TELKOM
NIKA
, Vol.13, No
.1, March 2
0
1
5
, pp. 1~12
ISSN: 1693-6
930,
accredited
A
by DIKTI, De
cree No: 58/DIK
T
I/Kep/2013
DOI
:
10.12928/TELKOMNIKA.v13i1.1180
1
Re
cei
v
ed O
c
t
ober 2, 20
14;
Revi
se
d Ja
n
uary 4, 2015;
Acce
pt
ed Jan
uary 15, 201
5
Variable Step Size Perturb and Observe MPP
T
for PV
Solar Applications
A
w
a
n
g Bin J
u
soh*
1
, Ome
r
Jamal Eldin Ibrahim Mohammed
2
, Tole Sutikno
3
1,2
F
a
cult
y
of El
ectrical En
gin
e
e
rin
g
, Univers
i
ti T
e
knologi Ma
l
a
y
s
ia
813
10 Jo
hor B
ahru, Mal
a
ysia
3
Departme
n
t of Electrical En
gi
neer
ing, F
a
cult
y of
Industri
a
l
T
e
chnolog
y, U
n
iversit
a
s Ahm
ad Da
hla
n
,
Yog
y
ak
arta 55
164, Ind
ones
ia
*Corres
p
o
ndi
n
g
author, em
ail
:
aw
an
g@
fke
.
utm.my
1
, omer.j
amal.i
brah
im@
g
mail.c
o
m
2
, tole@e
e.uad.
ac.i
d
3
A
b
st
r
a
ct
In ord
e
r to
d
e
li
ver
max
i
mu
m
output
of p
hoto
v
oltaic
(PV) c
e
l
l
s, the
usa
g
e
o
f
maxi
mu
m po
w
e
r poi
n
t
tracking (MPP
T
) is essential.
T
he spee
d an
d stabil
i
ty
of the tracking tec
h
niq
ue ar
e hig
h
l
y
desire
d
. Perturb
and O
b
serve
(
P&O) is on
e of
the
most c
o
mmo
n
tracki
ng
t
e
chn
i
qu
es, b
u
t it suffers fro
m
the sl
ow
tracking
speed at s
m
all duty cycle step
and fluctuates w
hen subjected with large duty
step, which results
in
higher
losses
und
er d
y
na
mic w
eath
e
r
to w
h
ich the
photov
olta
ic
(P
V) cells
expos
ed. In this
pap
er, varia
b
le ste
p
si
z
e
Pertur
b an
d Observe is
in
troduce
d
thro
u
gho
ut Matl
ab/S
i
muli
nk si
mu
lati
on to ov
erco
me this pro
b
l
e
m to
achi
eve h
i
g
her
efficiency, rel
i
abl
e trackin
g
accura
cy
and
hig
her sp
ee
d
und
er fast cha
ngi
ng w
eath
e
r. In
comparis
on w
i
t
h
other v
a
ria
b
l
e
P&O techni
q
ues, the
pr
op
o
s
ed
meth
od fe
atures a
dyn
a
m
ic ste
p
si
z
e
f
o
r
mor
e
trackin
g
efficiency a
n
d
accuracy. D
oub
le d
i
od
e
mo
de
lli
ng is u
s
ed in th
is techni
que for b
e
tter
photov
olta
ic (P
V) cell
’
s
c
hara
c
teristic pr
edict
ion. In
this
stu
d
y, the
ad
apte
d
tech
niq
u
e
h
a
d
b
een
tested
to
w
i
de rang
e of sun irra
dia
n
ce a
nd op
erati
on te
mp
eratur
es.
Ke
y
w
ords
: do
ubl
e dio
de
mo
dell
i
n
g
, PV solar, MPPT
, perturb an
d obs
erv
e
, matl
ab/Si
mu
link
1. Introduc
tion
PV cells
are
recogni
se
d for havin
g no
n-line
a
r
cha
r
acteri
stics. A
t
one poi
nt, whi
c
h i
s
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 o
u
tp
ut [1]-[4]. Fig
u
re
1
sh
ows
the V-I
and
V
-
P cha
r
a
c
teri
stics
of a typical 1
50W PV cell. The V-I
and
V-P ca
n b
e
di
stingui
sh
ed from ho
w th
eir
load i
s
a
pplie
d,
either from
sh
ort circuit to o
pen ci
rcuit, i.e. from ze
ro l
oad
s to infinity.
