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
.
5
,
Octo
ber
201
9
, pp.
3359
~3
365
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v9
i
5
.
pp
3359
-
33
65
3359
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Sh
or
t
-
t
erm
optimal hyd
ro
-
the
rm
al sche
du
lin
g using c
lus
tered
adapti
ve teachin
g learnin
g based
optimiz
ation
Surender
R
e
d
dy
Sa
lk
ut
i
Depa
rt
m
ent
o
f
R
ai
lro
ad
and
E
lect
ric
a
l
Eng
ineeri
n
g,
W
oosong
Uni
ver
sit
y
,
Dae
je
on
,
Republic
o
f
Kor
ea
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Dec
23
, 201
8
Re
vised
A
pr
3
,
201
9
Accepte
d
Apr
11
, 201
9
In
thi
s
pape
r
,
Cluste
red
Adapt
ive
Teac
h
ing
L
ea
rning
B
ase
d
Optimiza
tio
n
(CATLBO)
al
g
orit
hm
is
prop
osed
for
de
te
r
m
ini
ng
the
op
t
imal
hour
l
y
sche
dule
of
po
wer
gene
r
at
ion
in
a
h
y
dro
-
the
r
m
al
power
s
y
st
em.
In
th
e
proposed
appr
oa
ch,
a
m
ult
i
-
rese
r
voir
ca
sca
d
ed
hy
dro
-
el
e
ct
ri
c
s
y
s
te
m
with
a
non
-
li
ne
ar
re
la
t
i
onship
bet
wee
n
wate
r
dis
cha
rg
e
rate,
n
et
h
ea
d
and
power
gene
ra
ti
on
is
co
nsidere
d.
Constr
ai
nts
such
as
power
bal
an
ce
,
wa
te
r
balanc
e
,
rese
rvoir
vo
lume
l
imits
and
ope
rat
ion
li
m
it
s
of
h
y
dro
and
the
rm
al
p
la
nts
are
conside
red
.
Th
e
fea
sibi
li
t
y
and
eff
ective
n
ess
of
the
proposed
a
lgori
thm
is
demons
tra
te
d
th
rough
a
t
est
s
ystem,
and
the
r
esult
s
ar
e
comp
are
d
wi
th
exi
sting
conve
n
t
iona
l
and
e
volut
i
onar
y
a
lgori
thm
s.
Sim
ula
ti
on
r
esult
s
rev
e
al
s
tha
t
th
e
propose
d
CATLBO
al
gorit
hm
appe
ars
to
be
the
best
i
n
te
rm
s
of
conve
rge
n
ce spe
ed
and
opt
imal c
ost c
om
par
ed
wi
th
oth
er
t
ec
hn
iqu
es.
Ke
yw
or
d
s
:
E
voluti
onary a
lgorit
hm
s
G
ene
rati
on sc
he
du
li
ng
Hydro
-
the
rm
al
s
che
duli
ng
M
ulti
-
chain
r
es
ervoirs
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
:
Su
r
en
der Re
dd
y Sal
ku
ti
,
Dep
a
rtm
ent o
f R
ai
lroad
a
nd E
le
ct
rical
En
gi
ne
erin
g,
17
-
2,
Woos
on
g Un
i
ver
sit
y,
Jay
ang
-
do
ng, Do
ng
-
gu, Dae
je
on
-
34606, Re
public o
f K
or
e
a.
Em
a
il
: sur
en
de
r@wsu.ac
.
kr
1.
INTROD
U
CTION
Hydro
po
wer
pl
ants
are
m
ulti
-
pur
po
se
pr
oj
ec
ts,
w
hich
are
not
only
gen
e
rat
e
the
el
ect
rical
powe
r
but
al
so
re
spo
ns
ibl
e
f
or
t
he
fu
l
fill
m
ent
of
irri
gation
re
qu
irem
ents
of
nea
rb
y
zon
e
[
1].
S
hor
t
te
rm
hydr
o
-
therm
al
sche
du
li
ng
(S
T
-
HT
S)
dete
rm
i
nes
t
he
op
ti
m
a
l
powe
r
genera
ti
on
of
the
hy
dro
an
d
t
her
m
al
ge
ner
at
or
s
,
s
o
as
t
o
m
ini
m
iz
e the total
co
st
of the
rm
al
g
ener
at
or
s,
w
hile sat
isfy
ing
t
he
c
on
st
ra
ints o
f hydro
-
therm
al
p
ower
sy
stem
.
This
is
one
of
the
c
onstr
ai
ned
power
syst
e
m
op
tim
i
zat
ion
pro
ble
m
,
wh
ic
h
has
com
plex,
non
-
li
nea
r
char
act
e
risti
cs
with
var
i
ou
s
ty
pes
of
co
ns
trai
nts
incl
ud
i
ng
powe
r
balance
,
water
balance
,
ph
ysi
cal
li
m
it
a
ti
on
s
on
t
he
rese
rvoi
r
an
d
tur
bin
e
flow
rate,
wat
er
trans
port
delay
betwee
n
co
nnect
ed
reserv
oir
s,
an
d
loa
ding
lim
it
s
of
both
hy
dro
and
the
rm
al
plants
[
2].
I
n
ge
ner
al
,
the
ob
j
e
ct
ive
in
t
he
hy
dro
-
t
her
m
al
scheduli
ng
pro
ble
m
is
to
m
ini
m
iz
e
the
total
fu
el
cost
of
th
erm
al
gen
erati
ng
un
it
s
.
In
t
he
li
te
ratur
e
,
v
a
rio
us
cl
assic
al
m
et
ho
ds
are
dev
el
op
e
d
f
or
so
lvi
ng
this
prob
le
m
.
However,
these
m
et
ho
ds
ha
ve
dif
fi
culti
es
in
hand
li
ng
co
ns
trai
nts
li
ke
non
-
co
nvex
and
pro
hib
it
ed o
pe
rati
ng r
e
gions
.
Ba
ckgrou
nd
:
I
n
recent
ye
a
rs
,
m
et
a
-
heu
risti
c
opti
m
iz
at
ion
al
gorithm
s
ha
ve
be
e
n
e
xten
sively
us
e
d
because
to
thei
r
feasi
bili
ty
,
ver
sat
il
it
y
and
r
obus
t
ness
in
re
achin
g
the
global
opti
m
a
l
so
luti
on.
T
hese
in
cl
ud
e
Gen
et
ic
Algo
rithm
s
(G
A)
[
3],
Evo
l
ution
a
ry
Pr
og
ram
m
ing
(
EP)
[
4],
Partic
le
Sw
arm
Op
ti
m
iz
at
ion
(P
SO)
[5
]
,
Im
pr
ov
e
d
P
S
O
[
6],
Sim
ula
te
d
A
nn
eal
i
ng
(S
A
)
[7
]
,
E
voluti
onary
St
r
at
egy
(ES)
[
8]
,
et
c.
Re
fer
e
nce
[
9]
pro
po
ses
a
M
od
i
fied
See
ker
Op
ti
m
iz
at
ion
Algorithm
(MSOA)
f
or
s
olvi
ng
the
S
hort
-
Term
Hydr
o
T
her
m
al
Sche
du
li
ng (S
T
-
H
TS) pro
blem
co
ns
ide
rin
g
op
e
rati
onal
constraints
. In
[
10]
, a
Mod
i
fied Diffe
re
ntial
Ev
olu
ti
on
(MDE
)
al
gorithm
is
dev
el
op
e
d
f
or
s
olv
i
ng
S
T
-
H
TS
pr
ob
le
m
.
A
two
-
phas
e
neural
netw
ork
base
d
opti
m
iz
at
ion
al
gorithm
fo
r
ST
-
HTS
pro
bl
e
m
is
pro
posed
in
[11].
I
n
[12
]
,
an
ef
fici
ent
op
ti
m
iz
ation
proce
dure
base
d
on
t
he
cl
on
al
sel
ect
ion
al
gorithm
(CSA
)
is
pro
pose
d
for
the
so
luti
on
of
S
T
-
H
TS
pro
ble
m
.
In
[13],
B
end
e
rs
Deco
m
po
sit
io
n
m
e
tho
d
im
pr
ov
e
d
by
Ba
ct
e
rial
Foragi
ng
or
ie
nted
by
Pa
rtic
le
Sw
arm
Op
ti
m
iz
ation
m
et
ho
d
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.
