TELKOM
NIKA Indonesia
n
Journal of
Electrical En
gineering
Vol.12, No.5, May 2014, pp
. 3728 ~ 37
3
6
DOI: http://dx.doi.org/10.11591/telkomni
ka.v12i5.5107
3728
Re
cei
v
ed
No
vem
ber 1
1
, 2013; Re
vi
sed
De
cem
ber 2
3
,
2013; Accep
t
ed Jan
uary 8
,
2014
An Optimization Model of Coal Allocation in a Group
Jianjian Zha
o, Yunhua G
a
n*, Xiaoqian Ma, Zeliang Yang
Schoo
l of Elect
r
ic Po
w
e
r, Sout
h Chi
na Un
iver
sit
y
of T
e
chnol
og
y
Guangzh
ou 5
106
40, Ch
ina,
T
e
l/fax: +
86-20
-871
106
13
*Corres
p
o
ndi
n
g
author, e-ma
i
l
:gan
yh
@scut.
edu.cn
A
b
st
r
a
ct
T
he frame of a
n
opti
m
i
z
at
ion
mo
de
l of coal
a
llocti
on i
n
a p
o
w
e
r gener
atio
n grou
p w
a
s introduce
d
in pres
ent stud
y. T
he influenc
e of mu
lti-c
oals
blen
din
g
an
d transp
o
rtatio
n w
e
re all consi
d
ered i
n
the mo
del.
F
i
rstly, the i
m
p
a
ct of eac
h pr
ocess
of c
oal
usin
g o
n
the c
o
st of pow
er
g
ener
ation
an
d
boil
e
r saf
e
ty in
the
combusti
on pr
ocess w
e
re a
n
a
ly
z
e
d,
an
d th
e deta
ile
d
mat
h
e
m
atic
al d
e
sc
riptio
ns w
e
re g
i
ven. A
mu
lti-c
oals
ble
ndi
ng
math
ematica
l
mod
e
l
base
d
o
n
sa
fty, envir
on
me
ntal prot
ection
and cost w
a
s propos
ed. T
h
e
opti
m
i
z
at
ion
mode
l of c
o
a
l
tr
ansp
o
rt
atio
n
i
n
the pow
er ge
nerati
on gro
u
p
to ac
hiev
e th
e
most b
enefit
w
a
s
establ
ishe
d.
T
he alg
o
rith
ms
of
thes
e
mo
del
s w
e
re stud
ie
d
an
d fo
und
tha
t
the g
e
n
e
tic
al
gorith
m
is
on
e
of
the su
itab
le
methods
to s
o
lv
e the
mo
dels.
An o
p
ti
mai
z
t
i
o
n
syste
m
f
o
r t
he c
oal
a
lloc
a
t
i
on
in
the
p
o
w
e
r
generation gr
oup was deve
loped based on the
mode
ls
and algorit
hm
s. T
he syst
em
adopts friendly
softw
are structure a
nd ca
n p
r
ovid
e pers
o
n
a
li
z
e
d
f
uncti
on
s for the pow
er ge
nerati
on
grou
p an
d p
o
w
er
plants.
Ke
y
w
ords
:
pow
er ge
ner
ation
grou
p, mu
lti-
coals
bl
end
i
ng, coa
l
al
loc
a
tion,
mu
lti-ob
j
e
ctive
o
p
ti
mi
zation
mo
de
l
Copy
right
©
2014 In
stitu
t
e o
f
Ad
van
ced
En
g
i
n
eerin
g and
Scien
ce. All
rig
h
t
s reser
ve
d
.
1. Introduc
tion
In China,
co
al is the p
r
i
m
ary ene
rgy
used
i
n
the
indu
stry pro
d
u
ction, an
d the po
we
r
indu
stry is a
larg
e
cou
s
u
m
er
of
coal
reso
urce
s. Ho
wever, th
e di
stributio
n im
b
a
lan
c
e
of co
al
resou
r
ces a
nd p
o
wer
g
eneration
en
terpri
se
s i
s
very serio
u
s in
Chia
n.
The l
ong
te
rm
perfo
rman
ce
of electri
c
al coal sup
p
ly an
d deman
d is that the coal sho
u
ld be tra
n
sp
orted fro
m
the we
st to the east a
s
well
as from the n
o
rth to the so
uth [1].
On the other hand, with the increa
sin
g
of
coal pri
c
e an
d the sho
r
tage of the co
al
transpo
rtation
,
more
a
nd m
o
re
thermal
p
o
we
r
pl
ant
ha
s n
o
lo
nge
r
si
ngly u
s
ed
the
de
sign
ed
co
al,
but adopted
the multi-coa
l
s blendi
ng tech
nolo
g
y to sovle these
problem
s. Accordi
ng to the
operation ex
perie
nces from many thermal po
we
r plants, the technolo
g
y can expa
nd
the
extensio
n of purcha
s
in
g coal s
ource
s, optimize
th
e coal
tran
spo
r
t, then achi
e
v
e the goal
of
redu
cin
g
the operation cost of thermal pow
e
r
plant
s a
nd po
wer g
e
n
e
ration g
r
o
u
p
[2].
For a p
o
wer
gene
ration g
r
oup, co
al su
p
p
ly is
a larg
e
system an
d i
n
volves a nu
mber of
different a
r
ea
s, incl
uding
coal
su
pplie
r, logisti
cs
ent
erp
r
ise, tran
sit logistics
ce
nter an
d ea
ch
power pl
ant i
n
the g
r
ou
p. Scientific
an
d effici
ent
co
al allo
cation
sho
u
ld p
r
edi
ct coal d
e
ma
n
d
based
on th
e
“Coal
-furnace" couplin
g,
coal
storage
si
tuation a
nd l
o
ad trend
in
e
a
ch
po
we
r
pl
ant
in the gro
up, integrate the
coal
sou
r
ces
inform
atio
n o
f
coal su
pplie
r,
optimizethe deployme
nt
of
variou
s tran
sportation m
e
ans, an
d ma
ke coal s
upp
lier, logisti
cs
provide
r
and
thermal po
wer
plant form a close
d
cycl
e [3].