(a)
(b)
Figure 1. V-I and V-P ch
aracteri
stics of
a
typical 150
W PV cell; (a). V-I (b). V-P
The p
o
int of
maximum o
p
e
ration
shifts with the
ch
a
nge
s of the
sun i
r
radia
n
ce, sola
r
panel surfa
c
e
temperatu
r
e
and deg
re
e of the sun i
rra
dian
ce [5]. Figure 2 illu
stra
tes the V-I and
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 1, March 2
015 : 1 – 12
2
V
-
P
ch
ar
act
e
rist
ic
s wit
h
re
spe
c
t
t
o
sol
a
r
ir
radi
an
ce
whe
n
the te
mperature
is fixed. Wh
en
the
sola
r irradian
ce ch
ang
es,
the point of maximum
shi
fts; for example, the PV cell output current
decrea
s
e
s
d
r
amatically
wh
en the
sola
r irradia
n
ce de
creases. In Fi
g
u
re
3, it is not
icea
ble that t
he
increa
se of PV cell temperature cau
s
e
s
a signific
ant drop in the o
u
tput voltage as well as out
put
power.
(a)
(b)
Figure 2. V-I and V-P ch
aracteri
stics un
der
vari
able radiation
s
; (a
). V-I (b). V-P
(a)
(b)
Figure 3. V-I and V-P ch
aracteri
stics un
der
vari
able t
e
mpe
r
ature;(a). V-I (b). V-P
2.
PV cell equi
v
a
lent circuits
The PV
cell
can b
e
m
odell
ed eith
er u
s
in
g si
ngle
dio
d
e
for ea
se
of
cal
c
ulatio
n, o
r
d
ouble
diode fo
r
better a
c
cu
ra
cy, efficien
cy an
d predi
ction
of
the cha
r
a
c
teristics du
rin
g
pa
rtial sha
d
ing
[4],[6].
2.1 PV cell
single diode equiv
a
lent circuit
As illustrated in Figure 4,
the PV cell
ca
n be impl
emented by
integrating a current
sou
r
ce, one
exponential
diode, an
d
parall
e
l-se
ri
es e
quivalen
t
resi
stan
ce,
whe
r
e V a
nd I
rep
r
e
s
ent the
terminal voltage and the
current. For certain tem
peratu
r
e an
d
irradia
n
ce, the
output cu
rren
t of the solar
panel
can b
e
cal
c
ulate
d
as
follows:
1
(1)
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Variabl
e Step Size Perturb and Ob
se
rve MPPT for PV
Solar Appli
c
a
t
ions (A
wan
g
Bin Jusoh
)
3
,
(2)
Figure 4. Single diod
e equ
ivalent circuit of Solar cell
2.2 PV cell d
ouble diode
equiv
a
lent circuit
In the d
oubl
e
diod
e
circuit
as sho
w
n i
n
Figu
re
5, th
e solar cell i
s
rep
r
e
s
ente
d
by two
expone
ntial d
i
ode
s. One is
an ideal dio
d
e
, while an
other is n
o
n
-
ide
a
l.
Figure 5.
Dou
b
le diod
e equ
ivalent circuit of solar
cell
IN
I
I
I
I
(3)
I
I
,
K
T
T
(4)
I
I
e
1
(5)
I
I
e
1
(6)
I
(7)
whe
r
e
current gene
rate
d by incide
nce light,
,
seri
es a
nd shu
n
t equiva
lent
resi
st
an
ce,
,
diode current,
1
.602
∗
1
0
(elect
r
o
n
cha
r
ge
),
1
.38
∗
10
/
,
cell surfa
c
e
temperature i
n
Kelvin
,
refe
ren
c
e tem
perature
in Kelvin, (
25°C)
,
temp
eratu
r
e co
efficient of SC in perce
nt chang
e per deg
re
e,
irra
dian
ce value W/m3
,
irradian
ce nomi
nal value (10
00W/m
3
)
,
number of se
rie
s
cell
s
,
numbe
r of parallel ce
lls
,
diode q
ual
ity factor, and
,
=
diode saturation
cu
rrent
.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 1, March 2
015 : 1 – 12
4
3.