5
,
Oct
ober
20
19
:
3
3
5
9
-
3
3
6
5
3360
(BDI
-
BFP
SO)
is
us
e
d
for
s
olv
in
g
AC
c
onstrai
ne
d
hydr
o
-
t
her
m
al
generati
on
sche
du
li
ng
pro
blem
.
In
[
14
]
,
gen
et
ic
al
gorithm
is
app
li
ed
t
o
s
olv
e
the
hy
dro
-
t
her
m
al
scheduli
ng
(H
T
S
)
prob
le
m
with
optim
al
powe
r
fl
ow
(O
P
F).
The
hy
dro
s
ub
-
pr
ob
le
m
is
so
lved
usi
ng
gen
et
ic
al
gorithm
,
and
the
therm
al
su
b
-
pro
blem
is
s
olve
d
us
in
g
lam
bd
a
it
erati
on
t
ech
niq
ue
without
li
ne
losses.
Re
f
eren
ce
[
15]
presents
a
cl
on
al
real
-
co
ded
quantum
-
insp
ire
d
e
vo
l
ut
ion
ary
al
go
rithm
(CRQEA)
with
Ca
uc
hy
m
utati
on
f
or
so
lvi
ng
ST
-
HTS
pro
blem
.
In
t
his
al
gorithm
, r
eal
-
co
de
d
r
ule is a
dopted
for ha
ndli
ng conti
nu
ous
var
ia
bles.
The
Prob
le
m
:
Re
fer
e
nce
[16]
de
velo
ps
a
S
T
-
H
TS
form
ulati
on
,
w
hich
ta
kes
i
nto
c
ons
iderati
on
of
sche
du
li
ng
t
he
therm
al
un
it
s
a
s
well
as
t
he
hy
dro
a
nd
ther
m
al
gen
erati
on
s
in
a
sche
duli
ng
ho
rizo
n
c
onsist
ing
of
a
num
ber
of
inter
vals.
I
n
[
17
]
,
PS
O
is
app
li
ed
to
de
te
r
m
ine
the
op
ti
m
al
ho
ur
l
y
schedule
of
power
gen
e
rati
on
in
a
hydro
-
t
her
m
al
power
syst
em
.
Re
fer
ence
[
18
]
dev
el
ops
a
m
od
el
f
or
deali
ng
with
the
ST
-
HTS
pro
blem
,
inco
rpor
at
in
g,
as
a
whole,
thre
e
prob
le
m
s
tr
aditi
on
al
ly
an
al
yz
ed
separ
at
el
y:
sh
or
t
-
te
r
m
hydro
therm
al
scheduli
ng
(
HTS),
un
it
-
c
omm
itmen
t,
a
nd
eco
nom
ic
disp
at
ch.
A
n
e
nh
a
nce
d
dif
fer
e
ntial
evo
l
utio
n
(EDE)
al
gorit
hm
to
so
lve
HTS
pro
blem
us
in
g
cha
os
th
eor
y
to
ob
ta
in
sel
f
-
ada
ptive
par
am
et
er
set
tin
gs
i
n
diff
e
re
ntial
ev
olu
ti
on
(D
E
)
i
s
pro
posed
in
[19].
A
c
ultura
l
al
gorithm
to
so
lve
the
opti
m
al
daily
gen
e
rati
on
sche
du
li
ng
of
hydro
-
the
rm
al
powe
r
syst
e
m
s
,
wh
ic
h
ta
kes
the
water
tra
nsport
delay
tim
e
between
c
on
necte
d
reserv
oirs
int
o
co
ns
i
der
at
io
n,
a
nd
can
conve
niently
deal
with
t
he
com
plica
te
d
hydrauli
c
c
ouplin
g
si
m
ulta
neo
us
ly
, is
propose
d
i
n [
20
]
.
The
Pro
posed
So
luti
on:
I
n
re
cent
ye
ars,
opt
i
m
iz
ation
m
e
t
hod
kn
own
as
Teachin
g
Lea
rn
i
ng
Ba
se
d
Op
ti
m
iz
ation
(
TLBO)
has
be
com
ing
m
or
e
popu
la
r
,
an
d
has
been
us
e
d
in
m
any
pract
ic
al
cases,
m
ai
nl
y
because
it
has
dem
on
strat
e
d
good
r
ob
us
t,
conve
rg
e
nce
pro
per
ti
es,
an
d
is
pr
i
ncipall
y
easy
to
unde
rs
ta
nd
.
TLBO
is
a
re
centl
y
dev
el
oped
e
vo
l
ution
a
r
y
al
go
rithm
ba
sed
on
t
wo
ba
sic
con
ce
pts
of
e
ducat
ion,
nam
ely
te
aching
ph
as
e
an
d
le
a
rn
i
ng
ph
a
se
[21].
In
first
ph
a
se,
le
a
rn
e
rs
im
pr
ove
their
knowle
dg
e
or
a
bili
ty
thr
ough
the
te
achin
g
m
et
hodo
l
og
y
of
te
acher,
a
nd
in
seco
nd
par
t
l
earn
e
rs
inc
reas
e
their
kn
ow
le
dg
e
by
interac
ti
ons
a
m
on
g
t
hem
se
lves.
T
he
al
go
rithm
do
es
no
t
req
ui
re
any
al
gorithm
sp
eci
fic
par
am
et
ers
wh
ic
h
m
akes
the
al
gorithm
ro
bust.
In
[
22]
,
te
a
ching
le
ar
ni
ng
base
d
opt
i
m
iz
ation
(TL
BO)
t
o
s
olv
e
ST
-
HT
S
pro
blem
consi
der
i
ng
no
n
-
li
nea
riti
es
li
ke
valve
point
l
oad
i
ng
ef
fects
of
t
he
t
her
m
al
un
it
a
nd
pr
oh
i
bited
discha
rg
e
zo
ne
of
wate
r
rese
r
vo
i
r
of
the
hydro
pla
nts
is
pro
po
se
d.
An
appr
oach
f
or
so
lvi
ng
s
hor
t
-
t
erm
HTS
us
in
g
an
integrate
d
al
gorithm
based
on
te
aching
le
arni
ng
base
d
opti
m
iz
at
ion
(TLBO)
an
d
op
po
sit
ion
al
base
d
le
arn
i
ng
(O
BL
)
is
pro
pose
d
in
[23
]
.
In
this
pa
pe
r,
Cl
us
te
red
Ad
a
ptive
Teachi
ng
Learn
in
g
Ba
s
ed
O
pti
m
iz
at
io
n
(CAT
LBO
)
al
gorithm
is
pro
po
se
d
t
o
s
olv
e
the
short
-
te
rm
HTS
prob
le
m
.
The
pro
posed
al
gori
thm
is
app
li
ed
t
o
s
olv
e
th
e
dail
y
gen
e
rati
on
sc
he
du
li
ng
of
a
te
st
hydr
o
syst
e
m
with
four
interco
nnect
ed
casca
de
hydro
plants.
Sim
ulati
on
resu
lt
s
dem
on
s
trat
e
the
eff
ect
iveness
,
feasi
bi
li
t
y
and
validi
ty
of
the
pro
pose
d
m
e
tho
d
i
n
te
rm
s
of
so
luti
on
pr
eci
sio
n, w
he
n
c
om
par
ed wi
th all
o
th
er al
gorithm
s in
the
li
te
ratur
e.
The
rest
of
t
he
pa
per
is
orga
ni
zed
as
fo
ll
ow
s:
Sect
ion
2
pr
esents
the
pro
blem
fo
rm
ulatio
n
f
or
sho
rt
te
rm
hydr
o
the
rm
al
sched
ulin
g
(
ST
-
HTS).
S
e
ct
ion
3
prese
nts
the
resu
lt
s
and
disc
us
sio
n.
Finall
y,
Sect
ion
4
su
m
m
arizes t
he
co
ntributi
ons
w
it
h
c
oncl
udi
ng r
em
ark
s.
2.
S
HO
RT TER
M
H
Y
DRO
-
T
HER
MA
L
SCHE
DU
LI
NG
(
ST
-
HTS
): P
R
OBL
EM
F
O
R
MU
L
ATIO
N
The
ST
-
H
TS
pro
blem
aims
at
al
locat
ing
the w
at
er
disc
ha
rge
a
m
on
g
s
horte
r
tim
e
intervals
in
order
t
o
m
ini
m
iz
e
the
fu
el
cost
of
the
r
m
al
gen
erato
rs
du
ri
ng
the
sch
edu
li
ng
inter
va
l,
wh
il
e
sat
isfyi
ng
va
rio
us
eq
ualit
y
and ine
qu
al
it
y
const
raints.
