TELKOM
NIKA
, Vol. 11, No. 5, May 2013, pp. 2648 ~
2656
ISSN: 2302-4
046
2648
Re
cei
v
ed
Jan
uary 17, 201
3
;
Revi
sed Ma
rch 1
6
, 2013;
Acce
pted Ma
rch 2
5
, 2013
Demand Discipline of Air Transportation for
Passen
gers
Yunqian Qu,
Jinfu Zhu
Coll
eg
e of Civil
Aviation, Na
nji
ng Un
iversit
y
o
f
Aeronautics,
Astronautics
Nanj
in
g, 210
01
6, Chin
a
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: qu
yun
q
i
ann
u
aa@
163.com
A
b
st
r
a
ct
T
he Yan
g
t
z
e
River D
e
lta (Y
RD) regi
on i
n
Chin
a is take
n
for an exa
m
p
l
e to rese
arch
de
man
d
discip
l
i
ne of ai
r transportatio
n
for passe
ng
ers in this p
a
p
e
r. T
he discip
l
i
ne inc
l
ud
es thr
ee parts: de
ma
n
d
gen
eratio
n, d
i
s
t
ributio
n
and
fl
ow
. F
i
rst of al
l, the k
e
y fact
ors infl
uenc
in
g
d
e
man
d
gen
era
t
ion
are
got
w
i
th
infor
m
ati
on
ent
ropy. A thres
h
old
is set to c
h
oose
more
i
m
p
o
rtant factors a
nd w
e
ig
hts of t
hese fact
ors ar
e
calcul
ated
at the sa
me ti
me.
Contri
butio
n ra
te of a factor is define
d
to
an
a
l
y
z
e
the d
i
ffere
nce of the fact
ors
for different cit
y
. By compari
n
g de
man
d
distr
i
buti
on
w
i
th a
i
r
ports
’
thro
ug
hp
ut, w
e
can an
a
l
y
z
e
the
de
ma
nd
flow
. T
here are some co
ncl
u
si
ons. T
he factors have diffe
re
n
t
contributio
n rate to different
city for deman
d
gen
eratio
n. Air
c
raft move
me
nt is t
he most imp
o
rtant fact
or influ
enci
ng
de
ma
nd flow
. It provides
mo
re
accurate basis
for m
a
rk
et positioning and dev
elop
ing, airport layout in a
m
u
lti-airport system
.
Ke
y
w
ords
: air
transportatio
n
;
deman
d disci
p
line; i
n
for
m
atio
n entropy
Copy
right
©
2013 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
Dema
nd
of ai
r tra
n
sportati
on in
a
regi
on
is th
e b
a
si
s
of the pl
annin
g
an
d d
e
velo
pment of
an ai
rpo
r
t. In
a region
la
cki
ng in
ai
r tra
n
sportation
de
m
and, it
won’t
have a
n
y be
n
e
fits even
if
we
provide mo
re
and better
air tran
spo
r
ta
tion serv
i
c
e. So demand
analysi
s
is o
ne of the most
importa
nt work of
air t
r
an
sp
ortation
devel
opment.
Dem
and
of air tra
n
sp
ortation
is clo
s
ely
relat
e
d
to so
cial a
c
ti
vity, economi
c
devel
opme
n
t, env
ironm
ent and
avia
tion su
pply. Dema
nd of
air
transpo
rtation
for pa
sseng
ers
have b
e
e
n
re
sea
r
ch
e
d
in lots of bi
bliographi
es [
1
-5]. But the
s
e
resea
r
ch is mainly about
demand g
e
neratio
n.
With contin
uou
s developme
n
t
of cities and
airpo
r
ts, mult
i-airport
syst
ems h
a
ve a
ppea
red i
n
Chin
a. A mul
t
i-airp
ort sy
st
em mea
n
s
many
airpo
r
ts
exist in
a regi
on of clo
s
ely-relate
d
citi
e
s
. In th
e multi-airp
ort
system,
anal
ysis
of dem
a
nd
discipline of
air transpo
rtation sh
oul
d cont
ain th
ree pa
rts: d
e
mand g
ene
ration, dema
nd
distrib
u
tion
and dem
and
flow. This paper m
a
i
n
ly analyze
s
demand di
sci
pline of
air
transpo
rtation
for
pa
sseng
ers,
which p
r
ovide
s
a m
o
re
a
c
curate
ba
sis for th
e pla
nnin
g
a
n
d
developm
ent of airpo
r
ts in the increa
sin
g
l
y multi-airpo
r
t systems.
2. Factor
s Choosing and
Weight
Calc
ulation
In orde
r to ref
l
ect problem
s more
comp
rehen
sively a
nd accu
rately
, we nee
d to con
s
id
er
many variabl
es in the research of e
c
o
nomy
and m
anag
ement. It increa
se
s t
he com
p
lexity of
analysi
s
inevi
t
ably and make
s us differe
nt to find
the
real characte
ristics an
d inhere
n
t discipl
i
ne
of things. In previou
s
bi
bliographi
es,
princi
pal
co
mpone
nt an
alysis i
s
often adopte
d
[6-7].
Ho
wever, thi
s
way cann
o
t
be applied
to every ca
se. And whe
n
function lo
ad of prin
cip
a
l
comp
one
nt is positive or
negative, the
meaning
of comp
re
hen
si
ve evaluation
function is
no
t
clea
r and n
a
m
ed cla
r
ity is low.
