TELKOM
NIKA
, Vol.14, No
.2, June 20
16
, pp. 404~4
1
0
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v14i1.2958
404
Re
cei
v
ed
No
vem
ber 3, 20
15; Re
vised
F
ebruary 16,
2016; Accept
ed Marc 2, 20
16
Photovoltaic Array Maximum Power Point Tracking
Based on Improved Method
Jia Cunliang
*
1
, Wang Yan
x
iong
2
, Wan
g
Zerong
3
1,2
School of Informatio
n
an
d Electrical E
ngi
ne
er
in
g, Univers
i
ty of Mini
ng a
n
d
T
e
chnol
og
y,
Chin
a Xuzh
ou
, 0516-8
3
8
907
17
3
No.716 R
e
se
a
r
ch Institute Lia
n
y
un
ga
ng, Chi
na, 180
36
67
85
56
*Corres
p
o
ndi
n
g
author, e-ma
i
l
: jcumt@16
3.com
A
b
st
r
a
ct
At present, a goo
d dea
l of meth
ods for the Maxi
mu
m P
o
w
e
r Point T
r
acking (MPPT
) has be
en
used i
n
en
gin
e
e
rin
g
app
lic
atio
ns. How
e
ver, Matlab si
mula
ti
on prov
ed that
they w
e
re difficult to harmon
i
z
e
the stability and speed
ability of system
. In addition,
in or
der to m
a
ximi
z
e
t
he
use
of PV panels
‘
p
ower, the
pap
er focuse
d on an
eo
natal
al
gorith
m
for Ma
ximu
m Pow
e
r Point T
r
ackin
g
(MPPT
). Based on the al
gor
ithm,
this pa
per
desi
gne
d a
n
i
m
pro
v
ed a
nd fe
asi
b
le var
i
ab
le
ste
p
pertur
bati
on and obs
ervatio
n
metho
d
w
h
ic
h
w
e
ll all
e
viat
ed
the conflict tha
t
the maxi
mu
m pow
er poi
nt
trackin
g
cou
l
d n
o
t take into ac
count the sta
b
i
lity
and sp
ee
d of respo
n
se effici
e
n
tly.
Ke
y
w
ords
:
PV
pow
er gen
erat
ion, MPPT
, Variabl
e st
ep p
e
rturbati
on a
nd o
b
servati
on
met
hod
Copy
right
©
2016 Un
ive
r
sita
s Ah
mad
Dah
l
an
. All rig
h
t
s r
ese
rved
.
1. Introduc
tion
In photovoltai
c
p
o
wer
gen
e
r
ation
syste
m
s, a
c
tual o
u
tp
ut po
wer of of
ten de
pend
s
on light
intensity, bat
tery tempe
r
a
t
ure a
nd l
o
ad imp
eda
n
c
e [1]. In t
he three fa
ctors
of out
put
cha
r
a
c
teri
stic of the PV ce
lls, light a
nd t
e
mpe
r
at
ure i
s
external e
n
v
ironme
n
t wh
ich i
s
u
nable
to
control. Therefore, adju
s
ti
ng the imped
ance value
o
f
load is the only choi
ce to make PV cells
output maximum po
wer i
n
the pra
c
tical work, mea
n
whil
e, this is the essen
c
e of photovoltaic
array maxim
u
m po
we
r p
o
i
nt trackin
g
.
At pre
s
ent, t
he MPPT m
e
thods u
s
ed
b
y
resea
r
chers at
home a
nd a
b
roa
d
in
clud
e: con
s
tant
-voltage me
th
o
d
, pertu
rbati
on an
d ob
se
rvation meth
od,
synovial co
ntrol,
in
crement
al
cond
ucta
n
c
e method,
fuzzy
cont
rol
and neu
ral n
e
twork predi
ction
method [2-4]. Perturb
a
tion
and ob
serva
t
ion is the m
o
st co
mmonl
y used meth
od in pra
c
tice
for
its sim
p
licity
and
ea
se of i
m
pleme
n
tatio
n
[5]. Ho
wev
e
r, it suffers f
r
om th
e
slo
w
tracking
spee
d at
small d
u
ty cycle
step a
nd fluctuate
s
whe
n
su
bje
c
t
ed
with large d
u
ty step, whi
c
h
results in
hig
h
e
r
losse
s
unde
r dynamic we
ather to whi
c
h the
photo
v
oltaic PVcel
ls exposed [6]. This pap
er
focu
sed o
n
a
n
improve
d
a
nd feasi
b
le variabl
e step p
e
rturbation a
nd ob
servatio
n method he
re.