Perturb a
nd Observ
e
(P&O)
Perturb
an
d
Observe
is the mo
st
com
m
only u
s
ed
method i
n
p
r
actice fo
r its
simpli
city
and e
a
se of
impleme
n
tation [7]-[10]. In this m
e
tho
d
, the output
voltage of t
he solar pan
el is
pertu
rbe
d
pe
riodi
cally, an
d then the o
u
tput power
is co
mpa
r
ed
to the previous
cycle.
Next,
corre
c
tive act
i
on is ta
ken t
o
enforce the
volt
age to move towa
rd th
e maximum o
peratio
n outp
u
t
voltage. The
comp
ari
s
o
n
can dete
r
min
e
the po
sition
of the op
eration poi
nt an
d
the directio
n
of
pertu
rbatio
n [11]-[14].
In total, there are th
ree
positio
ns a
n
d
tw
o dire
ctio
ns involved.
Figure 6 an
d
Table 1
demon
strate
the po
ssi
ble
po
sitions a
nd di
re
ct
ion
s
duri
ng the
pertu
rbatio
n
and o
b
servat
ion
pro
c
e
s
s an
d
also
the
req
u
i
red
actio
n
[1
5]-[18] in
ea
ch case to
achieve the
ma
ximum po
we
r
output.
Figure 6.
P&O MPPT posi
t
ioning
Table 1.
P&O
MPPT positioning a
nd taken actio
n
in e
a
ch
ca
se
Position
Variations
Action
1
0dP
No action taken
2
+dP /+dV
increase V
3
+dP /-dV
decrease V
4
-dP /-dV
increase V
5
-dP /+dV
decrease V
4.
Proposed v
a
riable step si
ze Perturb a
nd Observ
e
Larg
e
step d
u
ty cycle
has
a fast re
spon
se but
expe
ri
ences n
o
tice
able
fluctu
ation aro
u
n
d
the point
of
maximum
out
put of a PV
module. M
e
a
n
whil
e, sm
all
step
duty cy
cle h
a
s a
slo
w
er
tracking
spe
ed at the start of
operatio
n
and
un
der
dynamic
wea
t
her, but results in smoot
he
r
pertu
rbatio
n .The vibratio
n
of
the large
step an
d
the slo
w
sp
eed o
f
asso
ciate in
powe
r
lo
ss a
r
e
not
de
sire
d. Benefiting fro
m
t
he high
speed
of larg
e duty cycl
e
s
an
d the
so
ft tracki
ng of
the
small d
u
ty cycle, the p
r
opo
se
d vari
able Pert
u
r
b
and
Ob
se
rve
method
states that if
the
pertu
rbatio
n is towa
rd
the
maximum po
wer p
o
int, t
he duty cycle
is
increa
s
ed by
multiplicatio
n by
a factor (A
)
i.
e.
∆
P>0. Ho
wev
e
r,
if
∆
P<0, the
duty
cycl
e
sh
ould
be
divide
d
b
y
(A), in
whi
c
h
this
facto
r
is
a co
nsta
nt an
d greater tha
n
1. Fi
gu
re
7(a)
sho
w
s the
flow cha
r
t of the conventio
nal
P&O, whe
r
e fixed du
ty incre
m
en
t
is applie
d. In Figu
re
7(b
)
, the
flow
cha
r
t
of
the
prop
osed
techni
que
i
s
pre
s
ente
d
. In compa
r
i
s
o
n
, other
va
ri
able
P&Os,
as m
ention
e
d in
reference [13], utilize two
different
duty cycles, one i
s
relatively la
rge, used at the beginni
ng of
pertu
rbatio
n, and the othe
r is small, u
s
ed whe
n
the
system
vibra
t
es
aro
und m
a
ximum power
point.
Thi
s
method ha
s a
dynami
c
st
ep size
of
th
e duty
cycle,
whi
c
h i
n
cre
a
se
s it
s
spe
ed,
resulting in hi
gher effici
en
cy.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Variabl
e Step Size Perturb and Ob
se
rve MPPT for PV
Solar Appli
c
a
t
ions (A
wan
g
Bin Jusoh
)
5
(a)
(b)
Figure 7.
Flowchart of
Perturb a
nd Ob
serve;
(a
) conv
entional (b) M
odified
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 1, March 2
015 : 1 – 12
6
5.