2.1.
Mathem
ati
cal
f
ormula
tio
n for S
T
-
HTS
The
ST
-
HT
S
pro
blem
is
aimed
to
m
ini
m
iz
e
the
total
t
her
m
al
po
we
r
ge
nerat
ion
cost,
w
hi
le
m
aking
us
e
of
t
he
a
vaila
bili
ty
of
hydr
o
re
source
as
m
uch
as
possi
ble.
T
he
obj
ect
ive
f
un
ct
i
on
f
or
ST
-
H
TS
pro
bl
e
m
i
s
form
ulate
d
as [24
]
,
m
ini
m
iz
e, total
p
r
oductio
n
c
os
t (
F), i.e.
,
=
∑
∑
(
)
=
1
=
1
(1)
wh
e
re
t
is
t
he
ind
e
x
for
ti
m
e
i
nter
val,
T
is
t
he
total
num
ber
of
tim
e
interv
al
s
for
sc
he
du
l
ing
per
i
od,
M
is
the
total
num
ber
of
t
her
m
al
plant
s,
is
the
the
rm
al
po
wer
gen
e
r
at
ion
of
i
th
the
r
m
al
plant
duri
ng
ti
m
e
t,
(
)
i
s
the
pr
oduction
c
os
t
f
or
ge
ner
at
in
g
the
powe
r
.
I
n
ge
ne
ral,
t
he
f
uel
c
os
t
of
t
her
m
al
ge
ne
rato
rs
ca
n
be
expresse
d
as
a
qu
a
drat
ic
f
unct
ion
of
power g
ener
at
io
n [25],
and is
giv
e
n by,
(
)
=
+
+
(
)
2
(2)
wh
e
re
,
an
d
are the
fuel c
os
t coe
ff
ic
ie
nts
of
i
th
therm
al
p
ower
p
la
nt.
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
Shor
t
-
te
r
m o
ptima
l
hy
dro
-
the
r
ma
l s
che
duli
ng u
si
ng clustere
d ada
ptive
teac
hing
.
..
(
Su
re
nder Re
ddy
Sa
lk
uti)
3361
2.2.
Equalit
y C
onstrain
ts for
th
e ST
-
HT
S Pr
obl
em
2.2.1.
Sy
s
tem
po
w
er b
alanc
e constr
aint
s
The
total
pow
er
ge
ner
at
io
n
f
ro
m
hydr
o
a
nd
therm
al
un
it
s/
p
la
nts
is
the
su
m
of
total
s
yst
e
m
load/
dem
and
p
l
us
s
yst
e
m
losses
in
each h
our of t
he
sc
he
du
li
ng i
nter
val [26
]
.
+
∑
=
+
,
=
1
,
2
,
…
,
=
1
(3)
wh
e
re
N
is
t
he
total
num
ber
of h
yd
ro p
la
nts,
is
t
he
syst
em
load/
dem
and
du
rin
g
ti
m
e
per
io
d
t,
a
nd
,
is
the
tra
ns
m
issi
o
n
lo
sses o
f
the syst
e
m
du
ri
ng tim
e
per
iod
t. Th
e h
yd
ro
po
w
er g
ene
rati
on
(
)
is
e
xpresse
d
as
a f
un
ct
io
n of w
at
er d
isc
harge
rate an
d
st
or
a
ge
volum
e as [
24
],
=
1
2
+
2
2
+
3
(
)
+
4
+
5
+
6
(4)
Her
e
,
1
,
2
,
3
,
4
,
5
and
6
are
the
powe
r
g
e
ner
at
io
n
c
oe
ff
ic
ie
nts
of j
th
hydro
p
l
a
nt.
2.2.2.
Water
d
ynamic
b
alan
ce (
or)
hydra
u
li
c con
tin
uity
constrain
t
The
st
or
a
ge res
ervoir
volum
e lim
it
s ar
e ex
pressed
with
give
n
init
ia
l an
d final v
olu
m
es as foll
ow
s:
=
,
−
1
+
∑
(
,
−
+
,
−
)
+
−
−
=
1
(5)
wh
e
re
is
t
he
set
of
upstrea
m
un
it
s
directl
y
ab
ove
t
he
hy
dro
-
plant,
is
the
water
del
ay
tim
e
betwe
e
n
reserv
oir
a
nd
it
s
upstream
.
is
the
nat
ur
al
i
nf
l
ow
into
rese
rvoir
j
at
ti
m
e
int
erv
al
t,
is
the
water
discha
r
ge
of
hy
dro
plant
j
at
ti
m
e
interv
al
t,
is
the
wat
er
s
pill
age
of
hy
dro
plant
j
at
tim
e
interval
t,
an
d
is
the
water
volum
e o
f
r
ese
rvoir
j at t
he
e
nd of tim
e inter
val t.
2.3.
Inequ
alit
y Constr
aint
s
for S
T
-
H
TS Pr
ob
le
m
2.3.1.
T
herma
l gener
ators p
ower l
im
its
The ge
ner
at
io
n l
i
m
i
ts of
e
quiv
al
ent therm
al
g
ener
at
or
is
g
i
ve
n by [
27]
,
≤
≤
(6)
wh
e
re
an
d
are
m
ini
m
u
m
an
d m
axi
m
u
m
p
ower
gen
e
rati
on
of i
th
the
rm
al
p
ow
er
p
la
nt [28
]
.
2.3.2
.
Hydro
gener
ators
p
ower l
im
its
The o
per
at
in
g
l
i
m
i
t of
hydro
pl
ant is g
i
ven b
y
[24
]
,
≤
≤
(7)
wh
e
re
an
d
are
the m
ini
m
u
m
and m
axi
m
u
m
powe
r gene
rati
o
n o
f hyd
ro p
la
nt j.
2.3.3.
Reser
vo
ir
capaci
t
y
c
onstr
aint
The
operati
ng
vo
lum
e
of
reser
voir
sto
rage
lim
it
m
us
t
l
ie
in
between
m
ini
m
u
m
and
m
axi
m
u
m
capaci
ty
lim
i
ts, an
d
is
giv
e
n b
y,
≤
≤
(8)
wh
e
re
an
d
are
the m
ini
m
u
m
and m
axi
m
u
m
water
volum
e o
f
r
ese
rvoir
j.
2.3.4.
Hydro
w
at
er
dischar
ge
r
at
e
li
mi
ts
The
hydro
wa
te
r
discha
r
ge
r
at
e
lim
it
m
us
t
li
e
in
betwee
n
it
s
m
ini
m
um
and
m
axi
m
um
op
erati
ng
lim
it
s,
and is
gi
ven
by,
≤
≤
(9)
wh
e
re
an
d
are
the m
ini
m
u
m
and m
axi
m
u
m
water
discha
r
ge
of
hydro plan
t j.
The
ab
ove
obje
ct
ive
functi
on
is
so
lve
d
usi
ng
the
Cl
us
te
red
Ad
a
ptive
Teachin
g
Lear
ning
Ba
se
d
Op
ti
m
iz
ation
(C
ATLBO
)
al
gorithm
. Th
e
d
e
ta
il
ed
d
esc
ription o
f
C
ATLB
O
is
pr
es
ente
d i
n
[
29
-
30
]
.
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.
5
,
Oct
ober
20
19
:
3
3
5
9
-
3
3
6
5
3362
3.
RESU
LT
S
AND DI
SCUS
S
ION
To
te
st
t
he
ef
f
ect
iveness
of
t
he
pro
po
se
d
C
ATLB
O
al
gori
thm
fo
r
S
T
-
HT
S
pr
ob
le
m
,
a
te
st
syst
e
m
is
con
sidere
d
sa
m
e
as
in
ref
e
re
nce
[29
]
.
T
his
syst
e
m
con
sist
s
of
m
ulti
-
chain,
f
our
hydro
plant
casca
de,
and
an
equ
i
valent
t
herm
al
plant.
T
he
sche
duli
ng
pe
rio
d
c
onsidere
d
is
1
day
with
hourl
y
inter
vals.