PV cells a
r
e
recogni
zed f
o
r having
no
n-line
a
r
cha
r
acteri
stics. At one point, whi
c
h is
kno
w
n
as th
e maximum
power p
o
int
(MPP), the
cells a
r
e
cap
a
b
le to op
erat
e at maximu
m
efficien
cy an
d give the m
a
ximum outp
u
t [7-10]. According to th
e photovoltai
c
cells
equiv
a
lent
circuit mod
e
l [11], two typical output characteri
st
ic
cu
rves of PV cell
s can b
e
determined u
nde
r
a
certai
n tem
peratu
r
e
an
d illuminatio
n paramet
e
r
s: output
current-o
utput
voltage (I
-V)
cha
r
a
c
teri
stic curve a
nd o
u
tput
power-output voltag
e (P-V)
cha
r
acteri
stic
cu
rve are sho
w
n as
Figure 1.
(a) I - V cha
r
a
c
teri
stic
curv
e
(b) P - V chara
c
teri
stic
c
u
rv
e
Figure 1. Typical outp
u
t ch
ara
c
teri
stic
curve of PV cells
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Photovoltai
c
Arra
y Ma
xim
u
m
Power Po
int Trackin
g
Base
d on Im
prove
d
Method
(Jia
Cunli
ang
)
405
In the Figu
re
1, Isc a
nd Vo
c is
sh
ort ci
rc
uit curre
n
t an
d ope
n ci
rcuit voltage of PV cells
respectively. MPP (Maximum Power P
o
int) is
the
worki
ng p
o
int of maximum output po
we
r of
photovoltai
c
panel
s, then
the Vm is voltage an
d
Im
is
curre
n
t of MPP. Conn
ecting PV
ce
lls
terminal
s to l
oad resi
stan
ce, its externa
l
cha
r
a
c
te
ri
stic cu
rve is th
e origi
n
of the dotted line
in
Figure 1(a
)
. As the load resi
stan
ce is i
n
crea
sing fro
m
0 gradu
ally, the output voltage and o
u
t
put
power of the
PV cell
s a
r
e
also
in
cre
a
se
d sl
owly f
r
om
0;
Whe
n
the
load
resi
stan
ce i
n
cre
a
se
s
to
Rm in
Fig
u
re
1(a
)
, the
ou
tput po
wer of
photovoltai
c
pan
els incre
a
se
s to
pea
k,
namely i
n
t
he
MPP, and Pm = ImVm;
Contin
uing to
increa
se the
load
re
sista
n
ce, the
outp
u
t voltage of
PV
cell
s co
ntinue
s to increa
se,
but its output powe
r
begi
n
s
to decrea
s
e
.
Figure 1
(
b
)
i
s
the
P - V
ch
ara
c
te
risti
c
curve
ab
o
u
t the
chan
g
e
of the
rel
a
tionshi
p
betwe
en load
chan
ge
s of PV cells and
output volt
ag
e, and the MPP point corresp
ond
s to MPP
point in
the
Figure 1
(
a
)
.
Whe
n
the
lo
ad resi
stan
ce de
crea
se
s from i
n
finity to Rm, P - V
cha
r
a
c
teri
stic curve is
right
side pa
rt of the M
PP point in Figure 1
(
b
)
acco
rdin
gly; when the loa
d
resi
stan
ce
over
Rm contin
ues to
scale
down, co
rrespondi
ng P -
V cha
r
a
c
teri
st
ic curve i
s
th
e left
side p
a
rt of the MPP point in Figure 1(b
)
.
2. The Propo
sed Me
thod
In photovoltai
c
po
we
r g
e
n
e
ration
sy
ste
m
s,
maximu
m po
wer
poi
nt trackin
g
(MPPT) of
PV cell
s i
s
a
c
hieved
by DC/DC
co
nverte
r u
s
u
a
lly. To
be in
the i
m
pl
ementation
p
r
incipl
e, the to
tal
circuit
com
p
o
s
ed
of
DC/
DC
conve
r
ter a
nd lo
ad i
s
re
g
a
rde
d
a
s
the
system
eq
uivalent lo
ad
wh
ose
size is a ce
rtain function
of
duty-cycle
D of the workin
g DC/
DC converte
r. Thus, the out
put
power of PV
cell
s can b
e
ch
ang
ed b
y
adjusti
n
g
the duty-cycl
e D
simply,
and the
n
M
PPT
function i
s
implemente
d
. Boost converte
r
circuit dia
g
ram is shown in Figure 2.