Con
c
e
p
t of Entropy is o
r
ig
inated from t
her
m
odynami
cs,
which ref
e
rs to th
e de
gree
of
cha
o
s in the
system. It is extende
d a
more spe
c
ific co
ncept in
different su
bject
s
and i
s
an
importa
nt parameter in dif
f
erent fields.
Using
the
concept of thermo
dy
nami
c
s for refe
ren
c
e,
Clau
de Elwo
od Shan
non
named th
e a
v
erage
amo
u
n
t of informa
t
ion as i
n
formation ent
ro
py,
excludi
ng the redu
ndan
cy. And he g
o
t the mathem
atical e
q
u
a
tion for information ent
ropy
cal
c
ulatio
n. If the small
e
r in
formation
ent
ropy of so
me
factor i
s
, the
more i
n
form
a
t
ion is
suppli
e
d
,
and
th
e larg
e
r
function this factor is in the evaluat
ion. So the greate
r
its we
ight is.
Therefo
r
e, the
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NIKA
Vol. 11, No
. 5, May 2013 : 2648 – 265
6
2649
way to confirm weight ba
sed on info
rm
ation entro
py belon
gs to ob
ject
ive weig
ht. It is supe
rior to
other ways
in mos
t
c
a
s
e
s
[8].
Based
on
inf
o
rmatio
n ent
ropy, this p
a
p
e
r
c
onfi
r
ms factor weight
and give
s
a
minimum
weig
ht thre
sh
old. If weight
of a facto
r
is
smalle
r tha
n
the minimum
weig
ht thre
shold, the fa
ctor
will be ignored in order to filter
the criti
c
al factor. The following are the concrete steps [9]:
A multi-attribute d
e
ci
sio
n
con
s
ist
s
of
n o
b
ject
s a
n
d
m fa
ctors.
a
ij
is an
ori
g
i
nal value
of t
h
e
obje
c
t
i
in the fac
t
or
j
, which forms
dec
i
sion matrix
A
=(
a
ij
)
n*m
. Then
stand
ardi
ze it
, namely
A’
.
Do normali
zation for
A’
, and get matri
x
R
=(
r
ij
)
n*m
,
''
=1
=
n
ij
ij
i
j
i
ra
a
Calculate i
n
formatio
n en
tropy
E
j
of the jth factor,
1
ln
n
ji
j
i
j
i
E
kr
r
,
1l
n
kn
.
Calculate
weig
ht of the jth factor
w
j
,
1
(1
)
(
1
)
m
jj
j
j
wE
E
, and
01
j
w
,
1
1
m
j
j
w
。
Filter factors acco
rdin
g to weight
s, if
0.
0
1
j
w
,
delete the jth factor an
d ret
u
rn to
. Cal
c
ulate
weig
hts of th
e re
st facto
r
s again u
n
til weights
of
all the re
st facto
r
s are mo
re t
han 0.0
1
, whi
c
h
mean
s these factors
are effective.
To calculate
acc
o
rding to the s
t
eps from
to , we can get
c
o
rresponding weights
and
the key facto
r
s. The process of
cal
c
ulatio
n is simpl
e
, and the found
a
t
ions of filtering is cl
ear.
3. Demand G
e
nera
tion an
d Distribu
tio
n
It assum
e
s th
at the real
nu
mber
of air trans
portatio
n
for pa
sseng
e
r
s in
som
e
re
gion is
equal to it
s d
e
mand. A
c
co
rding
to the
key factors a
n
d
their
wei
g
h
t
s ch
oo
sing
from the
form
er
part, we
can
get the com
p
rehe
nsive ev
aluation valu
e
V
of each region
al unit a
bout dema
n
d
of
air tran
sp
orta
tion for passe
ngers.
Fa
ctor weight
s have
been given a
c
co
rdi
ng to their inform
atio
n,
so we ado
pt simple
weig
ht method for a
numbe
r of cri
t
ical facto
r
s o
f
each region
al unit.
'
=1
=
ij
ij
n
ij
i
x
x
x
n
(1)
'
m
ij
ij
j=
1
V=
w
x
(2)
First, a
c
cordi
ng to form
ula
(1),
we
ca
n
get the o
r
igin
al data n
ondi
mensi
onli
z
ed
by usin
g
equali
z
ation
method. Equ
a
lizatio
n method ca
n not
only eliminate the dimensio
n, but also
maintain th
e
distrib
u
tion of
values.
ij
x
is re
spe
c
tively the
origi
nal valu
e of re
gion
al
unit
i
in factor
j
, and
'
ij
x
is the value of of regional u
n
it
i
in factor
j
afte
r usin
g equ
ali
z
ation meth
o
d
. This pap
er
need
s to di
stribute the tot
a
l dema
nd of
the regio
n
b
e
long
regio
n
al units. So
we can turn
th
e
comp
re
hen
si
ve evaluation
value to a rate of comp
reh
ensive eval
ua
tion [10].
(t)
1
n
ii
i
i
VV
V
(3)
In formula (3),
i
V
is the dem
and sha
r
e of
air tran
sp
orta
tion for pa
ssenge
rs
of uni
t
i
in
the
total dem
and of
the
whole regi
on, namely dema
nd sha
r
e.
So
we ca
n
g
e
t
th
e
de
mand
of air
transpo
rtation
for passen
g
e
r
s of ea
ch uni
t.