The
hy
drauli
c
syst
e
m
con
sid
ered
is
ch
arac
te
rized
by
ri
ve
r
tra
nsport
de
la
y
betwee
n
su
ccessi
ve
re
s
ervoirs
,
var
ia
bl
e
head
hydro
pla
nts,
va
riable
natu
ra
l
inflow
rates
in
to
each
rese
rvoir,
pr
oh
i
bited
op
e
rati
ng
zo
ne
s
of
water
disc
harge
rates,
var
ia
ble
load
dem
and
over
sc
hedulin
g
interval.
The
qu
a
drat
ic
fu
el
cost
char
act
e
risti
cs
of
the
equi
valent
therm
al
u
nit i
s
giv
e
n by,
(
)
=
5000
+
19
.
2
+
0
.
002
(
)
2
(10)
The
lo
wer
a
nd
uppe
r
po
we
r
lim
i
ts
of
this
equ
i
valent
therm
al
gen
era
tor/unit
are
500
M
W
a
nd
2500M
W
respec
ti
vely
,
and
f
or
hydrauli
c
unit
s
are
0
M
W
an
d
500
M
W,
re
sp
ect
ivel
y.
Two
dif
fer
e
nt
case
stud
ie
s a
re c
on
sidere
d
to
d
em
on
st
rate t
he
ef
f
ect
iveness o
f
t
he pr
opos
e
d
C
ATLB
O
al
gori
thm
, an
d
th
ey
are:
-
Ca
se 1
: Sy
ste
m
w
it
h
qu
a
dr
at
ic
co
st c
urve a
nd w
it
ho
ut pr
ohibit
ed disc
harge z
on
e
s effect
.
-
Ca
se 2
: Sy
ste
m
w
it
h
proh
i
bited d
isc
harge
z
on
e
s effect
.
3.1.
Ca
se
1
This
case
c
ons
iders
quad
rati
c
cost
cu
r
ve
wi
t
hout
proh
i
bited
disc
ha
rg
e
z
ones
ef
fect.
Ta
bl
e
1
sho
ws
the
hour
ly
hydro
pla
nt
powe
r
ou
tp
uts,
an
d
total
therm
al
gen
erati
on
f
or
Ca
se
1.
The
m
ini
m
u
m
cost
ob
ta
ined
with
propose
d
CATLB
O
al
gorithm
is
9222
66.04
$.
H
ourly
hydro
plant
discha
rg
e
f
or
Ca
se
1
is
repo
rte
d
i
n
Table
2.
Ta
ble
3
s
hows
t
he
op
ti
m
u
m
cost
ob
ta
ine
d
w
it
h
oth
e
r
te
ch
niq
ues
r
ep
or
te
d
in
t
he
li
te
ratur
e
.
The
op
ti
m
u
m
costs
obta
ine
d
from
the
pr
opose
d
C
ATLB
O
al
gorithm
with
that
of
dy
nam
ic
pr
og
ra
m
m
ing
(D
P
),
Non
-
Lin
ear
Program
m
i
ng
(
NL
P)
,
E
vo
luti
on
a
ry
Pr
og
r
a
m
m
ing
(IFEP
),
an
d
Dif
fer
e
nt
ia
l
Evo
luti
on
(
DE),
Local
visi
on
of
PS
O
wit
h
in
erti
a
weig
ht
(
L
W
P
SO),
Im
pr
ove
d
Partic
le
Sw
a
rm
Op
ti
m
iz
at
ion
(I
P
S
O)
,
an
d
Mod
ifie
d
See
ke
r
O
ptim
iz
at
io
n
Al
gorithm
(MSOA)
are
pr
esented
i
n
Ta
bl
e
3.
T
he
pro
pose
d
ap
proac
h
yi
el
ds
bette
r
res
ul
t
than
DP,
NLP
,
IF
EP
,
DE,
IPSO,
and
MSO
A,
w
hi
le
sat
isfyi
ng
the
reser
voi
r
end
-
vo
l
um
e con
strai
nts
.
Table
1
.
Hyd
ro p
la
nt/rese
r
vo
i
r
pow
e
r o
utputs an
d
tota
l t
herm
al
g
ener
at
io
n f
or
Case
1
Hou
r
Hy
d
ro
Po
we
r Ge
n
e
ratio
n
s
(i
n
M
W)
T
h
e
rma
l P
o
we
r
Ge
n
e
ratio
n
s (
MW)
T
o
ta
l P
o
we
r Ge
n
e
ration
(MW)
P
lant
1
P
lant
2
P
lant
3
P
lant
4
1
8
5
.1
4
8
5
7
.8
8
2
0
.000
2
0
0
.099
1
0
2
6
.8
7
1
1
3
7
0
2
8
8
.2
1
5
5
2
.4
3
4
0
.000
1
8
7
.755
1
0
6
1
.5
9
7
1
3
9
0
3
8
0
.2
5
4
5
3
.9
1
8
0
.000
1
7
3
.733
1
0
5
2
.0
9
5
1
3
6
0
4
7
6
.9
8
0
5
8
.0
4
5
0
.000
1
5
6
.791
9
9
8
.185
1
2
9
0
5
7
5
.8
3
4
54
.2
5
3
2
4
.7
8
7
1
7
8
.741
9
5
6
.386
1
2
9
0
6
7
0
.8
4
5
5
6
.1
8
0
2
8
.8
4
9
1
9
8
.957
1
0
5
5
.1
6
8
1
4
1
0
7
7
1
.2
3
1
5
5
.9
8
4
3
1
.3
4
3
2
1
7
.440
1
2
7
4
.0
0
2
1
6
5
0
8
7
5
.2
1
1
6
2
.4
0
6
3
3
.4
5
9
2
3
4
.185
1
5
9
4
.7
4
0
2
0
0
0
9
7
6
.5
3
5
6
5
.9
5
7
3
5
.0
6
7
2
3
9
.065
1
8
2
3
.3
7
6
2
2
4
0
10
8
0
.1
6
2
6
8
.3
7
4
3
5
.1
0
3
2
4
3
.061
1
8
9
3
.3
0
0
2
3
2
0
11
7
9
.0
3
3
6
7
.0
0
3
3
6
.7
6
2
2
4
6
.302
1
8
0
0
.9
0
0
2
2
3
0
12
8
0
.3
1
3
7
1
.9
0
1
3
7
.7
4
4
2
5
1
.400
1
8
6
8
.6
4
3
2
3
1
0
13
7
9
.6
9
7
7
1
.7
4
7
3
7
.6
3
3
2
6
4
.148
1
7
7
6
.7
7
5
2
2
3
0
14
8
0
.3
0
1
7
0
.9
7
3
3
7
.0
5
4
2
7
2
.010
1
7
3
9
.6
6
1
2
2
0
0
15
8
0
.2
8
8
7
4
.3
9
1
3
7
.4
6
0
2
6
8
.170
1
6
6
9
.6
9
1
2
1
3
0
16
7
9
.8
7
4
7
4
.0
0
2
3
6
.6
6
3
2
7
0
.423
1
6
0
9
.0
3
9
2
0
7
0
17
7
7
.8
2
2
7
5
.4
3
6
3
8
.9
2
1
2
7
7
.736
1
6
6
0
.0
8
5
2
1
3
0
18
7
3
.7
5
4
7
5
.9
4
9
4
3
.1
9
7
2
8
2
.941
1
6
6
4
.1
5
8
2
1
4
0
19
7
7
.1
0
5
7
3
.0
8
8
4
6
.2
6
8
2
8
5
.244
1
7
5
8
.2
9
4
2
2
4
0
20
7
5
.3
5
2
7
6
.8
2
3
4
9
.1
4
1
2
8
8
.920
1
7
8
9
.7
6
4
2
2
8
0
21
7
4
.4
8
9
7
7
.2
98
5
0
.6
3
7
2
9
5
.627
1
7
4
1
.9
4
8
2
2
4
0
22
7
4
.7
0
6
6
7
.9
1
8
5
2
.7
2
8
2
9
9
.730
1
6
2
4
.9
1
7
2
1
2
0
23
5
8
.7
4
2
6
9
.5
4
4
5
4
.5
8
4
2
9
4
.779
1
3
7
2
.3
5
1
1
8
5
0
24
5
5
.0
3
3
7
0
.4
4
3
5
6
.0
6
9
2
9
5
.213
1
1
1
3
.2
4
3
1
5
9
0
T
o
tal G
e
n
e
ratio
n
C
o
st
=
922266
.0
4
$
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
Shor
t
-
te
r
m o
ptima
l
hy
dro
-
the
r
ma
l s
che
duli
ng u
si
ng clustere
d ada
ptive
teac
hing
.