PV
i
PV
e
e
i
L
i
Figure 2. Dia
g
ram of bo
ost
converte
r ci
rcuit in PV system
In Figure 2,
assumin
g
tha
t
Q is an ide
a
l
power swit
ch can obtai
n two kin
d
s
of state
equatio
n of Q in each
cycle
accordi
ng to the dynamic
circuit analy
s
i
s
method.
Q is co
ndu
ct
e
d
:
PV
1P
V
PV
d
d
d
d
L
L
e
Ci
i
t
i
Le
t
(1)
Q
is
c
l
os
ed
:
PV
1P
V
PV
d
d
d
d
L
L
e
Ci
i
t
i
Le
e
t
(2)
If the state
o
f
swit
ch i
s
p
r
oce
s
sed
by
averag
e met
hod in
ea
ch
swit
ch
cycl
e
and th
e
averag
e amo
unt of state is
represented
by overline, then
the re
sult
is as follo
ws:
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 2, June 20
16 : 404 – 41
0
406
PV
1P
V
PV
d
d
d
1
d
L
L
e
Ci
i
t
i
L
eD
e
t
(3)
In accord
an
ce with the sta
t
e-sp
ace averagi
ng meth
od
, (4) is the sta
t
e-sp
ace equ
ations:
PV
PV
T
11
PV
11
00
d
d
11
00
LL
ee
e
CC
i
ii
t
LL
(4)
e
T
=(
1-
D)
×e
, D is the duty ratio and
()
0
≤
D
≤
1
i
n
form
ula (4).Be
ca
u
s
e it is kno
w
n
that eT
is a functio
n
of D, therefore,
the system
working p
o
in
t will be c
han
ged with the
variation of d
u
ty-
ratio
D
whil
e t
he Bo
ost
co
n
v
erter i
n
Fi
gu
re
3 i
s
wo
rki
n
g. Wh
en
dP/d
e
PV
=0
, the func
tion of MP
P
T
is
realiz
ed [12].
3. Rese
arch
Metho
d
3.1. MPPT Fixed Step
Alg
o
rithm
In this
pap
er,
the PV cell i
s
con
n
e
c
ted to
a DC
bu
s through
the Bo
o
s
t
conve
r
ter. I
n
orde
r
to reali
z
e
MP
PT, the eq
uivalent lo
ad im
peda
nce
of t
he p
hotovolta
ic p
o
wer
gen
eration
sy
ste
m
can
be
chan
g
ed by
adju
s
ti
ng the
d
u
ty-ratio (D) of
the Bo
os
t co
n
v
erte
r
an
d
mak
i
ng
it th
e sa
me
a
s
the power so
urce internal resi
stan
ce. Al
t
hough a lot of the MPPT
method
s hav
e been u
s
ed
at
home an
d abro
ad, pert
u
rbatio
n and
obse
r
vation
method is commo
nly adopte
d
in the
engin
eeri
ng
pra
c
tice to
write co
ntrol p
r
og
ram
s
an
d
pertu
rbation
step is
usua
lly a fixed value
becau
se of the limitation
of processin
g
speed of curre
n
t micro
p
ro
ce
ssor.
Flow
graph
o
f
fixed step
si
ze
pertu
rb
ation a
nd
observation metho
d
is Figu
re
3.
I (k), V
(k) an
d P (k), resp
ectively
, is output cu
rre
n
t, output
voltage and
output po
we
r of photovoltaic
power
gene
ration sy
stem
in the
kth tim
e
sa
mplin
g.
ᇞ
P is th
e sy
stem po
wer dif
f
eren
ce
between
before and after
sa
mpling,
ᇞ
V is the ou
tput voltage of PV cells di
fference bet
ween befo
r
e
a
n
d
after samplin
g, D(k) i
s
d
u
ty-cycle
of dr
i
v
e sign
al of p
o
we
r ele
c
tron
ic devi
c
e a
n
d
ᇞ
D i
s
the fixed
variational
step of D(k).