In orde
r to improve the
competitivene
ss of
a po
we
r gene
ratio
n
grou
p, an opt
imization
model
of co
al
allocation
ba
sed
on
multi-coal
s bl
endi
n
g
techonol
gy
is a
ne
ce
ssity
[4-5].
Jin a
n
d
Kwang[6] prese
n
ted a methodol
ogy, multiagent
-sy
s
tem-ba
sed i
n
telligent ref
e
ren
c
e g
o
vernor
(MAS-IRG), to reali
z
e th
e
optimal ma
pping
by se
arching fo
r t
he be
st solu
tion to the
multi
obje
c
tive opti
m
ization
prob
lem that tackl
e
s
conf
li
cting
req
u
irem
ent
s an
d foun
d that MAS is a
n
efficient meth
odolo
g
ies to
desi
gn the
IC system f
o
r a
com
p
lex la
rg
e-scal
e p
o
we
r pla
n
t. A novel
plant o
p
timization te
chniq
ue
wa
s deve
l
oped
u
s
ing
geneti
c
alg
o
rithms
(GA) to maximi
ze t
h
e
overall
reve
nue
gen
erat
ed by
a
co
al p
r
epa
ratio
n
pla
n
t by
sea
r
ching
th
e be
st p
o
ssible
combi
nation
of overall
yield and
multiple p
r
o
duct q
uality con
s
traints [7]. Experimenta
l
investigatio
ns into the
igni
tion be
haviors of
pulve
rized
coal
s
and
co
al bl
end
s in a
d
r
op
tu
be
furna
c
e
u
s
ing
a flam
e m
oni
toring
sy
stem
we
re
carried
out. The
igniti
on b
ehavio
rs
of a
co
al bl
en
d
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
An Optim
i
zation Model of
Coal Allo
catio
n
in a Grou
p (Jianji
an Zha
o
)
3729
are foun
d to have simila
r cha
r
a
c
teri
stics as the
coal
of higher volatile matter in the blend a
nd
depe
nd
on it
s p
r
op
ortio
n
in the
blen
d.The
re
sults from thi
s
stu
d
y are
u
s
ed
to p
r
edi
ct t
h
e
operation of a
coal
fired p
o
we
r
pl
ant
in
terms of fuel
sele
ction,
fu
e
l
blen
ding,
an
d flame
sta
b
ility
[8].
Ba
s
e
d o
n
ther
ma
l po
we
r
ind
u
s
tr
y SO
2
e
m
issi
on
data
from stat
e de
partme
n
t aut
horitie
s,
con
s
id
erin
g the main
fact
ors of China'
s the
r
mal p
o
w
er indu
stry
SO
2
pre
d
ict
ed emi
ssi
on,
a
combi
nation
predi
ction m
o
del wa
s e
s
ta
blish
ed, co
nn
ecting g
r
ay p
r
edictio
n mod
e
l with BP ne
ural
netwo
rk mo
d
e
l to predict
SO
2
emissi
on
[9]. A study
pre
s
ente
d
an
investigation
on the influence
of hydroth
e
rmally treate
d
munici
pal
soli
d wast
e
(
MS
W) on
th
e co
-comb
u
stio
n chara
c
te
risti
c
s
with
different ra
nk coal
s, i.e. Indi
an, Indon
e
s
ian a
nd Au
stra-lia
n co
als.
These experimental re
sult
s
help to unde
rstand a
nd predict the beh
avior of
coal
and MSW bl
end
s in pra
c
t
i
cal appli
c
atio
ns
[10]. A numerical solutio
n
wa
s prese
n
ted to
the con
s
trai
ned
non-li
nea
r o
p
timization
o
f
a
multivariable
obje
c
tive fun
c
tion utili
zing
Exce
l spre
a
d
sh
eet. The
method i
s
ca
pable
of han
dling
any num
ber
of sou
r
ce
co
als
with different si
ze
fracti
ons [1
1]. Co
mparative co
mbustio
n
stu
d
ie
s
were pe
rform
ed on p
a
rti
c
les of pulve
ri
zed
coal
sa
mples f
r
om t
h
ree
different
ran
ks: a hi
gh-
volatile bituminou
s co
al, a sub
-
bitumi
no
us coal, and t
w
o lignite
coa
l
s [12].
Siti et al. [13] investig
ate
d
the b
ehavi
our of
Malay
s
ian
sub-bitu
minou
s coal
(Mu
k
ah
Balingian), oil palm biomass (em
p
ty fruit bunches (EF
B
),
kernel shell (PKS)
and mesocarp fibre
(PMF)) and t
heir respectiv
e
blend
s du
ri
ng pyroly
si
s
usin
g therm
o
gravimetri
c a
nalysi
s
(TGA
)
.
The
study inv
e
stigate
d
the
comb
ustio
n
p
r
ofiles of
tob
a
c
co
stem, hig
h
-sulfur bitum
i
nou
s
coal
an
d
their ble
n
d
s
by thermo
gra
v
imetric a
nal
ysis.
Ignition
and b
u
rn
out
perfo
rman
ce
s, heat relea
s
e
perfo
rman
ce
s, and gase
o
u
s
pollutant e
m
issi
on
s we
re also stu
d
ie
d by thermo
gravimetri
c a
nd
mass
sp
ectro
m
etry an
alyses [1
4]. Th
e
comb
usti
o
n
b
ehavior an
d e
x
cess
heat
re
lease d
u
rin
g
t
h
e
oxy-fuel com
bustio
n
of blended coal
s were inve
stigat
ed experim
e
n
tally using a
non-i
s
othe
rmal
thermo
gravim
etric analy
z
e
r
.