..
(
Su
re
nder Re
ddy
Sa
lk
uti)
3363
Table
2
.
H
ourly
p
la
nt/rese
rvo
ir d
i
sc
harge
(
×
10
4
3
)
f
or
Ca
se
1
Hou
r
Hy
d
ro
D
is
c
h
a
rg
e
s (
×
10
4
3
o
f wa
ter)
R
e
serv
o
ir
V
o
l
u
m
e
(
×
10
4
3
o
f wa
ter)
P
lant
1
P
lant
2
P
lant
3
P
lant
4
P
lant
1
P
lant
2
P
lant
3
P
lant
4
0
0
0
0
0
1
0
0
.0
8
0
.0
1
7
0
.0
1
2
0
.0
1
9
.80
7
.25
3
0
.0
1
3
.0
1
0
0
.20
8
0
.7
5
1
4
8
.10
1
0
9
.80
2
1
0
.6
4
6
.28
3
0
.0
1
3
.0
9
8
.5
6
8
2
.4
6
1
2
6
.30
9
9
.2
0
3
9
.02
6
.28
3
0
.0
1
3
.0
9
7
.5
4
8
5
.1
9
1
1
0
.11
8
7
.8
0
4
8
.53
6
.70
3
0
.0
1
3
.0
9
6
.0
1
8
7
.4
9
1
0
0
.0
7
4
.8
0
5
8
.48
6
.0
1
8
.3
0
1
3
.0
9
3
.5
3
8
9
.4
9
1
0
0
.0
9
1
.8
0
6
7
.69
6
.21
1
7
.4
1
1
3
.0
9
2
.8
4
9
0
.2
8
1
0
1
.39
1
0
8
.80
7
7
.74
6
.20
1
6
.8
5
1
3
.0
9
3
.1
0
9
0
.0
8
1
0
2
.72
1
2
5
.80
8
8
.36
7
.15
1
6
.0
7
1
3
.0
9
3
.7
4
8
9
.9
3
1
0
2
.34
1
4
2
.80
9
8
.495
7
.679
1
5
.3
5
7
1
3
.0
9
5
.2
4
5
9
0
.2
4
6
1
0
1
.934
1
4
8
.098
10
9
.025
7
.983
1
5
.3
7
6
1
3
.0
9
7
.2
2
0
9
1
.2
6
3
1
0
2
.117
1
5
2
.500
11
8
.618
7
.648
1
4
.9
4
8
1
3
.0
1
0
0
.602
9
2
.6
1
5
1
0
3
.815
1
5
6
.349
12
8
.775
8
.502
1
5
.1
9
2
1
3
.2
7
1
0
1
.827
9
2
.1
1
3
1
0
7
.327
1
5
9
.152
13
8
.545
8
.524
1
6
.0
5
2
1
4
.5
1
1
0
4
.281
9
1
.5
8
9
1
1
1
.876
1
5
9
.999
14
8
.491
8
.330
1
6
.6
8
2
1
5
.3
7
8
1
0
7
.790
9
2
.2
6
0
1
1
4
.618
1
5
9
.998
15
8
.389
8
.921
1
6
.9
7
3
1
4
.9
4
7
1
1
0
.401
9
2
.3
3
9
1
1
7
.692
1
5
9
.999
16
8
.265
8
.943
1
7
.4
3
2
1
5
.1
9
9
1
1
2
.136
9
1
.3
9
6
1
1
9
.275
1
5
9
.992
17
7
.921
9
.470
1
6
.9
4
1
1
6
.0
4
5
1
1
3
.215
8
8
.9
2
6
1
2
1
.053
1
5
9
.999
18
7
.323
1
0
.0
5
9
1
5
.8
0
1
1
6
.6
8
2
1
1
3
.891
8
4
.8
6
7
1
2
4
.439
1
5
9
.999
19
7
.819
9
.828
1
4
.7
7
6
1
6
.9
7
4
1
1
3
.072
8
2
.0
3
9
1
2
7
.527
1
5
9
.999
20
7
.609
1
1
.1
3
8
1
3
.5
4
3
1
7
.4
5
8
1
1
1
.463
7
8
.9
0
0
1
3
1
.777
1
5
9
.973
21
7
.499
1
1
.7
6
9
1
0
.0
0
1
1
8
.6
3
3
1
1
0
.964
7
6
.1
3
1
1
4
1
.654
1
5
8
.281
22
7
.517
9
.649
1
0
.0
0
1
1
9
.9
7
3
1
1
1
.447
7
5
.4
8
3
1
5
1
.091
1
5
4
.109
23
5
.445
1
0
.3
5
7
1
0
.0
1
0
2
0
.1
0
8
1
1
5
.001
7
3
.1
2
6
1
6
0
.719
1
4
8
.777
24
5
.001
11
.1
2
6
1
0
.0
0
5
2
2
.3
2
0
1
2
0
70
1
7
0
1
4
0
Table
3
.
C
om
par
iso
n of o
ptim
al
co
sts
for
te
st
syst
e
m
w
it
h
quad
rati
c cost
and no
pro
hib
it
ed discha
rg
e
z
on
e
s fo
r
Ca
se
1
Alg
o
r
it
h
m
Min
im
u
m
c
o
st
($
)
Alg
o
r
it
h
m
Min
im
u
m
c
o
st
($
)
DP [6
]
9
2
8
9
1
9
.1
5
L
WP
S
O
[32
]
9
2
5
3
8
3
.8
GA [
3
]
9
2
6
7
0
7
.0
0
DE
[6
]
9
2
3
5
7
4
.3
1
NL
P
[6
]
9
2
4
2
4
9
.4
8
MD
E
[1
0
]
9
2
2
5
5
5
.4
4
F
E
P
[3
1
]
9
3
0
2
6
7
.9
2
IP
S
O
[
6
]
9
2
2
5
5
3
.4
9
C
E
P
[3
1
]
9
3
0
1
6
6
.2
5
MS
OA
[9
]
9
2
2
3
5
5
IF
E
P
[
4
]
9
3
0
1
2
9
.8
2
C
AT
L
B
O
9
2
2
2
6
6
.0
4
3.2.
Ca
se
2
Table
4
prese
nt
s
the
ho
ur
ly
hy
dro
pla
nt
po
wer
outp
uts,
t
he
rm
al
po
wer
ge
ner
at
io
n,
an
d
total
powe
r
gen
e
rati
on
f
or
Ca
se
2.
T
he
m
ini
m
u
m
th
erm
al
gen
erati
on
c
os
t
ob
ta
i
ne
d
i
n
this
c
ase
is
912772
.3159
$.
The
opti
m
a
l
hydro
disc
harge
a
nd
st
or
a
ge
vol
um
es
ob
ta
ined
from
pr
opos
e
d
CATLBO
al
gorithm
are
pr
esented
in Ta
ble 5.
Table
4
.