St
a
r
t
V
(
k
)
,I(k
)
P(
k)
=
V
(
k)
×
I(
k
)
△
P
=
P
(
k
)
–
P(
k-
1)
△
V =
V
(
k
)
–
V(
k-
1)
△
P =
0
△
P >
0
△
V >
0
△
V >
0
D(
k
+
1
)
=
D(
k
)
-
△
D
D(
k+1)
=
D
(
k)
+
△
D
D(k
+
1)
=
D(
k)
+
△
D
D
(
k
+
1)
=
D(
k)
-
△
D
V
(
k-
1)
=
V
(
k)
I
(
k-
1)
=
I(k
)
P
(
k-
1)
=
P
(
k
)
Re
t
u
rn
Y
Y
YY
N
N
N
Figure 3. Flow gra
ph of fixed step
si
ze pertu
rbatio
n and ob
se
rvation method
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Photovoltai
c
Arra
y Ma
xim
u
m
Power Po
int Trackin
g
Base
d on Im
prove
d
Method
(Jia
Cunli
ang
)
407
The
simulatio
n
re
sult
s
can
be foun
d in
th
e last
se
ction
,
whe
n
ᇞ
D i
s
large
r
, the
re
spo
n
s
e
spe
ed of system unde
r illumination
ch
ange i
s
rapi
d
,
but the output power of
PV cells ha
s a
stron
g
sh
ock near the maximum power point. Wh
en
ᇞ
D begin to decrease, the oscillati
on
amplitude
is
decrea
s
e
d
o
b
viously; ho
wever, the
system tra
c
ks ill
umination
ch
ange
at a
slo
w
er
pace. It is visible that fixed
step
alg
o
rith
m is
difficult t
o
coo
r
dinate
betwe
en
stab
ility and
rapid
i
ty
for MPPT in the s
y
s
t
em.
3.2. MPPT Improv
ed Var
i
able Step Al
gorithm
In ord
e
r to o
v
erco
me th
e
sho
r
tage
of t
he
former pe
rturb
a
tion
an
d ob
se
rvation
method,
this p
ape
r im
proved
alg
o
rit
h
m ba
se
d o
n
the fix
ed ste
p
MPPT sim
u
lation
exp
e
ri
ments. With
t
h
e
cha
nge
of o
u
t
put voltage
(V), the
cha
n
g
e
rule
of outp
u
t po
we
r of P
V
cell
s
(P) ca
n be
summe
d
up in Figu
re
1(b
)
. As sh
o
w
n in Fig
u
re
4, M1, M2, M3 and M
4
are o
b
taine
d
from the P - V
cha
r
a
c
teri
stic curve in Fig
u
r
e 1(b).
Figure 4. Output cha
r
a
c
teri
stic curve tan
gent of PV module
By comp
ari
s
on, the
slo
p
e
rel
a
tionship
among
M1,
M2, M3
and
M4
can
be
o
b
tained
in
Figure 4.
M2
M2
M
1
M
1
M
3
M
3
M2
M2
M
1
M
1
M4
M4
dd
d
d
d
d
dd
d
d
d
d
P
VP
V
P
V
P
VP
V
P
V
(5)
Visibly, slope
of P - V characteri
stic
curv
e is minimum
only in the maximum power point
M2 (
M2
M2
dd
=
0
PV
), ma
kin
g
use of the
relation
ship
of
ᇞ
P/
ᇞ
V on th
e cu
rve a
nd
output po
we
r
P to adjust step value every time in pertur
bation m
e
thod, namel
y, the slope and di
sturb
a
n
ce
step have positive correl
a
tion,
then
the system
search of
st
ability and rapidity can
be
coo
r
din
a
ted. Therefore,
ᇞ
D(k) in the m
o
ment of k is
as follo
ws:
()
()
(
1
)
()
()
(
1
)
()
(
)
()
Pk
Pk
Pk
Vk
Vk
Vk
Dk
P
k
V
k
D
(6)
ᇞ
V(k) an
d
ᇞ
P(k), re
spe
c
ti
vely, is outpu
t voltage vari
ation an
d out
put po
wer va
riation of
PV cell
s, d
u
ri
ng the
mo
me
nt of (k - 1) to k d
ue to
ch
ange
of d
u
ty
ratio.
ᇞ
D i
s
the fixed
varia
t
ion
step of the previous p
e
rtu
r
bation an
d o
b
se
rvation m
e
thod. Co
nse
quently,
ᇞ
D(k) s
u
bs
titutes
for
ᇞ
D in Fig
u
re
3 according t
o
the formula
(6) in
the
cont
rol flow cha
r
t of improved a
l
gorithm.