=
i
i
DD
V
(4)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Dem
and Di
scipline of Air Tran
spo
r
tation
for Passen
ge
rs (Y
unqia
n
Qu)
2650
In formula (4),
i
D
is the dema
nd of air tran
spo
r
tation for passeng
ers for unit
i
,
D
is the
the total demand of the wh
ole regi
on.
4.
Application
4.1. Case Introduc
tion
Core a
r
ea
of
YRD in
Chi
na i
s
a
su
per urb
an
agglo
m
eratio
n. It h
a
s th
e core
city of
Shangh
ai, in
cludi
ng
other 15
citie
s
of
Na
njing,
Ha
ngzhou,
etc.
In 20
09, Y
R
D ha
d ten
civil
airpo
r
ts, formi
ng a a multi-a
i
rpo
r
t system.
S
hangh
ai Ho
ngqia
o
Airpo
r
t, Shanghai P
udon
g Airpo
r
t,
Han
g
zhou Xi
aoshan Airpo
r
t and
Nanjin
g Lukou Airp
ort are big o
nes. Th
ese a
i
rpo
r
ts
can h
a
ve
adeq
uate su
p
p
ly of air transpo
rtation.
At the same time, it is very
convenient to conne
ct cities
b
y
surfa
c
e tran
sportation in t
h
is re
gion. S
o
it is as
sum
ed that the actual finished
through
put o
f
air
transpo
rtation
for pa
ssen
g
e
rs of
YRD e
qual
s the d
e
m
and
of this
regio
n
. The
o
peratio
ns of this
regio
n
’s ai
rpo
r
ts are not o
p
timistic. Mo
st air
port
s
(ma
i
nly small an
d medium
-si
z
ed airp
ort
s
) a
r
e
in a deficit state in rece
nt
years. The
studie
s
of de
mand ge
nera
t
ion, demand
distributio
n and
flow of air tra
n
sp
ortation fo
r pa
ssenge
rs in the r
egi
on
can h
e
lp the
airpo
r
ts
kno
w
the market a
nd
analyze their
own a
d
vanta
ges a
nd
sho
r
tcomin
gs. A
ll
above can he
lp the airp
ort
s
confirm m
a
rket
positio
ning a
nd plan a rea
s
on
able ma
rketing strate
gy.
4.2. Case
Ca
lculation
Con
s
id
erin
g
the availability, 9 comm
on stati
s
tical indexe
s
are cho
s
e
n
: X
1
-
GDP(hundred million yuan), X
2
-G
DP p
e
r ca
pita(y
ua
n),
X
3
-di
s
p
o
sable
perso
nal
income
of u
r
ban
resi
dent
s(yu
a
n
), X
4
-re
s
ide
n
t pop
ulation
(
ten thou
sa
nd
), X
5
-nu
m
be
r of
empl
oyee
s(ten
th
ou
san
d
),
X
6
-the propo
rtion of tert
iary
indu
stry
(%
),
X
7
-the co
mpl
e
ted amo
unt
of
fixed investments(hun
dred
million yuan), X
8
-total exports(hundre
d million dollar), X
9
-num
ber
of foreign tou
r
ist
s
(t
en
thousand
). We can u
s
e
th
e data
of YRD(in
clu
d
ing
Shang
hai, Zh
e
jiang Province and
Ji
ang
su
Province) to f
ilter fa
ctors.
The time
is from 1
998
to 2
009, a
nd th
e
data i
s
g
o
t from the
stati
s
tical
yearbo
oks or statistical
b
u
ll
etin
s. Be
cau
s
e the
high
-sp
eed
rail
way i
s
ope
ned
in th
e YRD in
20
1
0
,
and the sp
ee
d and co
mfort of ordinary railway is infe
ri
or to air tran
spo
r
tation, so
we can
not take
railway tra
n
sportation
into
con
s
id
eratio
n
for th
e d
e
ma
nd of
air tran
spo
r
tation. T
a
ble 1
sho
w
s the
factors and
weights.
Table 1. Fa
ctors a
nd weig
ht
X
1
X
2
X
3
X
4
X
5
X
6
X
7
X
8
X
9
O
r
iginal
value
0.1489
0.1304
0.0804
0.0008
0.0019
0.0064
0.1688
0.3110
0.1515
Final
value
0.1502
0.1315
0.0811
-
-
- 0.1703
0.3139
0.1528
After ma
king
the calculatio
n of 9
o
r
iginal
fa
cto
r
s, th
re
e facto
r
s a
r
e
deleted. T
he
weig
hts
of rest facto
r
s are all more than 0.01
whi
c
h mea
n
s these facto
r
s are effective. Therefo
r
e,
key
factors that have impact o
n
deman
d of air tran
spo
r
ta
tion for passe
ngers in the core area of YRD
regio
n
a
r
e: G
D
P, pe
r
capit
a
G
D
P, urba
n pe
r
capita
disp
osable
i
n
come, co
mpl
e
ted
am
ount of
fixed investm
ents, total e
x
port, numb
e
r
of forei
gn t
ouri
s
ts. Th
e
six facto
r
s
a
r
e respe
c
tively
expre
s
sed a
s
x
1
, x
2
, x
3
, x
4
, x
5
, x
6
.
In market econ
omy, the developme
n
t
of
social eco
nomy sh
ould re
ply on the
harm
onio
u
s
developm
ent
of investment, consumpt
i
on and expo
rtation. (1) T
he are
a
of YRD
regio
n
only covers 2% of total territory o
f
Chi
na, but GDP of this region fro
m
20
01 to 2009
ke
eps
about
20% of
Chi
na, a
nd
region
al total
export a
s
a
share
of
Chin
a
'
s total
expo
rt ha
s in
crea
se
d
from 3
0
% to
40%.