For intera
ction
beh
aviors o
n
cha
r
acteri
stic te
mperature
s
,
the
volatile relea
s
e temp
eratu
r
e shows a
n
additive
beh
avior; however, ignition a
nd burnout t
e
m-
peratu
r
e
s
sho
w
non
-ad
d
itive behavio
rs f
o
r blen
ded
co
als [15].
In pre
s
e
n
t
study, the co
mbined
effect of
tran
sp
ortation an
d
multi-coal
s b
l
endin
g
comb
ustio
n
o
n
the total
co
st of po
we
r g
enerati
on
wa
s inve
stigate
d
ba
sed
on t
he meth
od of
coa
l
manag
eme
n
t and all
o
catio
n
in a p
o
wer
gene
ration
group. Th
e opti
m
ization
co
al
purcha
s
in
g a
nd
transpo
rtation
paths for ea
ch po
we
r pla
n
t in t
he gro
up we
re dete
r
mine
d by the introdu
ction
of
multi-coal
s bl
endin
g
com
b
ustion mo
del
and coal tran
spo
r
tation an
d allocation
model.
2. Frame
w
o
r
k of an Op
timal Allocati
on Model
Traditio
nal fuel coal tra
n
sp
ortation
model
often
use the lowe
st transpo
rt cost a
s
obje
c
tive fun
c
tion. But for a po
we
r ge
neratio
n g
r
ou
p, it is more
compl
e
x due
to many fact
ors,
such as deli
v
ery fee, arrival time, coal dema
nd, coal species,
auxilia
ry unit costs, and
environ
menta
l
requi
reme
nts. It is a multi-obje
c
tive
opt
imal issue. T
he wh
ole
pro
c
e
ss of fuel coal
prod
uctio
n
a
nd u
s
ing, f
r
o
m
minin
g
to
comb
ustio
n
, i
n
volves m
a
n
y
factors, an
d it is
sh
own
by
Figure 1.
Figure 1. Flow Ch
art of Co
al Flow Di
re
ct
ion
For the
pro
c
ess of coal
comb
ustio
n
i
n
a po
we
r
plant, usi
ng
multi-coal
s b
l
endin
g
technology may
affect combustion stability,
er
osion, fouling,
overte
m
perature of
heating
surfa
c
e
s
,
sl
aggin
g
in t
he furnace,
unit ope
ra
tion efficien
cy and i
n
creasi
ng el
ect
r
icity
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3728 – 37
36
3730
com
s
um
ption
of auxiliary equipme
n
ts. Thu
s
, th
e co
al spe
c
ie
s ad
apta
b
ility and the
comp
re
hen
si
ve benefits all
shoul
d be co
nsid
ere
d
in the model.
For the
pro
c
e
ss
of co
al tra
n
sp
ortation, t
he fi
nal lo
we
st comp
reh
e
n
s
ive price at th
e entry
of ea
ch
plant
is the m
a
in
goal. A
c
cordi
ng to
t
he
sel
e
cted
coal
speci
e
s by th
e allo
catio
n
and
blendi
ng syst
em
in ea
ch power plant,
the
opt
im
al
purch
asi
ng
scheme
can
be a
c
hi
eve
d
by
filtering co
al quality, sele
cting pat
h, and
predi
cting the
additional
co
st.
So, for the
whole
pro
c
e
ss of coal
allo
cation in
a
group, the
mod
e
l mu
st
con
s
i
der th
e
most suitable
coal
spe
c
ie
s and the b
e
st
transpor
t
a
tion path, and t
hen obtai
n th
e bigge
st be
n
e
fit
for the powe
r
gene
ration
group. The
optimal me
thod ro
ute of coal allo
cati
on in a power
gene
ration g
r
oup is
sho
w
n
by Figure 2.
Figure 2. Optimal Method
Route of Coal
Allocation in
a Grou
p
3. Optimizati
on Multi-co
a
l
s Blending
Model in a Po
w
e
r Plant
Multi-coal
s
blendi
ng mo
del ne
ed
consi
deri
ng
many facto
r
s, such
a
s
ignition,
comb
ustio
n
, slag
ging
cha
r
acteri
stics, a
nd is a
multi-obje
c
tive optimization i
s
su
e [6]. The model
can b
e
de
scri
bed by,
12
(
(
)
,
()
,
,
()
)
mi
n
y
x
x
x
p
ff
f
(
p<K
)
(1)
Subject
to:
12
()
(
(
)
,
()
,
,
()
)
0
xx
x
x
L
gg
g
g
(2)
Whe
r
e,
x
i
s
v
a
riabl
e vecto
r
,
f
(
x
) i
s
obje
c
t
i
ve function,
y
is
obje
c
tive function ve
ct
or,
g
(
x
)
is co
nst
r
aint condition.
3.1. Objectiv
e Functio
n
s
The
com
p
re
hen
sive e
c
o
nomic indi
ca
tors of p
o
wer
plant
we
re u
s
ed
a
s
obje
c
tive
function
s, incl
uiding
coal p
r
ice cost, ope
ration co
st an
d environ
men
t
al cost.
(1)
Coal p
r
ic
e
cost
If the standa
rd co
al p
r
ice
i
s
P
0
(Yua
n/ton), the ave
r
a
ge calorifi
c v
a
lue of bl
end
ed coal i
s
Q
avg
(kCal/kg
),
the price i
s
P
avg
,
the saving co
st und
e
r
a ce
rtain m
u
lti-co
als bl
e
nding meth
o
d
is
expre
s
sed a
s
,
10
7
000
avg
avg
P
PP
Q
(3)
(2) O
p
e
r
ation
cost
The op
eratio
n co
st mainly
inclu
d
e
s
the
power
con
s
u
m
ptions
of co
al pulveri
zin
g
system
and fan
sy
ste
m
. Assuming
the po
we
r
comsu
m
ption
per
wei
ght p
u
lverized
coa
l
for ea
ch
ki
n
d
of
fuel coal all i
s
s (kWh/ton
)
, the increa
sin
g
of operatio
n
cost is
cal
c
ul
ated by,
2
7000
(1
)
y
av
g
P
sP
Q
(4)
Whe
r
e, P
y
denotes the p
r
ice of electri
c
ity to power g
r
id
, Yuan/kWh.