In the improv
ed algo
rithm,
Matlab simu
lation experi
m
ent found t
hat
ᇞ
P(k)/
ᇞ
V(
k
)
w
e
r
e
so large be
cause of the in
terferen
ce e
ffect that
ᇞ
D(k) we
re too
large a
nd the
system co
ntrol
were di
so
rde
r
ed,
so it ne
eds to
imp
r
o
v
e algo
rithm
again.
Com
b
ining th
e chara
c
te
risti
c
s of
previou
s
fixe
d step
with p
e
rturbation m
e
thod, tw
o
ki
nds
of duty ratio ca
n be
set in the co
ntrol
algorith
m
to
regulate th
e
step len
g
th. Th
e larger
on
e
adju
s
ts
step t
o
cope
with t
he mutatio
n
s
of
illumination i
n
MPPT cont
rol to make the system
search the maxi
mum power point qui
ckly
while
the smalle
r is used for the
control after the MPP
is found to red
u
ce the system
power o
scill
ation
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 2, June 20
16 : 404 – 41
0
408
near th
e ma
ximum power point. The Figure 4 sh
o
w
s that
ᇞ
P(
k
)
/
ᇞ
V(k) h
a
s
a positive an
d a
negative o
n
b
o
th sid
e
s of
MPP, the wo
rking
point of
PV cells is i
n
verse
d
alte
rn
ately betwe
e
n
left
and ri
ght
sid
e
s of MPP u
nder t
he pe
rt
urbatio
n met
hod. Somethi
ng can b
e
ob
tained
whe
n
the
workin
g point
is cha
nge:
(1
)
(
)
0
(1
)
(
)
(1
)
(
)
0
(1
)
(
)
Pk
Pk
S
Vk
Vk
Pk
Pk
S
Vk
Vk
(7)
In formula (7
), S is a positive that shows t
he wo
rking
point of PV cells did n
o
t re
ach the
MPP,
ᇞ
D sh
ould choo
se
a larg
e ste
p
to track the
ex
ternal e
n
vironment va
ria
t
ion rapi
dly. The
workin
g poi
nt of PV cell
s j
u
mps from
o
ne si
de to th
e
other
sid
e
of
MPP whe
n
S inverses f
r
om a
positive to a negative. Th
at is to say, t
he workin
g point of PV cell is nea
r the MPP and
ᇞ
D
should choose a small
step
to reduce
the
po
wer oscillation of the
system
near the M
PP.
Acco
rdi
ng to
the above
ste
p
conve
r
sion
method t
he
seco
nd imp
r
ov
ement of algo
rithm ha
s be
e
n
compl
e
ted.
Judgin
g
wheth
e
r S i
s
a
posit
ive or n
o
t bef
ore j
udgin
g
ᇞ
P=
0 in the flow chart of figure
3 to choo
se t
he be
st
ᇞ
D.
4. Results a
nd Discu
ssi
on
As sh
own in
Figure 5, se
t up the max
i
mu
m po
we
r
point tra
cki
ng
(MPPT)
sim
u
lation
model
of ph
otovoltaic
po
we
r g
ene
ration
system b
a
sed
on Matla
b
/Simulink softwa
r
e
enviro
n
me
nt
and ma
ke u
s
e of the main circuit of Boo
s
t conve
r
ter d
e
sig
ned in
se
ction 2.
Figure 5. Maximum po
wer
point trackin
g
simulation m
odel
4.1. Simulation of Fixed
Step Perturb
a
tion and O
b
serv
a
tion Method
In Figure 5, t
he op
en
circuit voltage p
hotovol
taic
p
anel
s Voc =
22.1V, the sh
ort ci
rcuit
curre
n
t Isc = 5.95 A, volt
age of
maxi
mum p
o
we
r
point Vm
=
18.2 V, curre
n
t Im =
5.55
A,
temperature
photovoltai
c
panel
s in th
e
maximum p
o
w
er point i
s
2
5
℃
. Building
a control
module
according
to t
he flo
w
cha
r
t
of figure
3 to
simulate
an
d
get the
simul
a
tion dia
g
ram
abo
ut tra
cki
n
g
the output po
wer of ph
otovoltaic
pa
nel
s as sho
w
n in
Figure 6.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Photovoltai
c
Arra
y Ma
xim
u
m
Power Po
int Trackin
g
Base
d on Im
prove
d
Method
(Jia
Cunli
ang
)
409
0
0.
