Int
ern
at
i
onal
Journ
al of
P
ower E
le
ctr
on
i
cs a
n
d
Drive
S
ystem
(I
J
PE
D
S
)
Vo
l.
11
,
No.
3
,
Septem
be
r 2020
, pp.
1305
~
13
12
IS
S
N:
20
88
-
8694
,
DOI: 10
.11
591/
ij
peds
.
v11.i
3
.
pp
1
30
5
-
13
1
2
1305
Journ
al h
om
e
page
:
http:
//
ij
pe
ds
.i
aescore.c
om
Senso
rless dual
axis sol
ar tr
ac
k
er
usi
ng impro
ved s
un
position
algorith
m
Chan
Men
Lo
on
1
, Mu
hama
d Za
l
an
i
Da
u
d
2
1
,2
Ea
stern
Corri
dor
Rene
wab
le
Ene
rgy
(
ECRE
)
,
Univer
sit
i
Mal
a
ysia
T
ere
ngg
anu
,
Mal
aysia
1,
2
Facul
ty
of
Oc
ea
n
Engi
ne
eri
ng
Te
chno
logy
a
nd
Inform
atics
(FT
KK
I),
Univer
sit
i
Mala
ysi
a Te
ren
gganu,
Ma
la
ysi
a
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Dec
4
, 2
01
9
Re
vised
M
a
r
17
, 2
0
20
Accepte
d
A
pr
8
, 2
0
20
Thi
s
p
ape
r
pre
s
ent
s
d
eve
lop
ment
of
a
p
rotot
yp
e
sensorl
ess
dua
l
axi
s
sol
a
r
tra
ck
er
for
ma
x
i
mum
ext
r
action
of
solar
ene
rgy
.
To
prov
e
the
c
once
pt
and
eva
lu
at
e
the
pro
posed
al
gori
thm,
a
low
cost
wid
el
y
av
ailabe
mat
eri
a
ls
were
used
which
was
progra
mmed
base
d
on
Arduino
microc
ontro
ll
er
.
The
porposed
al
gorit
h
m
work
s
base
d
on
two
sea
rch
me
th
ods
nam
el
y
the
g
loba
l
se
arch
that
appr
oxi
m
at
es
the
best
poi
nt
lo
ca
t
ion
in
a
reg
ion,
and
loc
a
l
sea
r
ch
th
at
furthe
r
d
eterm
i
nes
the
actua
l
sun’s
positi
on.
E
xper
imental
result
s
sh
owed
th
at
th
e
prop
osed
a
lgori
t
hm
giv
es
b
et
t
er
per
forma
n
ce
com
par
ed
to
the
exi
sting
sun
po
siti
on
al
go
rit
h
m
(SP
A)
-
base
d
me
thod
as
well
as
the
fix
ed
panel
sys
te
m
.
I
n
te
r
ms
of
tot
a
l
output
power
,
th
e
proposed
al
gorit
h
m
give
s
17.
96%
mor
e
eff
icient
th
an
th
e
f
i
xed
sys
te
m
and
6
.
38%
bet
t
er
tha
n
the
SP
A
-
base
d
sys
te
m
.
Furth
erm
o
re,
the
pe
rce
n
tage
err
or
of
the
exp
eri
m
enta
l
measured
ang
le
to
th
e
a
ct
ua
l
sun
az
im
u
th
angl
e
was
rel
a
ti
ve
ly
mi
n
imal
(l
ess
tha
n
3%)
during
clea
r
day
oper
ation.
The
s
ystem
was
prove
n
to
b
e
eff
ec
t
ive
in
tracki
n
g
the
sun
for
improved
ene
rgy
p
roduc
ti
on
of
solar
PV
p
ane
ls
and
th
e
propos
ed
al
gori
thm
al
s
o
c
an
b
e
us
ed
f
or
designi
n
g
the
tra
ck
er
wi
th l
arg
er
si
ze of
sol
a
r
PV
sys
te
ms.