(3) Environm
ental co
st
Environme
n
tal co
sts co
n
s
ist pri
m
arily
of
desulfuri
zation co
sts,
limestone costs
a
nd
se
wag
e
ch
arges. Assu
mi
ng the de
sul
f
urizatio
n e
ffic
i
enc
y
is
s
a
me for different units
. The
environ
menta
l
cost
s ca
n be
expresse
d a
s
:
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
An Optim
i
zation Model of
Coal Allo
catio
n
in a Grou
p (Jianji
an Zha
o
)
3731
2
3
(
700
0
/
)
SO
av
g
a
v
g
PU
S
Q
(5)
Whe
r
e,
2
SO
U
(Yua
n/ton) de
note
s
the fixed de
sulfuri
z
atio
n cost in
cludin
g
l
i
meston
e sl
urry
co
st and emi
ssi
on
s enviro
n
mental cha
r
ge.
S
avg
(%) is the average
sulfur.
Based
o
n
t
he a
bove th
ree
si
de
s, the o
b
je
ctive functio
n
of multi-coal
s blen
din
g
optimizatio
n model is a
s
followin
g
:
12
3
F
PP
P
mi
n
(6)
3.2. Cons
trai
nt Con
d
ition
s
The m
a
in
co
n
s
traint
conditi
ons a
r
e
all
co
me fro
m
coal
quality. The
n
online
a
r
rel
a
tinshi
ps
are u
s
ed to calcul
ate co
al quality, and a
r
e expe
re
sse
d
as follo
wing
:
Calorific value:
,,
,
,
,
Aq
i
i
i
i
i
i
B
Qf
X
Q
M
A
V
F
Q
(7)
Sulfur:
,
As
i
i
B
Sf
X
S
S
(8)
Mois
ture:
,,
,
Am
i
i
i
i
B
M
fX
A
V
F
M
(9)
Volatile matter:
,,
,
,
Av
i
i
i
i
i
B
Vf
X
M
A
V
F
V
(10)
As
h:
,,
,
Aa
i
i
i
i
B
A
fX
A
V
F
A
(11)
Ash melting p
o
int:
As
t
ST
f
(X
i
)
B
ST
(12)
Whe
r
e,
A
is l
o
wer limit,
B
is up
per li
mit,
i
is ea
ch
co
al
sampl
e
. And
f
is no
nlinea
r f
unctio
n
of ea
ch
variable, an
d can b
e
cal
c
ul
ated by the neural n
e
two
r
k method.
4. Transpor
tation Model
4.1. Objectiv
e Functio
n
s
In orde
r to o
b
tain the obj
ective fun
c
tions of coal transportatio
n
model, the f
o
llowin
g
factors shoul
d be co
nsi
dered.
(1) T
r
an
sp
ort
a
tion dista
n
ce
The tran
spo
r
tation di
stan
ce impa
cts di
rectly on
tran
spo
r
tation
co
st an
d tra
n
sp
ortation
time, which
were ge
ne
rally a linear
relati
onship with transport di
sta
n
ce.
(2) T
r
an
sp
ort
a
tion co
sts
For the e
c
o
nomic inte
re
st of the power
g
ene
rati
on gro
up, redu
cing the
cost of
transportation will bring
co
al cost reduction.
(3) T
r
an
sp
ort
a
tion time
Coal
needs t
o
be delivered to
power plant within the sti
pul
ated t
i
me; otherwi
se it will
affect the normal power
generation. Meanwhile
, the
shortening of
the transport time will reduce
the co
st of the whol
e tran
sport process.
The optim
al
obje
c
tive is t
he minim
u
m
total co
st and
the sh
orte
st
transpo
rt time, and i
s
expre
s
sed a
s
:
(13)
Where,
and
are weig
hing co
e
fficients,
is adju
s
ting co
efficient,
,1
k
ii
C
is the
transpo
rtation
expen
se
fro
m
no
de
i
to
nod
e
1
i
by t
r
an
spo
r
tation
mod
e
l
k
.
kl
i
p
is
the
transpo
rtation
expen
se at
node
i
by the
transpo
rtatio
n model
co
n
v
erting from
k
to
l
;when
,1
k
ii
x
is 1, it mean
s the tra
n
sportation model
use
d
is
k
from node
i
to node
1
i
;when
,1
k
ii
x
is
,1
,
1
,1
,
1
k
k
kl
kl
ii
ii
i
i
ik
i
k
l
k
k
kl
kl
ii
i
i
i
i
ik
ik
l
M
in
Z
x
C
y
p
xH
y
t
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3728 – 37
36
3732
0, it means n
o
t using tran
sportation m
o
del
k
;whe
n
kl
i
y
is 1, it means t
he tran
sp
orta
tion model
c
onverts
from
k
to
l
at no
de
i
; when
kl
i
y
is 0, it mean
s the tra
n
sp
ortation m
o
d
e
l
k
will
conve
r
t to another o
r
not coverting.