Devel
oped
e
c
on
o
m
y and
con
s
is
tent in
crea
se of exp
o
rt i
m
prove
s
regi
nal
eco
nomi
c
vitality and foreign e
c
on
om
ic rel
a
tion
s, whi
c
h will ef
fectively pro
m
ote dema
n
d
gene
ration
of air tra
n
spo
r
tation for
bu
si
ness p
a
sse
n
gers in thi
s
region.
(2)
G
D
P of YRD region
and u
r
ba
n p
e
r capita di
sposable in
co
me are more
than the nat
ional ave
r
ag
e. Numb
er of
air
leisu
r
e travel
ers h
a
s in
cre
a
se
d with the improv
em
e
n
t of reside
nts' paying a
b
ility.
Meanwhi
le,
with rich tourism re
so
urce
s in this re
gi
on, fo
reign t
ouri
s
ts nu
mb
er as a
sha
r
e of China h
a
s
increa
sed
fro
m
6% in
20
0
1
to 1
3
.88%
in 20
09.
Air trans
p
ortation is
th
e first i
m
porta
nt way of
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NIKA
Vol. 11, No
. 5, May 2013 : 2648 – 265
6
2651
traveling for foreign tourists. So
rapi
d devel
opment of touri
s
m will
strongly support
the
developm
ent of air tran
spo
r
tation for thi
s
regi
on.
(3
) The se
co
nd i
m
porta
nt factor is compl
e
ted
amount of fixed inve
stmen
t
s. It is the g
eneral
term
of the wo
rkl
o
ad an
d rel
a
tive expen
se
s of
fixed investm
ents that wa
s built an
d p
u
rcha
sed
re
cently in a ce
rtain pe
riod i
n
the form o
f
curre
n
cy. Increa
se of inve
stment
s will
bring
th
e gro
w
th of econo
my, employment and in
co
me.
All of these will promote d
e
m
and in
crea
se
of air tran
sp
ortation for p
a
ssen
gers.
Data of 16 cities in YRD
core re
gion i
s
used with f
o
rmul
a (1
) to formula (4
).
The
deman
d dist
ri
bution of air transportatio
n
for pa
sseng
er
s. Part of the results a
r
e listed in table 2.
Table 2. De
m
and ge
neratio
n
of each
city(million p
eopl
e)
i
city 2001
2003
2005
2007
2009
1 Shanghai
1014.53
1160.55
1966.83
2447.68
2764.48
2 Nanjing
188.63
211.49
366.06
473.74
551.63
3 Wuxi
172.45
221.32
377.71
496.78
564.82
4 Changzhou
88.39
115.43
203.30
274.11
340.64
5 Suzhou
294.36
438.91
801.92
1064.74
1238.84
6 Nantong
83.23
97.67
177.58
251.07
328.92
7 Y
angzhou
60.46
71.19
125.94
186.06
248.36
8 Zhenjiang
71.87
84.59
144.35
201.93
266.04
9 Taizhou
47.60
54.99
100.62
152.46
204.33
10 Hangzhou
253.58
275.96
451.36
597.70
699.36
11 Ningbo
173.69
219.11
379.05
506.27
593.18
12 Jiaxing
99.61
125.91
218.97
278.19
327.01
13 Huzhou
59.00
73.16
124.71
161.23
192.51
14
Shaoxing
98.74
129.31
213.40
284.49
337.49
15 Zhoushan
46.55
56.97
99.88
147.26
183.98
16 Taizhou2
78.40
99.44
160.56
212.75
234.52
4.3. Case
An
aly
s
is
It takes citie
s
as no
de
s of
deman
d di
stri
bution i
n
the
regio
n
. And
a
i
rpo
r
ts
are
no
des
for dema
nd flow.
4.3.1. Deman
d
Gener
a
tio
n
(1)
Differe
nces of key factors. De
gre
e
of influen
ce on dem
an
d gene
ratio
n
of key
factors is different from o
ne anothe
r. So it is
nece
s
sary to make quantitative analysis o
n
the
differen
c
e
s
in
key factors i
n
each city. Here a
ne
w index is defin
ed, name
s
a
s
contrib
u
tion
rate.
the sha
r
e
of weight of f
a
ctor j i
n
co
mpre
hen
sive
evaluation
value of city i is define
d
as
contri
bution
rate of factor j to city i.
'
=
ij
j
i
ij
Cw
x
V
(5)
In Equation
(5),
C
ij
i
s
the
contri
bution
rate of facto
r
j
to c
i
ty
i
, other
symbol
s h
a
ve the
same
me
anin
g
a
s
above. The
ta
ble 3 shows
th
e con
t
ri
bution
rate
of differe
nt fa
ctors to
16
cities
in 2009.
Table
3 sho
w
s th
at the
same
facto
r
has
differe
nt impa
ct to d
i
fferent citie
s
. For
Han
g
zhou
(
i
=10), the orde
r of contri
buti
on rate of si
x
factors i
s
: numbe
r of foreign tou
r
i
s
ts >
compl
e
ted a
m
ount of fixed investme
nts > total
exp
o
rt > G
D
P
> per
capita
G
D
P > u
r
b
an
per
capita di
sp
osable in
come.