Ke
yw
or
d
s
:
Du
al
-
axis
so
la
r
tracke
r
Sensorless
so
la
r
trac
ker
So
la
r
e
ne
rgy
Sun
po
sit
io
n
al
gorithm
(
S
PA)
This
is an
open
acc
ess arti
cl
e
un
der
the
CC
BY
-
SA
l
ic
ense
.
Corres
pond
in
g
Aut
h
or
:
M
uha
mad
Zala
ni D
a
ud,
Faculty
of
Oce
an
E
ngineeri
ng
Tech
nolo
gy a
nd Info
rmati
cs
(F
T
KKI)
Un
i
ver
sit
i
M
al
aysia Te
reng
ga
nu,
21030 K
uala
N
eru
s
, Te
reng
ga
nu, Mal
aysia.
Emai
l:
zala
ni@umt
.edu.
my
1.
INTROD
U
CTION
So
la
r
phot
ovol
ta
ic
(P
V
)
s
ys
te
m
is
hi
gh
l
y
po
t
entia
l
in
M
al
a
ysi
a
du
e
to
t
he
a
bundant
of
sun
sh
ine
with
aver
a
ge
ir
rad
ia
nce
pe
r
year
of
1643
kWh/m
2
[1,
2].
H
ow
e
ve
r,
m
os
t
s
olar
PV
in
sta
ll
at
ion
in
this
c
ountr
y
no
w
are
of
t
he
fixe
d
t
ype
s
ys
te
m
.
F
or
s
uc
h
i
ns
t
al
la
ti
on
s,
t
hough
it
sti
ll
giv
e
s
a
relat
ively
good
outp
ut,
howe
ve
r
hav
i
ng
it
fixe
d
at
an
a
ng
le
will
aff
ect
the
ye
arly
ou
t
pu
t
due
to
the
movem
ent
of
s
un.
I
n
gen
e
ral,
e
ff
ect
i
ve
s
un
tracker
ca
n
ma
ximize
t
he
e
ne
rgy
extracti
on
in
wh
ic
h
t
he
outp
ut
a
nd
perf
ormance
ca
n
be
imp
r
ov
e
d
up
to
40%
as
com
par
e
d
t
o
the
fi
xed
pa
nel
[3].
Alth
ough
ma
ny
res
earche
rs
re
por
te
d
that
so
la
r
tracker
ca
n
im
pro
ve
the
eff
ic
ie
ncy,
howe
ver
few
ot
her
fact
or
s
suc
h
as
c
os
t,
s
pa
ce
and
mec
hani
cal
aspects
ne
ed
to
be
c
on
si
der
e
d
wh
e
n
a
pp
l
ying
s
uch
te
c
hnol
ogy
[
4].
S
olar
trac
ker
ca
n
be
cat
eg
or
iz
e
d
into
fe
w
t
ypes
na
mely
the
a
ct
ive
trackin
g,
passi
ve
trac
k
in
g,
s
emi
passive
tr
ackin
g,
m
an
ua
l
trackin
g
an
d
chro
no
l
og
ic
al
trackin
g
.
The
so
la
r
trackr
ei
ther
c
an
w
ork
i
n
cl
ose
d
lo
op
trac
kin
g
or
ope
n
lo
op
trac
king
[
5
-
7].
T
he
m
os
t
c
ommo
n
t
yp
e
of
so
la
r
tracker
is
ei
the
r
a
si
ngle
axis
or
do
ub
le
a
xes
.
Ge
ne
rall
y,
t
he
track
e
r
ca
n
be
move
d
horiz
on
ta
ll
y,
ver
ti
ca
ll
y
or
bo
t
h [8].
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
11
, N
o.
3
,
Se
ptembe
r
2020
:
1
30
5
–
13
1
2
1306
In
previ
ous
res
earches
,
t
her
e
are
var
i
ou
s
ty
pe
s
of
s
olar
trac
ker
de
velo
pe
d
by
di
ff
e
ren
t
re
searche
rs
t
o
pro
ve
t
he
e
ff
ic
ie
ncy
of
so
la
r
tracker
ei
the
r
i
n
cl
os
e
d
l
oop
or
op
e
n
lo
op.