Hyd
ro p
la
nt
powe
r o
utputs a
nd tota
l t
her
m
al
g
enera
ti
on
for
Ca
se
2
Hou
r
Hy
d
ro
Po
we
r Ge
n
e
ratio
n
s
(i
n
M
W)
T
h
e
rma
l P
o
we
r Ge
n
e
ratio
n
s
(MW)
T
o
tal Po
we
r Ge
n
e
ration
(MW)
P
lant
1
P
lant
2
P
lant
3
P
lant
4
1
8
5
.8
4
5
6
3
.4
2
1
0
.000
2
0
3
.300
1
0
1
7
.4
3
4
1
3
7
0
2
9
1
.6
7
5
5
5
.6
3
6
0
.000
1
8
8
.290
1
0
5
4
.3
9
9
1
3
9
0
3
8
0
.9
1
4
5
1
.3
5
5
0
.000
1
7
3
.338
1
0
5
4
.3
9
3
1
3
6
0
4
8
6
.5
9
2
6
6
.6
6
0
0
.000
1
5
6
.278
9
8
0
.470
1
2
9
0
5
6
8
.0
4
7
5
8
.8
3
4
4
1
.5
9
7
1
7
8
.002
9
4
3
.519
1
2
9
0
6
6
7
.1
4
6
5
3
.3
8
4
0
.000
1
9
8
.094
1
0
9
1
.3
7
6
1
4
1
0
7
5
3
.6
2
3
7
0
.2
8
9
3
3
.9
4
0
2
1
5
.990
1
2
7
6
.1
5
9
1
6
5
0
8
6
3
.7
9
1
3
.649
4
1
.5
7
7
2
3
2
.178
1
6
0
8
.8
0
5
2
0
0
0
9
8
2
.6
3
4
5
2
.6
3
3
4
1
.7
7
1
2
3
2
.411
1
8
3
0
.5
5
1
2
2
4
0
10
8
5
.4
4
1
7
6
.6
0
9
4
2
.7
6
2
2
4
7
.716
1
8
6
7
.4
7
1
2
3
2
0
11
8
5
.2
0
6
5
3
.9
4
5
4
4
.9
9
2
2
5
2
.256
1
7
9
3
.6
0
1
2
2
3
0
12
5
6
.0
8
6
5
5
.3
8
6
4
5
.6
5
0
2
4
8
.401
1
9
0
4
.4
7
8
2
3
1
0
13
8
7
.5
3
5
5
7
.6
4
5
4
0
.6
3
4
2
5
0
.579
1
7
9
3
.6
0
8
2
2
3
0
14
6
6
.3
3
3
5
8
.7
4
6
3
3
.9
4
0
2
4
7
.386
1
7
9
3
.5
9
5
2
2
0
0
15
7
7
.8
4
5
7
1
.6
4
0
4
7
.8
0
2
2
5
0
.005
1
6
8
2
.7
0
8
2
1
3
0
16
6
9
.6
6
5
6
1
.9
5
9
4
5
.5
7
2
2
4
7
.070
1
6
4
5
.7
3
3
2
0
7
0
17
8
8
.9
9
2
8
5
.2
1
3
4
1
.0
1
8
2
6
9
.019
1
6
4
5
.7
5
8
2
1
3
0
18
7
5
.4
2
4
8
2
.3
4
2
4
2
.2
9
7
2
9
4
.238
1
6
4
5
.6
9
9
2
1
4
0
19
8
7
.2
7
0
5
6
.0
2
3
4
6
.6
9
3
2
5
6
.419
1
7
9
3
.5
9
6
2
2
4
0
20
5
5
.3
2
8
8
5
.6
1
8
5
1
.1
3
8
2
9
4
.337
1
7
9
3
.5
7
9
2
2
8
0
21
7
3
.2
0
0
8
4
.4
9
2
5
2
.6
1
9
2
7
3
.050
1
7
5
6
.6
3
8
2
2
4
0
22
7
3
.8
2
9
7
7
.2
5
3
5
4
.7
5
9
3
0
5
.385
1
6
0
8
.7
7
4
2
1
2
0
23
7
8
.3
3
2
7
3
.5
5
8
5
6
.0
4
6
2
9
2
.006
1
3
5
0
.0
5
9
1
8
5
0
24
6
7
.3
1
1
4
5
.1
1
7
5
8
.8
3
1
2
9
0
.436
1
1
2
8
.3
0
6
1
5
9
0
T
o
tal G
e
n
e
ratio
n
C
o
st
=
912772
.3
1
5
9
$
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.
5
,
Oct
ober
20
19
:
3
3
5
9
-
3
3
6
5
3364
Table
5
.
H
ourly
p
la
nt
discha
r
ge (
×
10
4
3
)
for
Ca
se
2
Ho
u
r
Hy
d
ro Disch
a
rges
(
×
10
4
3
o
f
water
)
Res
ervo
ir
Vo
lu
m
e
(
×
10
4
3
o
f
water
)
Plan
t 1
Plan
t 2
Plan
t 3
Plan
t 4
Plan
t 1
Plan
t 2
Plan
t 3
Plan
t 4
0
0
0
0
0
1
0
0
.000
8
0
.00
0
1
7
0
.000
1
2
0
.000
1
9
.96
4
8
.27
4
2
9
.65
9
1
3
.42
0
1
0
0
.036
7
9
.7
26
1
4
8
.441
1
0
9
.380
2
1
1
.77
4
6
.89
9
2
9
.98
3
1
3
.14
1
9
7
.26
3
8
0
.82
7
1
2
6
.658
9
8
.63
9
3
9
.25
3
6
.02
4
2
9
.72
5
1
3
.04
5
9
6
.01
0
8
3
.80
4
1
1
0
.897
8
7
.19
4
4
1
0
.93
4
8
.33
6
2
9
.14
0
1
3
.04
6
9
2
.07
6
8
4
.46
8
1
0
3
.804
7
4
.14
8
5
7
.37
4
6
.97
4
1
3
.28
2
1
3
.06
9
9
0
.70
2
8
5
.49
4
1
0
9
.674
9
0
.73
8
6
7
.25
1
6
.10
9
2
8
.67
4
1
3
.04
8
9
0
.45
1
8
6
.38
5
1
0
1
.958
1
0
7
.674
7
5
.32
7
9
.12
4
1
6
.23
6
1
3
.00
1
9
3
.12
4
8
3
.26
0
1
0
4
.432
1
2
4
.397
8
6
.53
3
6
.34
0
1
2
.49
4
1
3
.01
4
9
5
.59
1
8
3
.92
1
1
0
8
.163
1
4
0
.523
9
9
.60
1
6
.04
2
1
2
.21
4
1
3
.01
1
9
5
.99
0
8
5
.87
9
1
0
8
.384
1
4
0
.794
10
1
0
.13
8
1
0
.22
8
1
2
.97
0
1
3
.14
2
9
6
.85
2
8
4
.65
0
1
1
2
.072
1
5
6
.325
11
9
.91
1
6
.10
0
1
1
.73
9
1
3
.35
0
9
8
.94
1
8
7
.55
0
1
1
7
.274
1
5
9
.212
12
5
.32
9
6
.16
7
1
3
.69
4
1
3
.01
2
1
0
3
.612
8
9
.38
3
1
2
1
.760
1
5
8
.695
13
9
.99
2
6
.36
8
1
7
.34
2
1
3
.32
0
1
0
4
.620
9
1
.01
5
1
2
8
.558
1
5
7
.589
14
6
.42
6
6
.33
6
1
8
.74
2
1
3
.00
7
1
1
0
.1
95
9
3
.67
9
1
2
4
.245
1
5
7
.552
15
7
.92
3
8
.25
3
1
4
.09
7
1
3
.41
0
1
1
3
.272
9
4
.42
6
1
2
9
.307
1
5
5
.881
16
6
.73
2
6
.65
1
1
5
.40
5
1
3
.06
3
1
1
6
.540
9
5
.77
5
1
2
8
.696
1
5
6
.512
17
9
.68
1
1
1
.25
9
1
7
.12
5
1
5
.17
3
1
1
5
.858
9
1
.51
6
1
2
7
.831
1
5
8
.681
18
7
.50
2
1
1
.33
9
1
6
.71
3
1
8
.30
9
1
1
6
.357
8
6
.17
8
1
2
8
.102
1
5
9
.114
19
9
.44
3
6
.45
9
1
5
.10
6
1
3
.74
9
1
1
3
.914
8
6
.71
9
1
3
0
.329
1
5
9
.462
20
5
.06
9
1
3
.20
0
1
2
.91
0
1
8
.76
7
1
1
4
.845
8
1
.51
9
1
3
7
.179
1
5
6
.099
21
7
.23
2
1
4
.29
1
1
0
.09
4
1
5
.76
5
1
1
4
.613
7
6
.22
8
1
4
9
.868
1
5
7
.459
22
7
.30
3
1
2
.49
5
1
1
.74
0
2
1
.27
0
1
1
5
.310
7
2
.73
3
1
5
1
.655
1
5
2
.902
23
7
.91
5
1
2
.38
0
1
0
.78
4
1
9
.71
1
1
1
6
.394
6
8
.35
3
1
6
2
.303
1
4
8
.296
24
6
.39
4
6
.35
3
1
3
.89
7
2
1
.20
6
120
70
170
140
4.
CONCL
US
I
O
N
In
this
pa
per,
a
new
Cl
us
te
red
A
da
ptive
Teachin
g
Learn
i
ng
Ba
se
d
Op
ti
m
iz
at
ion
(CATLBO
)
al
gorithm
is
dev
el
op
e
d
to
so
l
ve
the
Shor
t
-
Term
Hydro
The
rm
al
Sche
duli
ng
(S
T
-
HTS)
pro
blem
.
The
pr
opos
e
d
al
gorithm
is
tested
on
a
sta
ndar
d
sam
ple
test
syst
e
m
con
siderin
g
th
ree
di
ff
ere
nt
case
stud
ie
s
.
