
o
f
ge
ne
r
a
to
r
unit
1
5.
CONC
L
USI
ON
E
s
ti
mation
o
f
the
input
-
output
c
ur
v
e
o
r
f
ue
l
c
os
t
c
u
r
ve
pa
r
a
mete
r
s
o
f
ther
mal
ge
ne
r
a
tor
us
ing
the
gr
e
y
wolf
opti
mi
z
e
r
(
GW
O)
method
is
pr
e
s
e
nted
in
thi
s
pa
pe
r
.
T
h
r
e
e
models
of
f
ue
l
c
os
t
c
ur
ve
s
with
thr
e
e
ther
mal
ge
ne
r
a
tor
s
with
dif
f
e
r
e
nt
f
ue
ls
type
ha
ve
be
e
n
tes
ted
us
ing
thi
s
method.
T
he
e
s
t
im
a
ted
pa
r
a
mete
r
is
obtaine
d
by
mi
nim
izing
the
tot
a
l
number
of
a
bs
olut
e
e
r
r
or
be
twe
e
n
the
a
c
tual
va
lue
a
nd
the
e
s
ti
mate
d
va
lue
of
the
ge
ne
r
a
tor
f
ue
l
c
os
t
f
unc
ti
on.
T
he
tes
t
r
e
s
ult
s
s
how
that
the
GW
O
method
is
mor
e
a
c
c
ur
a
te
f
or
e
s
ti
mating
pa
r
a
mete
r
of
the
input
-
output
c
ur
ve
of
ther
mal
g
e
ne
r
a
tor
unit
s
by
p
r
oduc
ing
s
maller
tot
a
l
a
bs
olut
e
e
r
r
or
s
c
ompar
e
d
to
thos
e
obtaine
d
us
ing
L
S
E
,
P
S
O
,
GA
,
AB
C
a
nd
C
S
methods
a
nd
s
li
ghtl
y
les
s
a
c
c
ur
a
te
c
ompar
e
d
to
thos
e
obtaine
d
us
ing
DE
met
hod.
RE
F
E
RE
NC
E
S
:
[1
]
D
ay
c
o
ck
C
.
,
D
e
s
j
ar
d
i
n
R
.
,
Fen
n
el
S.
,
“
G
en
era
t
i
o
n
Co
s
t
F
o
recas
t
i
n
g
U
s
i
n
g
O
n
-
l
i
n
e
T
h
erm
o
d
y
n
am
i
c
Mo
d
el
s
,”
E
l
ec
t
r
i
c
P
o
we
r
S
ys
t
em
R
es
e
a
r
c
h
,
v
o
l
.
1
,
p
p
.
1
-
9,
2
0
0
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N
:
1693
-
6930
T
E
L
KO
M
NI
KA
T
e
lec
omm
un
C
omput
E
l
C
ontr
o
l
,
Vol.
18
,
No
.
4
,
Augus
t
2020
:
2235
-
224
4
2244
[
2
]
El
-
H
a
w
a
r
y
M
.
E
.
,
M
a
n
s
o
u
r
S
.
Y.
,
“
P
e
r
f
o
r
m
a
n
c
e
E
v
a
l
u
a
t
i
o
n
o
f
P
a
r
a
m
e
t
e
r
E
s
t
i
m
a
t
i
o
n
A
l
g
o
r
i
t
h
m
s
f
o
r
E
c
o
n
o
m
i
c
O
p
e
r
a
t
i
o
n
o
f
P
o
w
e
r
S
y
s
t
e
m
s
,”
I
E
E
E
T
r
a
n
s
a
c
t
i
o
n
s
o
n
P
o
w
e
r
A
p
p
a
r
a
t
u
s
a
n
d
S
y
s
t
e
m
s
,
v
o
l
.
1
0
1
,
n
o
.
3
,
p
p
.
5
7
4
-
5
8
2
,
1
9
8
2
.
[3
]
El
-
Sh
i
b
i
n
i
M
.
,
O
s
man
Z
.
H.