4.2. Cons
trai
nt Con
d
ition
s
,1
1
k
ii
k
x
(14
)
1
kl
i
kl
y
i
(15
)
1,
,
1
2
kl
k
l
ii
i
i
i
x
xy
,,
ik
l
(16
)
1
0,
0
(17
)
,1
,1
k
k
kl
kl
ii
ii
i
i
ik
ik
l
x
Hy
t
T
(18
)
,1
,0
,
1
kk
l
ii
i
xy
,,
ik
l
(19
)
Whe
r
e, Equ
a
t
ion (1
4) i
ndi
cate
s only
on
e
mod
e
of
transportatio
n
can be used betwe
en
two tran
spo
r
t
a
tion node
s; Equation (1
5) indica
te
s onl
y one transfe
r model can
be used in o
ne
node; Eq
uati
on (16) is
used to e
n
sure
the con
s
is
te
ncy of the
whole tran
spo
r
t pro
c
e
ss, if t
he
transpo
rtation
model
conve
r
ts from
k
to
l
at
nod
e
i
, the
transpo
rtation
model
u
s
e
k
fr
o
m
th
e n
ode
i
-1 to node
i
;
Equation (1
7) indi
cate
s the sum of th
e
weight coe
fficients is 1;
Equation (1
8)
indicates the
total transp
o
rtation time sh
ould be
le
ss
than the spe
c
ified latest a
rrival time; and
Equation (19) means that t
he value
s
of variable
s
a
r
e
limited to 0 or 1.
5. Optimizati
on Model an
d Soft
w
a
re S
y
stem of Coal Allocation
in a Group
Based on
the above me
ntioned mo
de
ls, an optimi
z
ation model o
f
coal allo
cati
on in a
power g
ene
ration group
can
be o
b
ta
ined. In o
r
d
e
r to p
r
ovid
e su
ppo
rt a
nd gui
deline
of
engin
eeri
ng
pra
c
tice,
a
software
syst
em i
s
d
e
sin
ed b
a
sed
on
the o
p
timization mo
del.
The
sof
t
w
a
r
e
sy
st
em u
s
ing B
/
S
st
ru
ct
ur
e, u
s
ers can e
a
sil
y
visit it by
using
a web b
r
owse
r. And t
he
netwo
rk top
o
l
ogy dia
g
ra
m
of it is sho
w
n
in
Fi
gure 3. The software
syste
m
wa
s desi
gne
d
u
s
i
ng
modula
r
i
z
atio
n method, an
d the main m
odule
s
are sh
own in Fig
u
re
4.
The main fun
c
tion
s of the system a
r
e a
s
followi
ng
s:
(1)
Group
u
n
ified de
ploy
ment mod
u
le
. It is a key
deci
s
io
n-m
a
king
pro
c
e
ss and i
s
divided into the multi-coal
s blen
ding d
e
c
isi
on an
d tra
n
sp
ortation
ro
ute optimizati
on de
cisi
on.
Figure 3. System Netwo
r
k Topolo
g
y
i
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
An Optim
i
zation Model of
Coal Allo
catio
n
in a Grou
p (Jianji
an Zha
o
)
3733
Figure 4. System Functio
n
Diag
ram
(2) P
o
wer
pl
ant real-tim
e
status mo
dul
e.
It
is used to
sh
ow
unit
operation dat
a,
flow
dire
ction of coal, store
d
co
al status in th
e coal
yard a
nd ra
w co
al d
y
namic st
ratified inform
atio
n.
Unit o
peratio
n pa
ram
e
ters: di
splayin
g
the m
a
in
o
peratin
g p
a
rameters, the
cu
rrent
prog
ram
of coaling, o
pera
t
ion
optimization pa
ram
e
te
rs
of co
al
pul
verize
r a
nd p
r
oviding
ala
r
m
whe
n
there a
r
e exceptio
n d
a
ta.
Gene
ral situ
a
t
ion of coal yard: displayin
g
two-dim
e
n
s
ional map an
d three-dime
nsio
nal
map of coal y
a
rd, an
d the
name of
coal,
coal q
uality informatio
n, st
ored tim
e
, he
ight, area
s
ca
n
be displayed
on ea
ch
coal
dump. Wh
e
n
the co
al
yard op
erato
r
s finish coal pil
i
ng and
coali
ng,
they will upda
te the store
d
coal d
a
ta of coal yard.
Ra
w
coal
dy
namic:
re
al-ti
m
e tra
c
king
of diffe
re
nt ki
nds of
coal
enterin
g into
the coal
yard, and
cal
c
ulatin
g co
rre
s
po
ndin
g
co
a
l
height, layers, wei
ght and
coal q
uality data. Wh
en the
coal i
s
about
to chang
e, this mod
u
le can remi
nd th
e field operator of ope
rati
onal state of
the
main co
al pul
verize
r.
(3)
Data
entry module: thi
s
mod
u
le i
s
mainly
used t
o
provid
e gro
up fuel de
pa
rtment and
power plant staff
with
ente
r
ing ch
annel
of
coal
qu
ality information
and
real
-tim
e tran
sp
ort
cost.
The users can
ente
r
correspon
ding
inform
atio
n
at
spe
c
ifie
d lo
cation.
The
system
will
automatically update the d
a
t
abase to ensure
the a
c
curacy of deploy
ment re
sults.
(4) Data q
uerying mod
e
l:
this
mo
dule
i
s
u
s
e
d
to provid
e users
with daily
statisticalstat
e
ments,
which in
clud
es hi
stori
c
al
co
al
transpo
rt rout
e, hi
sto
r
i
c
al coal
pu
rcha
si
ng
amount, tran
sport co
st curv
e, pie cha
r
t of
powe
r
plant
store
d
co
al a
m
ount and
so
on.
(5)
Data
set
module: thi
s
module i
s
m
a
inly use
d
to p
r
ovide o
p
e
r
ati
on setup fun
c
tion for
privilege
d u
s
ers. It is
used
to set ble
ndi
ng bo
und
ary
con
d
ition
s
, transport o
p
tim
i
zation
bou
nd
ary
con
d
ition
s
, operatin
g pa
ra
meter ala
r
m value,
user inf
o
rmatio
n permissi
on an
d so on.