M
a
ny
of
them
us
e
d
li
ght
de
pe
nd
e
nt
resist
or
(L
DR)
as
se
ns
in
g
de
vi
ce
for
t
he
m
ot
or
to
ro
ta
te
t
he
so
la
r
pa
nel
t
o
the
res
pecti
ve
axis
[
9].
T
he
tr
acker
ro
ta
te
s
to
an
y
directi
on
base
d
on
the
data
feedbac
k
f
r
om
the
L
DRs.
Ba
sic
al
ly,
the
s
yst
em
com
pa
res
LDR
values
i
n wh
ic
h
the
er
ror
is
fe
d
to
the
mo
t
or
con
t
ro
ll
er.
T
he
mo
t
or
r
otate
s to
the
d
esi
red di
recti
on acco
rd
i
ng
l
y,
then
st
op
if
t
he
read
i
ng
of
LDRs
s
how
t
he
same
value
[10
-
12]
.
T
o
f
urt
her
imp
r
ov
e
the
data
ta
ken
from
the
L
DRs,
some
a
uthors
im
plement
the
f
uz
zy
l
og
ic
co
nt
ro
ll
er
to
bette
r
c
ontr
ol
the
m
ov
e
ment
of
t
he
so
la
r
pan
el
[8].
In
s
uch
a
ppr
oac
h,
the
co
ntr
ol
la
w
is
desc
ribe
d
by
a
kn
ow
le
dge
-
base
a
nd
a
fu
zz
y
lo
gic
in
f
eren
ce
mecha
nism
[
13,
14]
.
Re
ce
nt
researc
hes
al
so
intr
oduce
t
he
sens
or
le
ss
ap
proach.
It
is
sim
ple
as
no
se
nsors
are
need
e
d f
or
fee
db
ac
k
c
ontr
ol
of m
oto
r
s.
The
se
nsorless
sy
ste
m
c
omp
le
te
ly
de
pends
on
cal
c
ulati
on
of
sun
posit
ion
c
oor
din
at
e
s,
s
uch
a
s
the
sun
posit
io
n
al
gorith
m
(SPA)
[15
,
16].
SPA
is
use
d
t
o
cal
culat
e
the
so
la
r
azi
m
uth
and
ze
nith
an
gle
for
wh
ic
h
mo
st
of
this
syst
em
work
s
on
dual
a
xis
trac
ker
.
O
ne
mo
to
r
movin
g
horizo
ntal
axis
w
hich
is
dep
e
nd
o
n
zenit
h
a
ng
le
a
nd
an
oth
e
r
is
ver
ti
cal
axis
w
hich
is
base
d
on
the
azi
m
ut
h
a
ng
le
[
15
-
17
].
T
he
pro
ble
m
with
existi
ng
SPA
base
d
meth
od
is
the
occura
nce
of
misal
ig
nm
e
nt.
F
or
e
xa
mp
le
,
the
cal
culat
ion
of
t
he
su
n
locat
ion
is
done
us
i
ng
SP
A,
howe
ve
r
s
olar
tr
acker
di
recti
on
sho
uld
be
kn
own
fi
rst
to
pre
ven
t
misal
ig
nme
nt.
In
ot
her
w
ords
,
if
the
so
la
r
tr
acker
is
misal
ign
e
d
f
rom
0°
North
,
it
will
aff
ect
t
he
perf
ormance
of
th
e
so
la
r
tracker
due
t
o
that
misal
ignm
ent.
T
o
overc
ome
s
uch
misal
ignment
iss
ue,
an
im
pro
ved
al
gorithm
is
pr
opose
d
namely
the
gl
ob
al
a
nd
local
searc
h
to
trac
k
the
misal
ignment
of
t
he
s
olar
trac
ke
r.
T
he
e
xp
e
rime
ntal
su
n
po
sit
io
n
base
d
on
the
pro
po
se
d
al
gorithm
is
us
e
d
t
o
i
de
ntif
y
t
he
misal
ignment
fro
m
t
he
SPA
a
nd
fix
th
e
er
ro
r
so
t
hat p
a
nel c
an be
re
-
al
ig
ne
d
to
0°
North
.