For Na
njing
(
i
=2), the o
r
d
e
r is: amo
unt
of fixed asset investment
>
GDP >
numb
e
r of forei
gn t
ouri
s
ts
> pe
r
capita
GDP
> total export > urban p
e
r
capita dispo
s
a
b
le
incom
e
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Dem
and Di
scipline of Air Tran
spo
r
tation
for Passen
ge
rs (Y
unqia
n
Qu)
2652
Table 3. Co
ntribution
rate o
f
key factors
i x
1
x
2
x
3
x
4
x
5
x
6
1 12.40%
3.12%
1.99%
10.35%
53.24%
18.90%
2 17.47%
15.25%
8.81%
26.23%
15.15%
17.09%
3 20.13%
17.91%
8.44%
22.93%
20.85%
9.73%
4
16.84%
20.81%
13.29%
27.15%
14.45%
7.46%
5 14.23%
12.34%
4.05%
12.99%
41.69%
14.70%
6
19.90%
15.25%
12.17%
29.72%
15.39%
7.58%
7
17.03%
20.79%
14.90%
23.23%
7.31%
16.74%
8
14.32%
25.65%
15.00%
20.60%
6.02%
18.40%
9
18.41%
21.66%
16.86%
30.96%
9.34%
2.77%
10 16.61%
13.33%
7.32%
17.77%
17.59%
27.38%
11 16.19%
15.98%
8.79%
18.33%
29.50%
11.22%
12 13.36%
20.56%
14.39%
20.46%
17.08%
14.15%
13 13.15%
25.39%
23.04%
17.97%
9.59%
10.86%
14 16.03%
20.07%
15.17%
16.96%
21.14%
10.63%
15
6.60%
37.35%
24.94%
11.81%
9.20%
10.09%
16 19.68%
18.69%
19.85%
19.29%
19.43%
3.06%
(2)
Dem
and
pre
d
ication.
Wh
en
maki
ng ma
rket p
l
an, airpo
r
ts sh
ould
not
only
con
c
e
n
trate
on its thro
ug
hput, but also sho
u
ld con
s
ide
r
the de
mand of air t
r
an
spo
r
tation
for
passe
nge
rs i
n
airpo
r
t cove
rage. In that ca
se, it
is ne
ce
ssary to predict dem
and
. First, accord
ing
to key factors and histo
r
ica
l
data, fit
the curv
e of de
m
and ge
neration. Then p
r
e
d
icted value
o
f
key factors a
r
e take
n into the curve e
quation,
so
we ca
n get the future de
mand ge
nera
t
ion.
T
a
k
i
ng
C
h
ang
z
h
o
u
(
i
=4
) a
s
example
and
usin
g multipl
e
linea
r
reg
r
e
ssi
on a
nalysi
s
with
data f
r
om
2001 to 20
09,
we ca
n get the followi
ng e
quation:
=
1984.
53
+
27.
51
+
30.
82
-2741.
21
-7214.
71
+
88092.
82
-1109843
44
1
4
2
4
3
4
4
4
5
4
6
Dx
x
x
x
x
x
(6)
Symbolic me
aning
of equ
ation (6
) i
s
the same
as above. R=0
.
992, R
2
=0.984. It
mean
s
the model
i
s
wel
l
con
s
i
s
tent with
samp
l
e
data he
re.
Whe
n
si
gnifi
cant level
is0.025,
F=70.
9. It rep
r
esents that i
ndep
ende
nt v
a
riabl
es
ha
s remarka
b
le
lin
ear rel
a
tion with
dep
end
ent
variable. The
r
efore, we co
u
l
d
a
c
curately predi
ct
d
e
ma
nd of
Cha
n
g
z
hou i
n
the fut
u
re
by eq
uati
o
n
(6).
4.3.2. Deman
d
of Distribu
tion
With the d
e
velopme
n
t of eco
nomy, de
mand of ai
r t
r
an
spo
r
tation
for pa
sseng
ers i
n
each
city ha
s b
een
in
cre
a
sin
g
yea
r
b
y
year. Bu
t region
al
d
e
m
and dist
ributi
on
i
s
cent
rali
zed.
From
200
1 to
2009, Sh
ang
hai ha
s th
e m
o
st de
man
d
all the time,
whi
c
h d
e
ma
n
d
sh
are is ab
out
one-thi
r
d
of t
o
tal dem
and
in YRD
co
re
regio
n
. And
Zhou
sha
n
h
a
s the
le
st de
mand,
whi
c
h
only
take
s up 2%.
Table 4. De
m
and share
Shanghai
Nanjing
Hangzhou
Changzhou
Zhoushan
2001
35.84%
6.66%
8.96%
3.12%
1.64%
2002
34.62%
6.21%
8.83%
3.19%
1.63%
2003
33.78%
6.16%
8.03%
3.36%
1.66%
2004
33.81%
6.03%
7.80%
3.29%
1.62%
2005
33.27%
6.19%
7.63%
3.44%
1.69%
2006
32.16%
6.14%
7.81%
3.42%
1.81%
2007
31.64%
6.12%
7.73%
3.54%
1.90%
2008
31.17%
6.09%
7.76%
3.70%
1.98%
2009
30.46%
6.08%
7.71%
3.75%
2.03%
Table 4 sho
w
s
the dema
nd sha
r
e of some
cities. As
the
re
pre
s
entative
s
of
larg
e
cities, dem
an
d sha
r
e of Shang
hai, Nan
jing and
Han
g
zh
ou ha
s b
een de
crea
si
ng year by year.