Simulation and Res
u
lts
Photovoltaic sola
r system
con
s
i
s
ts of
PV
a
rray, switching co
nve
r
ter,
MPPT controlle
r
and loa
d
, whi
c
h may b
e
a
DC lo
ad, batt
e
ry ch
arg
e
r
o
r
inverte
r
. Th
e input(s)/
out
put(s) of MP
PT
controlle
r ma
y differ from one metho
d
to anothe
r.
Figure 8
sho
w
s the block dia
g
ram of a typical
MP
P
T
sy
st
e
m
.
Figure 8. Solar MPPT System
In this study, a 150 W sola
r panel
with the followi
ng specifi
c
ation
was u
s
ed in th
e simulatio
n
.
Table 2. PV solar pa
nel dat
a
Parameter Value
Max
i
mum po
wer
(P
ma
x
)
150 W
Voltage at Pmax
(V
MP
P
)
34.5 V
Curre
nt at Pma
x
(I
MP
P
)
4.35
A
Short circuit curr
ent (I
SC
)
4.75 A
Open circuit voltage (V
oc
)
40 V
Tempar
ature coe
fficient of I
SC
0.065±0.15
%/°
C
Tempar
ature coe
fficient of V
oc
-160±20mV/
°C
Tempar
ature coe
fficient of pow
e
r
-0.5±0.05
%
/°
C
Ideality
factor
1.2
In this
simul
a
tion, bo
ost
co
nverter
wa
s u
s
ed
to
fu
nctio
n
in
co
ntinuo
us
cu
rrent mo
de. T
h
e
values of the
ca
pa
citor
an
d
indu
ctor
ha
d be
en
ca
l
c
ul
ated for suita
b
le rang
e of
duty cycl
e, a
nd
then the
hig
h
e
st valu
es
were
cho
s
en.
Table
3
sh
ows the
spe
c
ification of
the
b
oost
co
nverte
r’s
comp
one
nts. Switchin
g fre
quen
cy of 40 kHz wa
s u
s
e
d
slightly abo
ve the audiab
le noise.
Table 3. Boost co
nverte
r comp
one
nt specifi
c
ation
Parameter Value
Sw
itching freq
ue
ncy
40 kHz
Inductor
160 µH
Capacitor
65
µF
Load resistance
1000
Ω
Output voltage
ripple
0.01V
Figure 9
sho
w
s
the
Simuli
nk
mod
e
l su
bsyste
m of doubl
e diod
e
modele
d
PV sola
r,
whe
r
e
the
m
a
thematical
e
quation
an
d
PV panel
pa
ramete
rs a
r
e
re
pre
s
e
n
ted
.
V-I an
d V-P
cha
r
a
c
teri
stics of PV mod
u
le ha
d be
en
determin
ed t
h
rou
gh o
pen
to
short
-
ci
rcui
t
test by rangi
ng
the output re
sistan
ce from
zero to infinity.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
9
30
Variabl
e Step Size Perturb and Ob
se
rve MPPT for PV
Solar Appli
c
a
t
ions (A
wan
g
Bin Jusoh
)
7
Figure 9.
Simulink
sub
s
yst
e
m of PV module
Figure 10
sh
ows the V
-
I
and V-P
ch
aracteri
stic
un
d
e
r Stand
a
rd
Test
Con
d
itio
n (ST
C
),
w
h
er
e
G=
1
000
W
/
m
2
and T
=
25
°C. Thi
s
result was o
b
tained fro
m
the circuit as
sh
own in Fig
u
re
9.
(a)
(b)
Figure 10. V-I and V-P cha
r
acteri
stic u
n
d
e
r Standa
rd
T
e
st Co
ndition
(STC); (a) V-I
characte
ri
stic
of PV module (b) V-P characteri
stic of P
V
module
Figure 11
shows the com
p
rehensive system blocks
of the solar
module, Pert
urb and
Observe alg
o
r
ithm and the
swit
chin
g co
n
v
erter.
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
9
30
TELKOM
NIKA
Vol. 13, No. 1, March 2
015 : 1 – 12
8
Figure 11. Simulink m
odel
of P&O MPPT system
Figure 12
sh
ows the outp
u
t of
large
step duty cycl
e. It had a fast re
sp
on
se
but also
experie
nced
notica
b
le flu
c
tuation a
r
ou
n
d
the p
o
in
t of
maximum
o
u
tput of a PV
modul
e. It took
0.05 m seconds
but stabil
i
zed
at
0.5 m
illi-second. T
h
e
vibration
of the large step reduced t
h
e
efficien
cy, wh
ich was n
o
t desired.