,
“
A
N
o
v
el
T
ech
n
i
q
u
e
t
o
E
s
t
i
ma
t
e
t
h
e
Fu
e
l
Co
s
t
Fu
n
ct
i
o
n
s
fo
r
E
c
o
n
o
mi
c
O
p
era
t
i
o
n
o
f
Po
w
er
Sy
s
t
em
s
,”
In
t
e
r
n
a
t
i
o
n
a
l
Jo
u
r
n
a
l
o
f
E
l
ec
t
r
i
ca
l
P
o
wer
a
n
d
E
n
er
g
y
S
ys
t
em
s
,
v
o
l
.
11
,
n
o
.
2
,
p
p
.
1
0
9
-
1
1
4
,
1
9
8
9
.
[4
]
Ch
en
H
.
Y
.
K
.
,
Po
s
t
el
C
.
E.
,
“
On
-
l
i
n
e
Paramet
er
Id
e
n
t
i
fi
cat
i
o
n
o
f
In
p
u
t
-
o
u
t
p
u
t
Cu
rv
e
s
fo
r
T
h
erma
l
U
n
i
t
s
,”
IE
E
E
Tr
a
n
s
a
c
t
i
o
n
s
o
n
P
o
we
r
A
p
p
a
r
a
t
u
s
a
n
d
S
ys
t
em
s
,
”
v
o
l
.
1
,
p
p
.
2
2
1
-
2
2
4
,
1
9
8
6
.
[5
]
So
l
i
ma
n
S
.
A
.
,
E
mam
S
.
E
.
A
.
,
Ch
r
i
s
t
en
s
e
n
G
.
S.
,
“
O
p
t
i
mi
zat
i
o
n
o
f
t
h
e
O
p
t
i
mal
Co
eff
i
ci
e
n
t
s
o
f
N
o
n
-
m
o
n
o
t
i
c
al
l
y
In
creas
i
n
g
In
cremen
t
al
C
o
s
t
Cu
rv
e
s
,”
E
l
ect
r
i
c
P
o
wer
S
y
s
t
e
m
R
es
e
a
r
c
h
,
v
o
l
.
21
,
n
o
.
2
,
p
p
.
99
-
1
0
6
,
1
9
9
1
.
[6
]
L
i
an
g
Z
.
X
.
,
G
l
o
v
er
J
.
D.
,
“
Imp
ro
v
ed
Co
s
t
Fu
n
ct
i
o
n
fo
r
E
co
n
o
m
i
c
D
i
s
p
a
t
ch
Co
m
p
u
t
at
i
o
n
s
,
”
IE
E
E
Tr
a
n
s
a
ct
i
o
n
s
o
n
P
o
we
r
S
ys
t
em
s
,
v
o
l
.
6
,
n
o
.
2
,
p
p
.
8
2
1
-
8
2
9
,
1
9
9
1
.
[7
]
T
ay
l
o
r
F
.
J
.
,
H
u
an
g
C
.
H.
,
“
Recu
rs
i
v
e
es
t
i
ma
t
i
o
n
o
f
i
n
cre
men
t
a
l
co
s
t
cu
r
v
es
,”
Co
m
p
u
t
e
r
s
a
n
d
E
l
ect
r
i
c
a
l
E
n
g
i
n
ee
r
i
n
g
,
v
o
l
.
4
,
n
o
.
4
,
p
p
.
2
9
7
-
3
07
,
1
9
7
7
.
[8
]
So
l
i
ma
n
S
.
A
.
,
A
l
-
K
a
n
d
ar
i
A
.
M.
,
“
Ka
l
m
a
n
F
i
l
t
e
r
i
n
g
A
l
g
o
r
i
t
h
m
f
o
r
O
n
l
i
n
e
Id
en
t
i
f
i
c
a
t
i
o
n
o
f
I
n
p
u
t
-
o
u
t
p
u
t
Cu
r
ves
f
o
r
Th
er
m
a
l
P
o
we
r
P
l
a
n
t
,”
P
r
o
cee
d
i
n
g
s
8
th
M
ed
i
t
e
r
r
a
n
i
a
n
E
l
ec
t
r
o
t
ec
h
n
i
ca
l
Co
n
f
er
e
n
ce
,
p
p
.
1
5
8
8
-
1
5
9
3
,
1
9
9
6
.
[9
]
D
u
ra
n
-
Paz
J
.