6. Results a
nd Analy
s
is
6.1. Information Databas
e
and Algorithm
The information a
bout
20
kin
d
s of
coal
s i
s
sh
o
w
n i
n
T
able
1, and
it wa
s u
s
e
d
a
s
a
databa
se to
cal
c
ulate the
optimization
model of
co
al allocation
in a gro
up. And the gen
etic
algorith
m
[7] wa
s used to solve the abov
e mentione
d model
s.
T
h
re
e ki
n
d
s of
coal
were
sele
cted
to be blen
d
ed in the o
p
t
imization p
r
oce
s
s, the
sea
r
ching
ran
ge of the ratio
is from 10% to 90% and th
e accuracy of
the ratio is 1
%
.
The co
nst
r
ain
t
condition
s were set as foll
owin
g:
Calorific value
:
50
25
≤
Qn
et
≤
5735(kCal/k
g)
(20)
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3728 – 37
36
3734
Moisture
:
0<M
≤
7(%)
(21)
Volatile
:
23
≤
V
≤
28(%
)
(22)
Sulfur
:
0<
S
≤
1.3(%)
(23)
As
h
:
0<
A
≤
19(%)
(24)
The
stand
ard
co
al p
r
ice i
s
assum
ed
as 100
5
yuan/t
on, the
elect
r
ic
ity co
nsum
ption of
auxiliary is 11.5kWh/t, the desulfurization fixed co
st
is 100 yuan/t, desulfuri
zat
i
on efficiency is
90%, and the
electri
c
ity pri
c
e is 0.4 yua
n
/kWh.
Table 1. Co
al
Information
Datab
a
se
Coal
Number
Heat
(k
Cal
/
k
g
)
Moisture
(%)
Volatile
(%)
Ash
(%)
Sulfur
(%)
Price
(Y
u
an/t)
01 5032
1.13
18.27
29.54
1.48
580
02 4398
13.53
35.84
2.02
0.41
450
03 4979
8.56
24.77
10.39
0.21
560
04 4843
9.69
27.43
8.70
0.25
530
05 4742
10.66
30.66
6.96
0.29
490
06 5539
9.76
29.27
7.56
0.27
630
07 5333
5.31
33.53
16.70
0.39
595
08 5433
5.08
25.45
16.61
0.87
610
09 5029
6.03
28.21
15.14
0.78
570
10 4572
11.70
41.99
4.73
0.10
460
11 5597
1.67
11.87
22.70
0.44
640
12 5825
1.43
14.93
21.50
0.35
670
13 5709
0.78
15.72
21.82
1.08
650
14 6167
1.36
18.79
20.25
1.16
750
15 4665
0.83
23.45
31.83
2.98
480
16 6150
2.06
10.00
17.39
0.25
690
17 4761
9.63
32.31
6.46
0.68
500
18 4780
8.07
29.14
13.79
0.25
520
19 4831
8.24
28.41
12.86
0.24
525
20 4924
8.43
26.77
11.40
0.22
540
6.2. Calculated Res
u
lts o
f
Optimi
za
tio
n
Multi-coals
Blending M
odel
The ge
netic
algorith
m
wa
s used to solv
e the mo
del and the
Matlab software
wa
s
applie
d in th
e process
of cal
c
ul
ation.
The m
a
ximu
m num
ber of
iteration
s
was
sel
e
cte
d
as
terminatio
n condition
s. It was set as 50
0
.
Part of
the computation
a
l results is
sho
w
n in Tabl
e 2
.
Table 2. The
Operation Re
sults of Ge
ne
tic Algorithm
No. Individual
code
Heat
value
(k
Cal
/
k
g
)
Water
content
(%)
Fugitive
constituent
(%)
Ash
content
(%)
Sulfur
content
(%)
Target
value
1
1
19
11 34 33
33
5151
3.66
19.504
21.78
0.72
115.1
2
20 19
1
30 27
43
4943
5.24
23.588
19.59
0.767
111.1
3
11 7
2 40
40
20
5252
5.498
25.328
16.17
0.412
170.2
4
10 16
1
34 37
29
5290
5.068
23.275
16.61
0.555
162.7
5
2 20
6 39
22
39
4957
10.94
31.282
6.244
0.314
196.2
6
17 4
3 35
47
18
4838
9.566
28.659
8.22
0.393
186.6
7
11 20
1
38 36
26
5209
3.964
18.898
20.41
0.631
127.9
8
20
8
13 34 31
35
5357
4.714
22.493
16.66
0.723
124.8
6.3. Calculated Res
u
lts o
f
Trans
porta
tion Model
The ge
netic
algorith
m
wa
s also u
s
ed
in t
he solvin
g of the mo
del. For exa
m
ple, the
transpo
rtation
network of coal to a powe
r
plant
is assumed a
s
Fig
u
re 5. There are thre
e mo
des
of tran
sp
ort
o
p
tions:
rail
tra
n
sp
ortation,
road t
r
an
sp
ort
a
tion a
nd
se
a
tran
sp
ortatio
n
. The
cost
s
of
variou
s tran
sportation
mo
des a
r
e
sh
o
w
n i
n
T
able
3, the
nee
d t
i
mes of va
rio
u
s t
r
an
sp
orta
tion
mode
s are shown in Tabl
e 4,
the fee
of transshi
p
m
ent betwe
en
variou
s tran
sportation m
o
des
are
sh
own in
Table
5, an
d the tan
s
shi
p
ment
time
s
betwe
en va
ri
ous t
r
an
sp
ort
a
tion mo
de
s
are
sho
w
n in Ta
b
l
e 6.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
An Optim
i
zation Model of
Coal Allo
catio
n
in a Grou
p (Jianji
an Zha
o
)
3735
Figure 5. Tra
n
sp
ortation
Network
Table 3. Unit
Tran
sp
ortatio
n
Co
sts of Va
riou
s Tra
n
spo
r
t Mode
s (yua
n/t)
Transpo
rtation M
ode
Coal mine - 1
1-2
2-3
3-4
4- po
w
e
r plant
Rail
62
39 40 45
53
Road
80
24 36 51
70
Sea
M
35 37 48
49
Whe
r
e, 1, 2, 3, 4 all respect differe
nt
transship
me
nt points. M mens that the se
a
transpo
rtation
cann
ot be used.