1
0.
2
0.
3
0.
4
0.
5
0.
6
0
20
0
40
0
60
0
80
0
100
0
t/(s
)
(a) T
he outpu
t power
cu
rve
of photovoltaic pan
els
0
0.
1
0.
2
0.
3
0.
4
0.
5
0.
6
0
0.
1
0.
2
0.
3
0.
4
0
.
5
t/(s
)
(b) T
he chan
ge of duty ratio
Figure 6. Simulation curve
of fixed step si
ze pertu
rb
ation and o
b
servation metho
d
(
ᇞ
D =0.0
02
)
Figure above
is the simul
a
tion wavefo
rm wh
en
ᇞ
D=0.002. The
initial illumination is
1000
W/m
2
b
u
t it decrea
s
es to
900
W/
m
2
whe
n
t=0
.
35s. Th
en it
is b
a
ck to
1000
W/m
2
when
t=0.45
s. Figu
re 6(a)
sho
w
s t
hat the out
put power of
photovolta
i
c
panel
s is m
a
ximum at t=0.03s
after the ru
n
n
ing of Boo
s
t
conve
r
ter a
n
d
the ma
xim
u
m po
wer t
r
a
cki
ng is
achi
eved. If the ligh
t
intensity changes sud
denly, Boost
converter can
quickly
track the illuminat
ion change and
contin
ue to
make
photov
oltaic pa
nel’
s
output maxi
mum po
we
r. But it is also
can b
e
seen
from
the figure, the maximum o
u
tput power
of photov
oltai
c
pan
els will
sho
c
k on a small scale wh
ose
amplitude
clo
s
e
s
to 20
W
due to the l
a
rge
step
le
n
g
th, whi
c
h
cannot me
et the re
qui
reme
nts
obviou
s
ly. In
orde
r to
redu
ce th
e
co
ncu
ssi
on
ᇞ
D
should be
reduced, for exampl
e, the
oscillati
on
amplitude
of
the sta
b
le
ou
tput of the
p
hotovoltaic
p
anel
s
i
s
controlled well as
ᇞ
D=0.00
02 a
n
d
the re
quirem
ents of
stabili
ty is meet. Howeve
r, t
he t
r
ackin
g
spe
e
d
is to
o
slow to ada
pt to the
weath
e
r who
s
e illumin
a
tio
n
is ch
ang
ed
freque
ntly.
4.2. Simulati
on of Improved Variable
Step Perturb
a
tion and O
b
serv
a
tion Method
Figure 7
i
s
MPPT sim
u
l
a
tion
cu
rve
of t
he im
pro
v
ed vari
able
step
p
e
rtu
r
b
a
tion a
nd
observation
method.
T
he large
r
ᇞ
D an
d
the smalle
r
ᇞ
D i
n
the
si
mulation
pro
g
r
am i
s
0.0
02
and
0.0002
re
sp
e
c
tively. It is shown in th
e f
i
gure
t
hat th
e maximum
power
output
of ph
otovoltaic
panel
s wa
s a
c
hieve
d
abou
t 0.05s with this algo
ri
thm.
Boost conve
r
ter could qui
ckly tra
ck th
e
cha
nge
of lig
ht intensity a
s
t=0.3s
and
t=0.5
s
a
nd
co
ntinued to
ma
ke
photovolta
ic pa
nel
s o
u
tput
the maximum
power. There is a nice stability in both cases.
(a) T
he outpu
t power
cu
rve
of photovoltaic
panel
s
(b) T
he chan
ge of duty ratio
Figure 7. Improved vari
abl
e step pe
rtur
bation ob
se
rvation simul
a
tion cu
rve
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 14, No. 2, June 20
16 : 404 – 41
0
410
5. Conclusio
n
Simulation e
x
perime
n
ts
shows that t
he
imp
r
oved
variable
st
ep pe
rturbati
on an
d
observation
method overcomes
the dra
w
ba
ck
of
the
two
p
r
eviou
s
pertu
rbatio
n and ob
servati
on
method
s
well
. In addition, i
t
alleviated th
e co
nflict that
the maximu
m po
wer tra
c
king
ca
nnot t
a
ke
into accou
n
t the accu
ra
cy of steady st
at
e and spee
d of resp
on
se e
fficiently.
Referen
ces
[1]
Qian Ni
ans
hu,
Liu Ku
o, Gu
o Jian
ye
. Re
search
of Ph
otovolta
ic cel
l
model
in
g an
d its outp
u
t
character
i
stic.