I
.
,
Perez
-
H
i
d
al
g
o
F
.
,
D
u
ran
-
Mart
i
n
ez
M.
J.
,
“
Bad
D
at
a
D
et
ect
i
o
n
o
f
U
n
eq
u
al
Mag
n
i
t
u
d
es
i
n
St
at
e
E
s
t
i
ma
t
i
o
n
o
f
P
o
w
er
Sy
s
t
em
s
,”
IE
E
E
P
o
wer
E
n
g
i
n
ee
r
i
n
g
R
evi
ew
,
v
o
l
.
2
2
,
n
o
.
4
,
p
p
.
57
-
60
,
2
0
0
2
.
[1
0
]
A
t
t
av
i
ri
y
an
u
p
a
p
A
.
,
K
i
t
a
H
.
,
T
a
n
ak
a
E
.
,
H
a
s
eg
a
w
a
J
.
,
“
A
H
y
b
r
i
d
E
P
a
n
d
S
Q
P
fo
r
D
y
n
am
i
c
E
c
o
n
o
mi
c
D
i
s
p
at
c
h
w
i
t
h
N
o
n
-
Smo
o
t
h
F
u
el
Co
s
t
F
u
n
c
t
i
o
n
,”
IE
E
E
Tr
a
n
s
a
ct
i
o
n
s
o
n
P
o
we
r
S
y
s
t
e
m
s
,
v
o
l
.
17
,
n
o
.
2
,
p
p
.
4
1
1
-
4
1
6
,
2
0
0
2
.
[1
1
]
Al
-
K
an
d
ari
A
.
M
.
,
E
l
-
N
ag
g
ar
K
.
M.
,
“
A
G
en
e
t
i
c
-
b
as
e
d
A
l
g
o
ri
t
h
m
f
o
r
O
p
t
i
ma
l
E
s
t
i
mat
i
o
n
o
f
In
p
u
t
-
o
u
t
p
u
t
C
u
rv
e
Paramet
ers
o
f
T
h
erma
l
Po
w
er
P
l
an
t
s
,”
E
l
ect
r
i
c
a
l
E
n
g
i
n
e
er
i
n
g
,
v
o
l
.
89
,
n
o
.
8
,
p
p
.
5
8
5
-
5
9
0
,
2
0
0
7
.
[1
2
]
Say
ah
S
.
,
H
amo
u
d
a
A
.
,
“
N
o
v
e
l
A
p
p
l
i
ca
t
i
o
n
o
f
D
i
ffere
n
t
i
al
E
v
o
l
u
t
i
o
n
A
l
g
o
r
i
t
h
m
fo
r
E
s
t
i
mat
i
n
g
Fu
e
l
C
o
s
t
Fu
n
ct
i
o
n
o
f
T
h
erma
l
G
en
erat
i
n
g
U
n
i
t
s
,”
Th
i
r
d
W
o
r
l
d
Co
n
f
er
e
n
ce
o
n
Co
m
p
l
ex
S
y
s
t
e
m
s
.
M
a
r
a
kech
M
o
r
o
cco
,
p
p
.
1
-
9,
2
0
1
5
.
[1
3
]
Say
ah
S
.
,
H
amo
u
d
a
A
.
,
“
E
ffi
ci
e
n
t
Me
t
h
o
d
fo
r
E
s
t
i
ma
t
i
o
n
o
f
Smo
o
t
h
an
d
N
o
n
s
mo
o
t
h
Fu
e
l
Co
s
t
C
u
rv
e
f
o
r
T
h
er
mal
Po
w
er
Pl
a
n
t
s
,”
In
t
e
r
n
a
t
i
o
n
a
l
Tr
a
n
s
a
c
t
i
o
n
s
o
n
E
l
ec
t
r
i
ca
l
E
n
e
r
g
y
S
y
s
t
e
m
s
,
v
o
l
.
28
,
n
o
.
3
,
p
p
.
1
-
14
,
2
0
1
8
.
[1
4
]
A
l
Ras
h
i
d
i
M
.
R
.
,
E
l
-
H
a
w
ary
M
.
E.