Figure 1
2
.
(a
) Solar mo
dul
e output po
wer (b
) Solar
module o
u
tpu
t
voltage
Con
d
ition: Du
ty cycle step
of
0.5V, 25°C, unity irradia
n
ce
Figure 1
3
sh
ows the
outp
u
t of small
st
ep d
u
ty
cycl
e
.
The
small
step duty
cycl
e ha
d a
slo
w
er tracki
ng sp
eed at
the st
art of operation an
d unde
r dyn
a
mi
c weathe
r, but resulte
d
in
smooth
e
r
perturbation i
n
changi
ng weat
her
at 2
5
°
C
and u
n
ity irra
dian
ce
.
Figu
re 14
show
s
t
h
e
output of the
variabl
e ste
p
duty cy
cle
P&O at
25
°
C
an
d u
n
ity irra
dian
ce. It wa
s o
b
serv
ed
fluctuating at the beginni
ng
of t
he operation but then rapidly settling
.
In compari
s
on to the sa
me
input a
s
appl
ied in the p
r
evious
conve
n
tional P&O,
the variabl
e P&O settled
after 0.025
milli-
se
con
d
.
Evaluation Warning : The document was created with Spire.PDF for Python.
TELKOM
NIKA
ISSN:
1693-6
930
Variabl
e Step Size Perturb and Ob
se
rve MPPT for PV
Solar Appli
c
a
t
ions (A
wan
g
Bin Jusoh
)
9
Figure 1
3
.
Output of small
step duty cycle; (a) Sola
r module o
u
tpu
t
power (b) S
o
lar mo
dule
output voltag
e - con
d
ition: duty cycle st
e
p
of 0.05V, 25°C, unity irra
dian
ce
Figure 14.
Plot of
(a) Solar module outp
u
t powe
r
(b)
Solar mod
u
le
output voltage
Con
d
ition: Variable d
u
ty cycl
e, 25°
C, uni
ty irradian
ce
Figure 15
sh
ows the
outp
u
t of
conventi
onal P&O
when
usin
g la
rge d
u
ty cycle
,
at 25°
C
and vari
able
irra
dian
ce.
It was obvio
us that t
he
system flu
c
tuated d
r
ama
t
ically with a
n
y
irra
dian
ce ch
ange. Thi
s
fluctuatio
n wa
s
asso
ciate
d
with powe
r
l
o
sse
s
. Figure 16
sho
w
s
t
h
e
output of co
n
v
entional P&O wh
en u
s
in
g small
duty cy
cle, at 25
°
C
an
d varia
b
l
e
irradian
ce.
Here,
Evaluation Warning : The document was created with Spire.PDF for Python.
ISSN: 16
93-6
930
TELKOM
NIKA
Vol. 13, No. 1, March 2
015 : 1 – 12
10
the sy
stem fl
uctuatio
n
wa
s le
ss ha
rsh
with a
n
y irrad
i
ance
chan
ge
. Ho
wever, th
e tra
cki
ng
sp
eed
wa
s slo
w
e
r
, in whi
c
h the p
o
we
r lost un
d
e
r ch
angi
ng
weath
e
r.
Figure 15.
Plot of
(a) Solar module outp
u
t powe
r
(b)
Solar mod
u
le
output voltage
Con
d
ition: Du
ty cycle step
of 0.5V
, 25°C, variable irra
dian
ce input
Figure 16.
Plot of
(a) Solar module outp
u
t powe
r
(b)
Solar mod
u
le
output voltage
con
d
ition: dut
y cycle step o
f
0.05V, 25°C, variable irra
dian
ce input
As sho
w
n in
Figure 17, under fa
st we
ather
chan
ge
, variable P&O delivere
d
the be
st
perfo
rman
ce,
ben
efiting fro
m
the
high
sp
eed
du
ring
th
e first sta
g
e
o
f
pertu
rbatio
n
and
fine
tuni
ng
arou
nd the p
eak p
o
wer po
int .
Evaluation Warning : The document was created with Spire.PDF for Python.