Me
a
n
w
h
ile
,
de
ma
n
d
sh
ar
e o
f
sma
ll an
d mid
d
l
e-s
i
ze
d c
i
ties
s
u
c
h
as
C
h
an
gz
h
o
u
a
n
d
Z
h
ou
sha
n
kee
p
s incre
a
s
ing
ra
pidly.
Contin
uou
s i
n
crea
se
of
d
e
mand
provides a g
ood
market b
a
si
s for
airpo
r
ts in
sm
all and middl
e-si
ze
d citie
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NIKA
Vol. 11, No
. 5, May 2013 : 2648 – 265
6
2653
4.3.3. Deman
d
Flo
w
In Table
5,
Pudong
airp
ort an
d Hon
gqiao
airp
ort
are
merged.
It is assu
m
ed that
airport
s
e
rv
es
the
c
i
ty, where it
is
loc
a
te
d f
i
rstly. Dema
nd fini
shin
g
rate
=airp
o
rt
throug
hput/d
emand
of th
e city where
airp
ort i
s
lo
cated.
Wh
en
the de
mand
finishin
g rate is
bigge
r th
an
1, the
airp
ort attract
s
de
mand
of
oth
e
r
citie
s
ex
cept the
lo
cal
dem
and. S
o
passe
nge
rs i
n
flow ha
ppe
n
s
. On the
co
ntrary, if
dem
and finishing
rate is l
e
ss t
han 1, it sh
o
w
s
that the airpo
r
t coul
dn't co
mplete local
deman
d.
And
this is outflo
w
. For exam
ple, the dem
and
finishin
g rate
of Na
njing
a
i
rpo
r
t in 2
0
0
1
is
149.2
4
%, whi
c
h
rep
r
e
s
ent
s
Nanjin
g airport
wa
s a
node
with p
a
s
seng
er inflo
w
in 2
001. T
h
e dema
nd fini
shin
g rate
of
Ning
bo ai
rpo
r
t in 2009 i
s
o
n
ly
67.96%, whi
c
h
sho
w
s th
at Ning
bo ai
rport was
pa
ssen
ger
outflo
w
no
de an
d
only com
p
let
e
d
67.96% of local deman
d.
Table 5. De
m
and flow
passenger thro
u
gh put (million)
demand finishing rate
2001
2005
2009
2001
2005
2009
Shanghai 2066.04
4146.23
5699.96
203.65%
210.81%
206.19%
Nanjing 281.52
538.59
1083.72
149.24%
147.13%
196.46%
Wuxi 3.63
61.91
221.79
2.10%
16.39%
39.27%
Changzhou
12.65
31.62
53.52
14.31%
15.55%
15.71%
Nantong
7.35
8.93
22.03
8.83%
5.03%
6.70%
Hangzhou
298.13
809.26
1494.47
117.57%
179.29%
213.69%
Ningbo
121.90
253.29
403.14
70.18%
66.82%
67.96%
Zhoushan
25.07
39.79
44.80
53.85%
39.84%
24.35%
Taizhou2 14.81
22.59
52.67
18.89%
14.07%
22.46%
Acco
rdi
ng to passe
nge
r throug
hput in 2
009,
the ten regio
nal airp
orts a
r
e cla
s
sified
into four types
by quick
c
l
us
ter:
1)
sup
e
r la
rge ai
rpo
r
t, includi
n
g
Pudong a
n
d
Hon
gqiao
a
i
rpo
r
ts;
2)
large ai
rp
ort, inclu
d
ing
Nanj
ing and
Han
g
z
ho
u airp
ort
s
;
3)
middle
-
si
zed
airpo
r
ts, in
clu
d
ing Wuxi an
d Ning
bo airp
orts;
4)
small ai
rpo
r
ts, the rest four airpo
r
ts.
The
real
p
a
ssenge
r th
rou
g
hout of
ea
ch
type of
airp
ort
s
kee
p
s
in
cre
a
sin
g
from 20
01
to
2009, b
u
t the
ch
angin
g
tre
nd of d
e
ma
n
d
finishi
ng
ra
te is diffe
rent
. Dema
nd fin
i
shin
g rates
of
ultra large a
n
d
larg
e airports are
more than 1.
So th
ese
airp
orts
are p
a
sse
n
g
e
r inflo
w
no
d
e
s.
The
ca
pacity
of two
airport
s
in
Sha
ngha
i is
gradua
lly
clo
s
e to
saturation, which i
nevitably lea
d
s
to red
u
ctio
n
of se
rvice
qu
ality
and in
crease of flight
delay. The
r
ef
ore, p
a
sse
n
g
e
rs be
gin to f
l
ow
to othe
r la
rge
airp
ort
s
re
ce
ntly. The ma
rket
con
c
e
n
tra
t
ion of two ai
rports in
Shan
ghai
starte
d t
o
redu
ce afte
r reaching its p
eak. Nanjing
and Ha
ng
zho
u
airpo
r
ts obt
ain a benefit firstly. Dema
nd
finishin
g rate
s of
sm
all a
nd mi
ddle
-
si
zed ai
rpo
r
ts a
r
e l
e
ss th
an
1, that me
an
s in
YRD
co
re
regio
n
, le
ss throu
gho
ut of
middle
an
d
small
ai
rport
s
i
s
n
o
t la
ck of dem
and,
but b
e
cau
s
e
of
deman
d outflow.