Table 4. The
Tran
sp
ortatio
n
Times of th
e Variou
s Tra
n
sp
ortation M
ode
s (day
s)
Transpo
rtation M
ode
Coal mine - 1
1-2
2-3
3-4
4- po
w
e
r plant
Rail
1.6
1.0 1.5 1.8
2.0
Road
2.3
0.8 1.2 1.5
2.0
Sea
M
1.1 1.3 1.7
1.8
Table 5. Tran
sshipme
n
t Ch
arge
s bet
wee
n
Variou
s Tra
n
sp
ortation M
ode
s (yuan/t)
Transpo
rtation
M
ode
Rail Road
Sea
Rail 0
5
4
Road
5
0
4
Sea 4
4
0
Table 6. The
Times of Tra
n
sshipm
ent b
e
twee
n
V
a
rio
u
s Tr
an
spo
r
t
Mode
s (d
ay
s)
Transpo
rtation
M
ode
Rail Road
Sea
Rail 0
0.7
0.5
Road
0.7
0
1
Sea 0.5
1
0
The Matlab software
wa
s use
d
to acco
mplish the op
timization pro
c
e
ss. The
co
ndition
s
were s
e
t as
following:
α
=0.7
(25)
β
=
0
.3
(26)
σ
=
10
(27)
And the cro
s
sover p
r
o
bab
ility was set as 0.8, mutation pro
babilit
y was set a
s
0.005, and the
whol
e tran
sp
ortation coal
wa
s set a
s
15
,000 tons.
The optimization result
of
t
he
tran
spo
r
ta
tion
path
is 1
-
1-2-3
-
3.
Tha
t
mean
s th
at
the rail
transpo
rtation
is use
d
from
coal mine to
point
2, the road tran
spo
r
tation sho
u
ld
be use
d
fro
m
point 2 to point 3, and the sea tra
n
sport
a
tion is
appli
e
d from point 3 to the powe
r
plant. The total
optimizatio
n goal value i
s
197.1, th
e frequ
en
cy
of operatio
n iteration i
s
51, the total
transpo
rtation
cost is 2
43 yuan/ton, and
t
he total transportation time
is 9 days.
7. Conclusio
n
An optimai
ztion sy
stem f
o
r the
co
al
a
llocation in
a po
we
r g
eneration g
r
oup
wa
s
develop
ed ba
sed
on the m
u
lti-co
als
ble
nding te
ch
n
o
l
ogy model,
coal tran
sp
ort
a
tion mod
e
l
and
the relevant
algo
rithm
s
. The
sy
ste
m
ad
opts fri
endly
softwa
r
e
stru
ctu
r
e
and
ca
n
pro
v
ide
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NI
KA
Vol. 12, No. 5, May 2014: 3728 – 37
36
3736
person
a
lized
function
s fo
r the po
we
r
gene
ra
tion g
r
oup and po
wer plant
s.
A
comp
utatio
nal
example sh
o
w
that
the op
timization re
sults
can
be
searche
d
o
u
t by the ge
neti
c
alg
o
rithm
i
n
a
very sho
r
t pe
riod of time, but the results co
ntai
n so
me sub
optim
al solution
s. In the practi
cal
appli
c
ation, t
he bala
n
ce b
e
twee
n search sp
eed a
nd
pre
c
isi
on
can
be a
c
hieved
by cho
o
sin
g
a
more a
ppreci
a
te algorith
m
according to the req
u
ireme
n
ts of com
put
ing time.
Ackn
o
w
l
e
dg
ements
This resea
r
ch is currently supp
orte
d b
y
the Nation
al Natu
ral Scien
c
e F
oun
dation of
Chin
a (No.51
3760
66, 508
0
6022
), the sp
ecial fou
ndat
i
on of Pearl Ri
ver Ne
w Star
of Scien
c
e an
d
Tech
nolo
g
y in Guang
zh
o
u
City (No. 2012J220
0
02), the Chi
na Schol
arship Cou
n
cil
(No.
2012
0844
007
2) a
nd Key
Labo
rato
ry of
Efficient an
d Cl
ean E
n
e
r
gy Utili
zatio
n
of G
uan
gd
ong
High
er Edu
c
a
t
ion Institutes
,
South Chin
a University of technolo
g
y (No.KLB10
0
0
4
).
Referen
ces
[1]
Peng
T
,
Herui
C. Econ
omic
Gro
w
t
h
a
nd
Coal
Co
nsum
ption
in
Ch
ina
.
T
E
LKOMNIKA Indo
nesi
a
n
Journ
a
l of Elec
trical Eng
i
ne
eri
n
g
. 201
3; 11(3)
: 1449-1
4
5
5
.
[2]
T
i
ejun C, Xi
ya
ng L. T
he Study
of Se
lectio
n
of Coal-F
ir
ed S
upp
lier i
n
T
hermal Po
w
e
r Ent
e
rprise B
a
se
d
on the E
x
tensi
on An
a
l
ysis Method.
T
E
LKOMNIKA Indone
sian Jo
urna
l of Electrical Eng
i
neer
ing
. 2
013;
11(7): 38
74-
38
85.
[3]
Moha
nta S, Ch
akrab
o
rt
y
S, M
e
ika
p
BC. Pre
d
ictio
n
of Econ
omic Oper
atin
g Co
nditi
ons fo
r India
n
Co
a
l
Prepar
ation Pl
ants.