Journ
a
l of pow
er
. 2012; 1
0
(5)
:
78-82.
[2]
Ahmed EM, Sho
y
am
a M. Singl
e varia
b
le b
a
sed va
r
i
ab
le
step size ma
xi
mum po
w
e
r p
o
i
nt tracker for
stand-a
l
on
e ba
tter
y
storag
e PV s
y
stems.
IEEE Internatio
nal C
onfere
n
c
e
on Industri
a
l
Technol
ogy
.
201
1; 32(2
1
): 210-2
16.
[3]
Barcho
w
s
k
y
A, Parvin J
P
, R
eed GF
, et a
l
.
A co
mp
arativ
e study
of MP
PT
meth
ods f
o
r distri
bute
d
photov
olta
ic ge
nerati
o
n
. IEEE PES Innovativ
e Smart Grid
Techn
o
lo
gi
es (ISGT
)
. 2012: 1-7.
[4]
Parlak KS, Can H.
A new
MPPT metho
d
for PV array syste
m
u
nder
partia
l
ly shad
ed co
nd
itions
. 2012
3rd IEEE
Inter
natio
nal
S
y
mp
osium
on
Po
wer El
ectr
onics
f
o
r Distri
bute
d
Generati
o
n
S
ystems (PEDG).
201
2: 437-
441.
[5]
Jusoh A, S
u
tik
no T
,
Guan T
K
, Mekhil
ef A. A Revi
e
w
on F
a
vora
ble
Ma
xi
mum Po
w
e
r P
o
int T
r
ackin
g
S
y
stems in S
o
l
a
r Energ
y
Ap
pl
icatio
n.
T
E
LKOMNIKA T
e
leco
mmu
n
icati
on,
Co
mp
uting, El
ectronics a
n
d
Contro
l
. 201
4; 12(1): 6-2
2
.
[6]
A
w
a
n
g
Bi
n J
u
soh, Omer J
a
mal El
di
n Ibra
him Mo
hamm
e
d, T
o
leSutikn
o. Vari
abl
e Ste
p
Size
Perturb
and Observ
e MPPT
for PV Solar Appl
ic
ations.
T
E
LK
OMNIKA T
e
lecommunic
a
tio
n
,
Comp
utin
g,
Electron
ics an
d Contro
l.
201
5; 13(1): 1-12.
[7]
Emilio M, Gio
v
ann
i P, Giov
ann
i S. T
w
o-steps
al
gorithm
improvi
ng th
e
P&O stead
y
state MPPT
efficienc
y.
App
l
ied En
ergy
. 20
14; 113: 4
14-4
21.
[8]
Jancar
le L
S
, F
e
rna
ndo
A, Ani
s
C, Cícero C
A
. Maximum p
o
w
e
r
poi
nt trac
ker for PV s
y
stems usi
ng
a
hig
h
performa
n
c
e boost co
nve
r
ter.
Solar Ener
gy
. 2006; 8
0
(7)
:
772-77
8.
[9]
Ali AG, Sey
ed MS, Asma S.
A high performanc
e maximum po
w
e
r point tracker for
PV s
y
stems
.
Internatio
na
l Journ
a
l of Electr
ical Pow
e
r & Energy Syste
m
s
. 2013; 53: 23
7
-
243.
[10]
Pall
avee
B, R
K
Nema. Ma
ximum p
o
w
e
r
poi
nt
trackin
g
control t
e
chn
i
qu
es: State o
f
art in P
V
app
licati
on.
Re
new
abl
e an
d Sustain
abl
e Ene
r
gy Revi
ew
s
. 2013; 23: 2
24-2
41.
[11]
F
u
W
ang, Z
h
o
u
Li
n, GuoK
e.
Photovo
l
taic c
e
ll e
n
g
i
ne
erin
g
i
n
math
ematica
l
mode
l res
earc
h
.
Jour
nal
of
Electrotech
n
ics
. 2011; 8(1
0
): 211-2
16.
[12]
Che
n
Guila
n, Sun
xia
o
, Li
Ran. Ma
xim
u
m po
w
e
r p
o
in
t tracking con
t
rol of photov
oltaic p
o
w
e
r
gen
eratio
n s
y
stems.
Electroni
c technol
ogy a
pplic
atio
n.
200
1; 8: 33-35.
Evaluation Warning : The document was created with Spire.PDF for Python.