Dema
nd flo
w
of air tran
sp
ortation fo
r p
a
ssen
ger i
s
actually the
airpo
r
t sele
ction of
passe
nge
rs.
In multi-airp
o
r
t system, there a
r
e more than one
airpo
r
ts a
r
ou
nd a city. So a
passe
nge
r who travel
s by
air
woul
d se
lect on
e ai
rp
ort to leave
or
sele
ct on
e
airp
ort to re
ach
whe
n
he retu
rns. Dem
and flow
ap
pea
rs whe
n
the ch
o
s
en
airp
ort i
s
not located in
the city wh
ere
the pa
sseng
er is. T
here
are m
any fa
ctors in
flue
n
c
ing
dema
n
d
flow, su
nch
as a
r
ea
of the
terminal, n
u
m
ber
of ch
eck-in
co
unte
r
s,
numbe
r of
b
oardi
ng g
a
te
s, su
rface tra
n
sp
ortation ti
me,
flight frequ
en
cy, navigatio
n co
ndition,
etc. Shan
gha
i, Nanjin
g,
Ni
ngbo and Ch
ang
zho
u
airp
orts
stand
for
the four cate
gori
e
s se
par
ately. There a
r
e
two
cu
rves in
Figures 1, 2,
3, 4, which a
r
e
prop
ortio
n
s o
f
throug
hput and
flight mo
vement
chan
ging
with tim
e
. Flight mov
e
ment
refle
c
ts the
numbe
r of ai
r ro
utes
and
flights of an
airpo
r
t.
It is the su
pply of air tra
n
sp
ortation from
the
airpo
r
t. Passenge
r thro
ug
hput sh
ows the outcome o
f
demand flo
w
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Dem
and Di
scipline of Air Tran
spo
r
tation
for Passen
ge
rs (Y
unqia
n
Qu)
2654
F
i
g
u
r
e
1
.
S
ha
ng
ha
i
a
i
r
p
or
t
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
12
3456
7
8
9
F
i
g
u
r
e
2
.
N
an
gj
i
n
g
a
i
r
p
or
t
0
0.02
0.04
0.06
0.08
0.1
0.12
0.14
1
2
3
456
78
9
F
i
g
u
re
3
.
N
i
n
gb
o
a
i
rp
ort
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
1
2
3
456
78
9
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 23
02-4
046
TELKOM
NIKA
Vol. 11, No
. 5, May 2013 : 2648 – 265
6
2655
F
i
g
u
r
e
4
.
C
ha
ng
z
h
ou
a
i
r
po
r
t
0
0.
001
0.
002
0.
003
0.
004
0.
005
0.
006
0.
007
0.
008
0.
009
12
345
678
9
Whateve
r
whi
c
h type of airports, chan
gi
ng tr
end of two curve
s
is
basi
c
ally co
n
s
iste
nt
whi
c
h sho
w
s there is la
rge co
rrelatio
n betwe
en p
a
ssen
ger th
rough
put and
flight movement
whe
n
d
e
man
d
is
sufficie
n
t. It mean
s the
numbe
r
of
ai
r ro
utes an
d fli
ghts
have
a
major imp
a
ct
on
regio
nal ai
r transportatio
n
deman
d flow.
In the
p
r
emi
s
e
of suffici
e
n
t
dema
nd, middle
a
nd small
airpo
r
ts
sh
oul
d plan m
a
rke
t
accurately
and
carry
out
marketing a
c
tively to increase air
rout
es
and imp
r
ove flight freque
ncy in order that
airpo
r
t woul
d
redu
ce de
ma
nd outflow.
5. Conclusio
n
Dema
nd di
sci
p
line of ai
r transportatio
n
for pa
sseng
e
r
s
con
s
i
s
ts of
three p
a
rt
s i
n
this
pape
r: de
ma
nd g
ene
ratio
n
, dema
nd
di
stributio
n a
n
d
dema
nd flo
w
. It takes the
YRD
co
re
re
g
i
on
as
example.
In dem
and
gene
ratio
n
analysi
s
, ba
sed on
information e
n
tro
p
y and
a
weight
threshold,
it
coul
d g
e
t wei
ght when
sel
e
cting
the
six
key
facto
r
s i
n
fluen
cing
re
gional
dem
a
n
d
gene
ration
of
air tra
n
sportation for pa
sseng
ers.
Th
e sele
ction
is b
enefi
c
ial
to sim
p
lify th
e
compl
e
xity of cal
c
ulatio
n a
nd can b
a
si
cally reflec
t th
e overall pe
rspective of the
pro
b
lem. Th
en,
a new ind
e
x - cont
ribution
rate is given
to analyze the differen
c
e
s
of the sam
e
factor to e
a
ch
city. Secondl
y, in demand distrib
u
tion a
nalysi
s
, dema
nd distri
butio
n in YRD core regio
n
pre
s
ents
centralization
.
Shanghai, the bigge
st city in
the area, take
s u
p
about one
-third d
e
man
d
.
Ho
wever, the
trend of ce
ntralization is
sl
owin
g do
wn
year by year.
Deman
d
gen
eration of
sm
all
and mi
ddle
-
si
zed
citie
s
in t
he a
r
ea i
s
in
cre
a
si
ng m
o
re and
more.