F
uel Proc
essin
g
T
e
chn
o
l
ogy.
201
1; 92 (
9
): 1696-
17
00.
[4]
Xi
jin G, Ming C
,
Jia
w
e
i
W
.
Co
al Ble
ndi
ng Op
timizatio
n
of Coal Pre
par
ation
Productio
n
Pr
ocess Base
d
on Improve
d
GA.
Procedia E
a
rth and Pla
neta
r
y Science
. 20
09; 1(1): 65
4-6
60.
[5]
Gupta V, Mo
hant
y MK.
Co
al Pr
epar
atio
n
Pla
n
t Op
timi
zation: A
Criti
c
al R
e
vi
e
w
of the E
x
istin
g
Methods.
Inter
natio
nal J
ourn
a
l of Miner
al Pr
ocessi
ng
. 20
06
; 79(1): 9-17.
[6]
Jin SH, K
w
a
n
g
YL. A Mu
ltia
gent-S
ystem-B
ased Int
e
ll
ige
n
t
Referenc
e G
o
vern
or for M
u
lti Obj
e
ctive
Optimal Po
w
e
r
Plant Operati
o
n.
IEEE Transactions on Ener
gy Conv
ersion
. 2008; 2
3
(4): 1
082-
109
2.
[7]
Gupta V, Moh
ant
y
M, Ma
ha
j
an A, Bis
w
a
l
SK.
Genetic Al
gorithms - A N
o
vel T
e
chni
qu
e to Optimiz
e
Coal Pr
epar
ati
on Pla
n
ts.
Internatio
nal J
ourn
a
l of Miner
al Pr
ocessi
ng
. 20
07
; 84(1-4): 133-
143.
[8]
T
i
an
y
a
ng
C, H
ong
jia
n Z
,
Yo
n
g
Y, H
o
n
g
li
an
g
Z
,
Han
g
Z
.
In
vestigati
ons
int
o
the
Ign
i
tio
n
Behav
iors
of
Pulveriz
ed
Co
a
l
s a
nd
Coa
l
Bl
e
nds
in
a Dr
op
T
ube F
u
rnace
Using
F
l
am
e M
onitor
i
ng
T
e
chniq
ues.
Fu
el
.
201
0; 89(3): 74
3-75
1.
[9]
Jian
guo Z
,
F
en Z
.
Sulfur Di
oxid
e Emissio
n
Co
mb
in
ation
Predictio
n Mo
del of C
h
in
a T
hermal P
o
w
e
r
Industr
y
.
T
E
LK
OMNIKA Indon
esia
n Journ
a
l o
f
Electrical Eng
i
ne
erin
g
. 201
3; 11(3): 115
7-1
165.
[10]
Marisam
y
M,
T
o
moaki N,
Kuni
o Y. C
harac
ter
i
stics
of Co-com
bus
tion a
n
d
Kin
e
tic Stud
y o
n
H
y
droth
e
rmal
l
y
T
r
eated Mun
i
cipa
l S
o
li
d W
a
ste
w
i
t
h
D
i
fferent
Rank
C
oals: A
T
hermogravim
etric
Anal
ys
is.
Appli
ed Ener
gy
. 201
0; 87(1): 14
1-1
48.
[11]
Moha
nta S, M
i
shra BK, B
i
s
w
a
l
SK. A
n
E
m
phas
is o
n
O
p
timum F
u
e
l
Producti
on for
Indi
an
Coa
l
Prepar
ation Pl
ants T
r
eating Mu
ltipl
e
Co
al S
ources.
F
uel
. 2
010; 89 (3): 7
7
5
-78
1
.
[12]
Yian
nis AL, Kulb
hush
an J, Reza K, Adel
F
S
.
Combustio
n
Behav
ior in
Air of Singl
e Particles fro
m
T
h
ree Differen
t
Coal Ra
nks
and from Su
g
a
rcan
e Bag
a
s
s
e.
Co
mbusti
o
n
and F
l
a
m
e
. 201
1; 158(
3):
452-
465.
[13]
Siti SI, Norazah AR, Khudzir I, Azil BA,
Z
u
lkifli
AR. Investiga
t
ion on T
hermo
chemic
al Beh
a
v
iour of Lo
w
Rank M
a
la
ys
ia
n Co
al, Oil P
a
lm Biom
ass a
nd T
heir Bl
en
ds dur
ing
P
y
ro
l
y
s
i
s via
T
hermogravim
etric
Anal
ys
is (T
GA).
Bioresourc
e
T
e
chn
o
lo
gy
. 201
0; 101(1
2
): 458
4–4
59
2.
[14]
Kaih
ua Z
,
Kai
Z
,
Yan C, Weipi
ng P. Co-c
om
busti
on C
h
aracteristics a
nd Ble
n
d
i
ng O
p
timizati
on of
T
obacco
Stem and Hig
h-su
lfur
Bitumin
ous C
oal
B
a
se
d o
n
T
hermo
gravim
etric an
d Mass S
pectrometr
y
Anal
ys
es.
Bior
esourc
e
T
e
chn
o
lo
gy
. 201
3; 131 (3): 325-
33
2.
[1
5
]
Yo
n
m
o
S,
C
h
eo
re
on
M, Se
on
gy
oo
l
A, Gy
un
g
m
in
C, D
u
c
k
jool
K. E
x
per
i
m
ental Stu
d
y
on Inter
a
ctio
n
and E
x
cess
He
at Rele
ase u
n
d
e
r Oxy-fue
l
Co
mbustio
n
of Bl
end
ed C
oals.
Korea
n
Jour
nal
of Che
m
ica
l
Engi
neer
in
g
. 2013; 30(
2): 337
-344.
Evaluation Warning : The document was created with Spire.PDF for Python.