This
al
gorithm
has
prov
i
ded
the
best
resu
lt
s
com
par
ed
to
oth
er
co
nvent
ion
al
an
d
m
e
ta
-
he
ur
ist
ic
al
gor
it
hm
s
li
ke
Dynam
ic
Program
m
ing
(DP),
N
on
-
Li
near
Pro
gr
am
m
ing
(
NLP),
Ev
olu
ti
onary
Pr
og
ram
m
ing
(I
F
EP),
Diff
e
re
ntial
Evo
l
ution
(
DE
)
,
Im
pr
ove
d
P
arti
cl
e
Sw
arm
O
pti
m
iz
at
ion
(IPS
O)
,
an
d
Mod
i
fied
Se
eker
Op
ti
m
iz
ation
Algorithm
(MSOA)
repo
rted
in
the
li
te
ratu
r
e.
This
C
ATL
BO
al
gorithm
can
easi
ly
be
e
xten
ded
to an
y
oth
e
r
c
om
plex
op
ti
m
izati
on
pro
blem
s
f
ace
d by the
ut
il
i
ti
es.
ACKN
OWLE
DGE
MENTS
This
resea
rch
work
has
bee
n
carried
out
ba
sed
on
the
sup
port
of
“
Woos
ong
U
niv
e
rsity
'
s
Acad
em
ic
Re
search
F
undi
ng
-
2019
”
.
REFERE
NCE
S
[1]
A.
Mahor
and
S.
Rangne
k
ar
,
“
Short
te
rm
g
ene
ra
tion
sche
duli
ng
of
ca
sc
ade
d
h
y
dro
el
e
ct
ri
c
s
y
s
te
m
u
sing
novel
se
lf
ada
pt
ive
ine
r
ti
a
weight
PS
O
,
”
Elec
tri
cal
Powe
r a
nd
Ene
rgy
Syst
e
ms
,
vol. 34, pp.
1
-
9,
2012
.
[2]
X.
Liao
,
et
al
.
,
“
An
ada
pti
ve
cha
ot
ic
ar
ti
fi
cia
l
bee
col
on
y
algorithm
for
short
-
te
rm
h
y
d
roth
ermal
gene
r
at
io
n
sche
duli
ng
,
”
Ele
ct
rical P
ow
er
an
d
Ene
rgy
Syst
ems
,
vol. 53, pp. 34
-
42,
2013
.
[3]
C.
E
.
Zoumas
,
e
t
al
.
,
“
A
gene
t
ic
al
gori
thm
solut
ion
appr
o
ac
h
to
the
h
y
dro
-
the
rm
al
coor
dination
proble
m
,
”
I
EEE
Tr
ansact
ions o
n
Powe
r Sy
st
ems
,
vol
/i
ss
ue:
19
(
3
)
,
pp.
1356
-
1364
,
2004
.
[4]
N.
Sinha
,
et
al
.
,
“
Fast
evol
u
ti
o
nar
y
progra
m
m
ing
te
chn
ique
s
f
or
short
te
rm
h
y
dro
-
th
ermal
sc
hedul
ing
,
”
I
EEE
Tr
ansacti
ons on Power
Syst
ems
,
vol
/i
ss
ue:
18
(
1
)
,
pp.
214
-
219
,
20
03
.
[5]
B.
Yu
,
e
t
a
l
.
,
“
Sho
rt
-
te
rm
hy
dro
-
th
ermal
sc
hedul
ing
using
par
ticle
sw
arm
opti
m
iz
at
ion
m
et
hod,
”
Ener
gy
Conve
rs
ion
and
Manage
ment
,
vo
l.
48
,
pp
.
1902
-
1
908
,
2007
.
[6]
P.
K.
Hota
,
et
al
.
,
“
An
improved
PSO
te
chni
que
for
short
-
te
rm
o
pti
m
al
h
y
dro
the
r
m
al
sche
duli
ng,
”
El
e
ct
ric
Pow
e
r
Syste
ms
Re
sear
c
h
,
vol
.
79
,
pp
.
10
47
-
1053,
2009
.
[7]
K.
P.
W
ong
and
Y.
W
.
W
ong
,
“
Shor
t
te
rm
hy
dr
o
-
the
rm
al
sche
d
uli
ng
par
t
-
I:
sim
ula
t
ed
anne
a
li
ng
appr
oac
h
,
”
IEE
Proce
ed
ings
-
G
ene
ration
,
Tr
ansm
ission and
Dist
ributi
on
,
vol
/
issue:
141
(
5
)
,
pp
.
4
97
-
501,
1994
.
[8]
T.
G.
W
ern
e
r
a
nd
J.
F.
Verste
ge
,
“
An
evol
uti
o
nar
y
str
at
eg
y
fo
r
short
te
rm
oper
at
ion
pl
anni
ng
of
h
y
dro
-
th
ermal
power
s
y
st
ems
,
”
IEEE
Tr
ansactions
Powe
r Sy
st
e
ms
,
vol
/i
ss
ue:
14
(
4
)
,
pp
.
1362
-
13
68,
1999
.
[9]
K.
R.
Krishnan
and
,
et
a
l
.
,
“
Optimal
short
-
t
er
m
h
y
drothe
rm
al
gene
r
at
io
n
sch
edul
ing
using
m
odifi
ed
see
k
e
r
opti
m
iz
ation
a
lg
orit
hm
,
”
Int
erna
ti
onal
Journal
o
f
Mode
l
ing,
Id
en
ti
ficati
on
and
C
ontrol
,
vol
/i
ss
ue
:
15
(
4
)
,
pp
.
250
-
258,
2012
.
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
Shor
t
-
te
r
m o
ptima
l
hy
dro
-
the
r
ma
l s
che
duli
ng u
si
ng clustere
d ada
ptive
teac
hing
.
..
(
Su
re
nder Re
ddy
Sa
lk
uti)
3365
[10]
L.
L
akshm
ina
ras
imm
an
and
S.
Subram
ani
an
,
“
Short
-
te
rm
sche
dul
ing
of
h
y
dro
the
r
m
al
power
s
y
ste
m
wit
h
ca
sca
ded
rese
rvoirs
b
y
usi
ng
m
odifi
ed
diff
ere
nt
ia
l
evol
u
ti
o
n
,”
I
EE
Proceed
ings
-
Gene
ratio
n,
Tr
ansm
ission
and
Distributi
on
,
vol
/i
ss
ue:
15
3(
6
)
,
pp
.
693
-
700
,
2
006.
[11]
R.
Nare
sh
and
J
.
Sharm
a
,
“
Two
-
Phase
Neura
l
Network
base
d
Soluti
on
T
ec
hni
que
for
Short
-
T
erm
H
y
droth
ermal
Schedul
ing
,
”
IE
E
Proc
ee
dings
-
Gene
ration, Tr
ansm
ission and
Distributi
on
,
vol
/is
sue:
146
(
6
)
,
pp.
657
-
663,
2006
.
[12]
R.
K.
Sw
ai
n
,
e
t
al
.
,
“S
hort
-
te
rm
h
y
droth
ermal
sc
hedul
ing
using
c
lona
l
sel
ection
a
lgori
thm
,
”
El
e
ctr
ic
al
Powe
r
an
d
Ene
rgy
S
yste
ms
,
vol.
33
,
pp
.
647
-
656,
2011
.
[13]
M.
Para
stega
r
i,
et
al
.
,
“
AC
con
strai
ned
h
y
dro
-
t
her
m
al
gene
r
ati
on
sche
duli
ng
p
roble
m
:
Applica
ti
on
of
Bende
rs
dec
om
positi
on
m
et
hod
improve
d
b
y
BF
PS
O
,
”
E
le
c
tric
al
Powe
r
and
Ene
rgy
Syst
ems
,
vol
.
49
,
pp
.
199
-
212,
2013.
[14]
V.
S.
Kum
ar
an
d
M.
R.
Mohan
,
“
A
gene
ti
c
al
gor
it
hm
soluti
on
to
the
opti
m
al
short
-
te
rm
h
y
dro
the
r
m
al
sche
duli
ng
,
”
El
e
ct
rica
l
Pow
er
and
En
ergy
S
yst
ems
,
vol
.
33
,
pp
.
827
-
835
,
2011
.
[15]
Y.
W
anga
,
et
al
.