Thro
ugh
re
co
gnisi
ng
curre
n
t
situation an
d
future tende
ncy of dema
nd distri
but
io
n in its servi
c
e ra
nge, airport ca
n make a
better pla
n
. Finally, in dem
and flow, the
most impo
rtant factor th
a
t
influence
s
region
al dema
n
d
flow is the n
u
mbe
r
of air route
s
and fli
ghts of
airp
ort. The more air ro
utes an
d the highe
r the
flight freque
n
c
y it has, the
more d
e
stin
ation t
he pa
ssen
gers
coul
d arrive
and
the more flig
hts
they coul
d select, the more
attrac
tive to passengers the airport
w
ill has. Therefore, large
airports
are pa
sseng
e
r
inflo
w
n
ode
s while mo
st
middle and small
airpo
r
ts are pa
sseng
e
r
outflo
w
n
o
d
e
s.
In YRD co
re
region, mid
d
le and sm
al
l airport
s
are
located in small and mid
d
le-sized citi
es.
These ai
rpo
r
t
s
have
defi
c
i
t
s be
cau
s
e
o
f
less thro
ug
hout. Actuall
y
, these ai
rp
orts
do n
o
t l
a
ck
deman
d. Through givin
g
accurate market or
ie
nta
t
ion and de
mand an
alysis, ca
rrying
out
marketing a
c
t
i
vely, seekin
g
more air
rout
es an
d
flights, middle and
small airport
s
wo
uld re
du
ce
passe
nge
r ou
tflow and in
crease its throu
ghout.
Referen
ces
[1]
W
i
ndle
R, Dresner M. Air
port Cho
i
ce i
n
Multip
le-a
irp
o
rt Regi
ons.
Journ
a
l of T
r
ansp
o
rtatio
n
Engi
neer
in
g
. 1995; 12
1(4): 33
2-33
7.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
2302-4
046
Dem
and Di
scipline of Air Tran
spo
r
tation
for Passen
ge
rs (Y
unqia
n
Qu)
2656
[2]
VA Profill
idis.
Econom
etric a
nd F
u
zz
y M
o
d
e
ls for th
e
F
o
r
e
cast of D
e
ma
nd i
n
the
Airp
o
r
t of Rho
des
.
Journ
a
l of Air T
r
ansp
o
rt Mana
ge
me
nt
. 2000;
6(2): 95-1
00.
[3]
Z
hang W
enc
h
ang. Patterns
and D
y
n
a
mics
of Regi
on
al Pa
sseng
er F
l
o
w
s
in Chi
na si
nc
e the 19
80s.
Scienti
a
Geogr
aph
ica Sin
i
ca
.
200
4; 24(4): 38
5-39
0.
[4]
Jiao Pe
ng
pe
n
g
. F
o
recastin
g Metho
d
an
d Its Mechan
ism of Impacts on Airport
Passen
g
e
r
T
h
roughp
ut.
Journa
l of T
r
ans
p
o
rtation
Syste
m
s E
ngi
ne
erin
g an
d Infor
m
ati
on T
e
c
hno
lo
gy
. 200
5; 5(1):
107-
110.
[5]
W
e
i Xi
n, F
eng
Xi
ngj
ie. Pred
i
c
ting the Air
p
o
r
t
Passeng
er T
h
roughp
ut Based on
a Sup
port Vecto
r
Regr
essio
n
Mo
del.
Jour
nal
of Civil Avi
a
tio
n
U
n
iversity of Ch
i
n
a
. 200
1; 22(3)
: 45-47.
[6]
Z
hou Y
i
xin
g
, Z
han
g
Li, W
u
Y
ue. Stud
y
of
C
h
in
a’s Urba
n C
entral
it
y
H
i
era
r
ch
y
.
Ar
ea
l R
e
search
an
d
Devel
o
p
m
ent
. 200
1; 20(4): 1-
5.
[7]
W
ang W
e
nfan
g. Res
earch
o
n
Meas
ur
em
en
t of Custom
er
Equit
y
of Airl
in
es. MaD T
hesi
s
. Nanj
in
g:
Coll
eg
e of Civil
Aviation, Na
nji
ng Un
iver
sit
y
o
f
Aeronautics
a
nd Astrona
utic
s; 2010.
[8]
Guo Yaj
un, Yi
Pingta
o
. A Mul
t
i-eval
uatio
n C
oncl
u
sio
n
Ag
gr
egati
on Met
h
o
d
Base
d o
n
Si
ngu
lar Va
lu
e
Decom
positi
o
n
.
Journal of No
rtheastern U
n
iv
ersity (Natura
l
Scienc
e)
. 200
7
;
28(2): 278-2
8
1
.
[9]
Ke Z
hao. T
he Rese
arch of Gra
y
A
l
gor
ithm
and Info
rm
atio
n Entrop
y i
n
R
oute Pl
ann
in
g Optimizatio
n
.
Indon
esi
a
Jour
nal of Electric
al
Engin
eeri
n
g
. 2
012; 10(
8):
[10]
Liu
Xi
aomi
ng.
Intellig
enc
e Comp
utation
a
l
Based Ai
r P
a
ssen
ger Flo
w
D
i
strib
u
tion
in Uncerta
i
n
Enviro
nment.
PhD T
hesis. N
anji
ng: C
o
ll
eg
e
of Civil
Av
iati
on, Na
nji
ng U
n
iversit
y
of Ae
rona
utics an
d
Astronautics; 2
008.
Evaluation Warning : The document was created with Spire.PDF for Python.