,
“
A
Su
rv
ey
o
f
Part
i
c
l
e
Sw
arm
O
p
t
i
mi
za
t
i
o
n
A
p
p
l
i
cat
i
o
n
s
i
n
E
l
ect
r
i
c
Po
w
er
Sy
s
t
em
s
,”
IE
E
E
Tr
a
n
s
a
c
t
i
o
n
o
n
E
v
o
l
u
t
i
o
n
a
r
y
Co
m
p
u
t
a
t
i
o
n
,”
v
o
l
.
13
,
n
o
.
4
,
p
p
.
9
1
3
-
9
1
8
,
2
0
0
9
.
[1
5
]
H
emamal
i
n
i
S
.
,
Si
mo
n
S
.
P.
,
“
A
rt
i
f
i
ci
a
l
Bee
Co
l
o
n
y
A
l
g
o
ri
t
h
m
f
o
r
E
co
n
o
m
i
c
L
o
ad
D
i
s
p
a
t
ch
Pro
b
l
em
w
i
t
h
N
o
n
-
s
m
o
o
t
h
Co
s
t
Fu
n
ct
i
o
n
,”
E
l
ect
r
i
c
P
o
wer
C
o
m
p
o
n
en
t
s
a
n
d
S
y
s
t
e
m
s
,
v
o
l
.
38
,
n
o
.
10
,
p
p
.
7
8
6
-
8
0
3
,
2
0
1
0
.
[1
6
]
Bas
u
M
.
,
Ch
o
w
d
h
u
ry
A.
,
“
Cu
ck
o
o
Search
A
l
g
o
r
i
t
h
m
fo
r
E
co
n
o
mi
c
D
i
s
p
a
t
ch
,”
E
n
er
g
y
,
v
o
l
.
60
,
p
p
.
99
-
1
0
8
,
2
0
1
3
.
[1
7
]
So
n
mez
Y
.
,
“
E
s
t
i
ma
t
i
o
n
o
f
Fu
e
l
Co
s
t
C
u
rv
e
Parame
t
ers
fo
r
T
h
ermal
P
o
w
er
P
l
an
t
s
U
s
i
n
g
A
BC
A
l
g
o
r
i
t
h
m
,”
Tu
r
k
i
s
h
Jo
u
r
n
a
l
o
f
E
l
ect
r
i
c
a
l
E
n
g
i
n
ee
r
i
n
g
&
Co
m
p
u
t
e
r
S
c
i
en
ce
s
,
v
o
l
.
21
,
p
p
.
1
8
2
7
-
1
8
4
1
,
2
0
1
3
.
[1
8
]
El
-
N
ag
g
a
r
K
.
M
.
,
A
l
-
Ras
h
i
d
i
M
.
R
.
,
A
l
-
O
t
h
man
A
.
K.
,
“
E
s
t
i
ma
t
i
n
g
t
h
e
I
n
p
u
t
-
ou
t
p
u
t
Paramet
er
s
o
f
T
h
ermal
P
o
w
er
Pl
an
t
s
,”
E
n
er
g
y
Co
n
ve
r
s
i
o
n
a
n
d
M
a
n
a
g
e
m
en
t
,
v
o
l
.
50
,
n
o
.
7
,
p
p
.
1
7
6
7
-
1
7
7
2
,
2
0
0
9
.
[1
9
]
A
l
ra
s
h
i
d
i
M
.
R
.
,
E
l
-
N
a
g
g
ar
K
.
M
.
,
A
l
-
O
t
h
man
A
.
K.
,
“
Part
i
c
l
e
Sw
arm
O
p
t
i
m
i
zat
i
o
n
Bas
e
d
A
p
p
r
o
ach
f
o
r
E
s
t
i
mat
i
n
g
Fu
el
Co
s
t
Fu
n
ct
i
o
n
Paramet
er
s
o
f
T
h
ermal
Po
w
er
Pl
an
t
s
w
i
t
h
V
a
l
v
e
L
o
a
d
i
n
g
E
ffec
t
s
,”
E
l
ect
r
i
c
P
o
we
r
Co
m
p
o
n
en
t
s
a
n
d
S
ys
t
em
s
,
v
o
l
.
37
,
n
o
.
11
,
p
p
.
1
2
1
9
-
1
2
3
0
,
2
0
0
9
.