,
“
A
cl
onal
real
-
code
d
quan
tum
-
inspire
d
evo
lut
i
onar
y
al
gor
it
hm
with
Cauc
h
y
m
uta
ti
on
for
short
-
te
rm
h
y
drotherm
al
g
ene
r
at
ion
sch
edul
ing
,
”
Elec
tri
cal
Powe
r and
E
nergy
Syst
ems
,
v
ol.
43
,
pp
.
1228
-
1240,
2012
.
[16]
S.
Y.
W
.
W
on
g
,
“
H
y
brid
Sim
ula
t
ed
Annea
li
n
g/
Gene
t
ic
Algor
it
hm
appr
oa
ch
to
Short
T
erm
H
y
dro
-
Th
ermal
Schedul
ing
with
Multi
ple
Th
erma
l
Plant
s
,
”
Inte
r
nati
onal
Journal
El
ectric
power
and
Ene
rgy
Syst
ems
,
vol.
23,
pp.
565
-
575,
2001
.
[17]
K.
K.
Manda
l
,
e
t
al
.
,
“
Parti
cle
sw
arm
opti
m
iz
at
i
on
te
chni
qu
e
ba
sed
short
-
te
rm
hy
droth
ermal
sch
edul
ing
,
”
Applie
d
Soft
Computing
,
vol.
8
,
pp
.
1392
-
1399,
2008
.
[18]
E.
Gil
,
e
t
al
.
,
“
Short
-
Te
rm
H
y
droth
ermal
Ge
ner
ation
Schedu
li
ng
Model
Us
ing
a
Gene
ti
c
Algorit
hm
,
”
IE
EE
Tr
ansacti
ons on Power
Syst
ems
,
vol.
18
,
pp
.
1256
-
1264,
2003
.
[19]
X.
Yuana
,
e
t
al
.
,
“
An
enha
nce
d
diffe
ren
ti
a
l
evo
lut
ion
a
lgori
thm
for
dai
l
y
opt
imal
h
y
dro
gene
r
ati
on
sche
duli
ng
,
”
Computers and Mathematics wi
t
h
Applications
,
v
ol.
55
,
pp
.
2458
-
2468,
2008
.
[20]
X.
Yuan
and
Y.
Yuan
,
“
Applicati
on
of
cultura
l
al
gori
thm
to
g
ene
ra
ti
on
sch
ed
uli
ng
of
h
y
dro
t
her
m
al
s
y
st
ems
,
”
Ene
rgy
Con
ve
rs
i
on
and
Manag
e
ment
,
vo
l. 47, pp
.
2192
-
2201
,
20
06
.
[21]
D
.
S
.
N
.
M
.
Rao
and
N.
K
um
ar,
“
Com
par
isiona
l
Inve
stig
at
ion
of
Lo
ad
Dispatc
h
Solu
tions
with
T
LB
O
,
”
Inte
rnational
Jo
urnal
of El
e
ct
ri
c
al
and
Comput
er
Engi
n
ee
ring
(
IJE
CE)
,
vol
/
issue:
7
(
6
)
,
pp
.
3246
-
3
253,
De
c
2017
.
[22]
P.
K.
Ro
y
,
“
Teac
hing
learni
ng
base
d
opti
m
izat
i
on
for
short
-
ter
m
h
y
drothe
rm
a
l
sche
duli
ng
pro
ble
m
conside
r
ing
val
ve
point
eff
e
ct
and
prohib
it
e
d
discha
rg
e
con
strai
nt
,
”
El
e
ct
ri
c
al
Powe
r
and
E
nergy
Syst
ems
,
vol.
53
,
pp
.
10
-
19,
2013
.
[23]
P.
K.
Ro
y
,
e
t
al
.
,
“
Optimal
short
-
te
rm
h
y
dro
-
ther
m
al
sche
dul
ing
usin
g
quasi
-
opp
ositi
onal
t
ea
ch
in
g
learni
ng
b
ase
d
opti
m
iz
ation
,”
E
ngine
ering
Appli
cat
ions o
f
Art
if
i
c
ial
In
te
l
li
gen
ce
,
vol.
26
,
pp
.
2516
-
2524
,
2013
.
[24]
S.
S.
Redd
y
,
e
t
al
.
,
“
Short
-
T
e
rm
Hy
dro
-
The
r
m
al
Schedul
ing
using
CMA
-
ES
with
Dire
cte
d
Ta
rge
t
to
B
e
st
Perturba
t
ion
Sch
eme
,
”
Inte
rnat
io
nal
Journal
of
Bio
-
Inspired
Computati
on
,
vol
/i
ss
ue:
7
(
3
)
,
pp
.
195
-
208,
2015
.
[25]
S.
S.
Redd
y
,
“
Optimal
React
iv
e
Pow
er
Schedu
li
n
g
Us
ing
Cuc
koo
Sear
ch
Alg
orit
hm
,
”
Int
ernati
onal
Journal
o
f
El
e
ct
rica
l
and
C
omputer
Engi
n
e
ering
(
IJE
CE)
,
v
ol
/i
ss
ue:
7
(
5
)
,
pp
.
2349
-
2356
,
Oc
t
2017.
[26]
S.
C.
Kim
and
S.
R.
Salkut
i
,
“
Optimal
pow
er
flow
base
d
conge
stion
m
an
age
m
ent
using
enha
nc
ed
gen
et
i
c
al
gorit
hm
s,
”
In
t
ernati
onal
Journ
al
of
Elec
tri
cal
and
Computer
Engi
nee
ring
(
IJECE)
,
vol
/i
ss
ue:
9
(
2
)
,
pp.
875
-
883
,
Apr
2019.
[27]
T
.
D.
Tha
nh,
e
t
al
.
,
“
Stocha
sti
c
cont
rol
for
optim
al
p
ower
flow
in
isla
nded
m
ic
rogrid
,
”
Int
ernat
ional
Journal
of
El
e
ct
rica
l
and
C
omputer
Engi
n
e
ering
(
IJE
CE)
,
v
ol
/i
ss
ue:
9
(
2
)
,
pp
.
1045
-
1057
,
Ap
r
2019.
[28]
S.
S.
Sakthi,
e
t
al
.
,
“
W
ind
I
nte
gra
te
d
The
r
m
al
Unit
Com
m
i
tment
Sol
ution
using
Gre
y
W
olf
Optimiz
er
,
”
Inte
rnational
Jo
urnal
of El
e
ct
ri
c
al
and
Comput
er
Engi
n
ee
ring
(
IJE
CE)
,
vo
l
/
issue:
7
(
5
)
,
pp
.
2309
-
2
320,
Oct
2017.
[29]
S.
S.
Redd
y
,
“
Optimal
sc
hedulin
g
of
wind
-
the
rm
al
power
s
y
stem
using
cl
ustere
d
ada
pt
ive
teac
h
in
g
le
arn
ing
base
d
opti
m
iz
ation
,
”
E
le
c
tric
al
Engi
n
e
ering
,
vo
l
/i
ss
ue:
99
(
2
)
,
pp
.
535
-
5
50
,
2017
.
[30]
S.
S.
R
edd
y
,
“
Cl
ustere
d
ada
p
ti
ve
te
a
chi
ng
-
l
ea
rn
in
g
-
base
d
opt
imiz
at
ion
al
gor
it
hm
f
or
solving
the
op
ti
m
al
g
ene
r
at
ion
sche
duli
ng
probl
em
,”
El
e
ct
rica
l Engi
ne
ering
,
vol
/i
ss
ue:
100
(
1
)
,
p
p.
333
-
346
,
201
8.
[31]
N
.
Sinha
,
et
al
.
,
“
Fast
evol
u
ti
o
nar
y
progra
m
m
ing
te
chn
ique
s
f
or
short
te
rm
h
y
dro
-
th
ermal
sc
hedul
ing
,
”
I
EEE
Tr
ansacti
ons on
Powe
r Sy
st
ems
,
vol
/i
ss
ue:
18
(
1
)
,
pp.
214
-
220
,
20
03.
[32]
B.
Yu
,
e
t
a
l
.
,
“
Short
-
te
rm
hy
dro
-
th
ermal
sc
hedul
ing
u
sing
par
ticle
sw
arm
opti
m
iz
at
ion
m
et
hod
,
”
Ener
gy
Conve
rs
ion
and
Manage
ment
,
vo
l.
48
,
pp
.
1902
-
1
908
,
2007
.
Evaluation Warning : The document was created with Spire.PDF for Python.