[2
0
]
A
l
Ras
h
i
d
i
M
.
R
.
,
E
l
-
N
ag
g
ar
K
.
M
.
,
A
l
H
aj
r
i
M
.
F.
,
“
Co
n
v
ex
a
n
d
N
o
n
-
co
n
v
e
x
H
eat
C
u
rv
e
Paramet
er
s
E
s
t
i
ma
t
i
o
n
U
s
i
n
g
C
u
ck
o
o
Searc
h
,”
Ar
a
b
J
S
c
i
E
n
g
.
,
v
o
l
.
40
,
n
o
.
3)
,
p
p
.
8
7
3
-
8
8
2
,
2
0
1
4
.
[2
1
]
Su
l
ai
m
an
M
.
H
.
,
Mu
s
t
affa
Z
.
,
Mo
h
ame
d
M
.
R
.
,
A
l
i
ma
n
O
.
,
“
U
s
i
n
g
t
h
e
G
ray
W
o
l
f
O
p
t
i
mi
zer
f
o
r
So
l
v
i
n
g
O
p
t
i
mal
React
i
v
e
Po
w
er
D
i
s
p
a
t
ch
Pro
b
l
em
,”
A
p
p
l
i
e
d
S
o
f
t
Co
m
p
u
t
i
n
g
,
v
o
l
.
32
,
p
p
.
2
8
6
-
2
9
2
,
2
0
1
5
.
[2
2
]
Su
l
t
a
n
a
U
.
,
K
h
a
i
ru
d
d
i
n
A
.
B
.
,
Mo
k
h
t
ar
A
.
S
.
,
Z
areen
N
.
,
Su
l
t
a
n
a
B.
,
“
G
rey
W
o
l
f
O
p
t
i
mi
zer
b
a
s
ed
Pl
aceme
n
t
an
d
Si
zi
n
g
o
f
Mu
l
t
i
p
l
e
D
i
s
t
ri
b
u
t
ed
G
en
era
t
i
o
n
i
n
t
h
e
D
i
s
t
r
i
b
u
t
i
o
n
S
y
s
t
em
,”
E
n
e
r
g
y
,
v
o
l
.
1
1
1
,
p
p
.
5
2
5
-
5
3
6
,
2
0
1
6
.
[2
3
]
Prad
h
a
n
M
.
,
Ro
y
P
.
K
.
,
Pal
T
.
,
“
G
rey
W
o
l
f
O
p
t
i
m
i
z
at
i
o
n
A
p
p
l
i
ed
t
o
E
c
o
n
o
mi
c
L
o
a
d
D
i
s
p
at
c
h
Pr
o
b
l
ems
,”
E
l
ec
t
r
i
ca
l
P
o
we
r
a
n
d
E
n
er
g
y
S
ys
t
em
s
,
v
o
l
.
83
,
p
p
.
3
2
5
-
3
3
4
,
2
0
1
6
.
[
2
4
]
M
i
r
j
a
l
i
l
i
S
.
,
M
i
r
j
a
l
i
l
i
S
.
M
.
,
L
e
w
i
s
A
.
,
“
G
r
e
y
W
o
l
f
O
p
t
i
m
i
z
e
r
,”
A
d
v
a
n
c
e
s
i
n
E
n
g
i
n
e
e
r
i
n
g
S
o
f
t
w
a
r
e
,
v
o
l
.
69
,
p
p
.
46
-
61
,
2
0
1
4
.
[2
5
]
Sh
o
u
l
t
s
R
.
R
.
,
Mead
M
.
M.
,
“
O
p
t
i
ma
l
E
s
t
i
ma
t
i
o
n
o
f
P
i
ecew
i
s
e
L
i
n
ear
I
n
cremen
t
al
C
o
s
t
C
u
rv
e
s
fo
r
E
D
C
,”
I
E
E
E
Tr
a
n
s
a
c
t
i
o
n
s
o
n
P
o
we
r
A
p
p
a
r
a
t
u
s
a
n
d
S
ys
t
em
s
,
V
o
l
.
PE
R
-
4
,
n
o
.
6
,
p
p
.
5
-
5
7
,
1
9
8
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.