2.
RESEA
R
CH MET
HO
D
This
resea
rc
h
introd
uces
a
n
imp
r
ov
e
d
method
for
the
existi
ng
S
PA
-
base
d
s
ys
t
em
[18
-
20
].
The
pro
posed
tracke
r
c
on
t
ro
l
scheme
has
be
en
im
pl
eme
nted
in
a
de
velo
pe
d
protot
yp
e
of
se
nsorless
du
al
axis
so
la
r
t
rack
e
r.
2.1.
System h
ardw
are d
e
sign
of
proposed
pr
otot
ype
The
c
on
tr
ol
c
ircuit
of
t
he
protot
yp
e
is
ba
sed
on
Ard
ui
no
mic
ro
c
ontrolle
r
a
nd
it
s
com
patible
com
pone
nts.
Figure
1(a
)
shows
t
he
sc
he
mati
cs
o
f
the
sy
ste
m
c
on
tr
ol
ci
rcu
it
.
T
he
ci
rcu
it
com
pr
ise
s
of
Ardu
i
no
boar
d
as
a
c
ontr
oller,
tw
o
ste
pper
mo
to
rs
(
28BY
J
-
48)
with
dri
ver
(
UL
N
2003
A)
eac
h,
the
c
urren
t
sens
or
(IN
A21
9) for
meas
ur
e
ment
of cu
rr
e
nt
/vo
lt
age
outp
ut
, and
a
12
V,
250 m
A (3 W
)
s
olar pa
nel.
(a)
(b)
(c)
Figure
1. Trac
ker sy
ste
m circ
uit, (
a
)
Sc
he
ma
ti
c (b
)
M
ot
or
a
nd drive
r (c)
C
urren
t
sens
or
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
Sensorless
du
al
a
xi
s s
ola
r
tra
cker usi
ng im
prove
d
s
un posit
ion
algorit
hm
(
Chan Me
n Lo
on
)
1307
The
A
rduin
o
boar
d
is
eq
uipp
ed
with
AT
me
ga32
8P
micr
oc
on
t
ro
ll
er
with
14
dig
it
al
pins,
6
a
nalo
g
pin
s
an
d
oth
e
r
f
un
ct
io
n
pi
ns
.
The
micr
oc
ontrolle
r
ca
n
be
pro
gr
am
med
usi
ng
A
rduin
o
I
DE
a
vaila
ble
f
rom
Ardu
i
no.cc
we
bs
it
e
[21].
T
he
ste
pp
e
r
m
oto
r
c
hosen
t
o
op
e
rate
the
s
ol
ar
pa
nel
is
of
28
B
YJ
-
48
ty
pe
with
the
co
rr
es
pond
ing
dr
i
ver
s
ho
wn
i
n
Fig
ure
1(b
).
T
he
pro
posed
prototype
sy
ste
m
has
tw
o
m
otors
that
a
re
us
e
d
to
co
ntr
ol
the
zenit
h
a
xis
a
nd
azi
m
uth
axi
s,
re
sp
ect
ivel
y.
The
ste
pper
mo
to
r
move
i
n
discrete
ste
ps,
th
us
accurate
a
nd
s
uitable
f
or
t
he
pro
po
se
d
s
ys
te
m.
F
or
t
he
m
ot
or
dr
i
ver,
it
is
base
d
on
UL
N
2003A
ste
pper
mo
t
or
dr
i
ver
wh
ic
h
is
commo
nly
use
d
to
dr
iv
e
th
e
28V
YJ
-
48
ty
pe
ste
pp
e
r
m
otor.
U
LN
2003
A
is
an
arr
a
y
of
s
even
NPN
Dar
li
ngto
n
tr
ansist
or
s
ca
pab
le
of
50
0mA
a
nd
50V
ou
t
pu
t,
it
ca
n
ac
hi
eve
a
very
hi
gh
c
urre
nt
a
nd
volt
age
ou
t
pu
t.
To
ma
ke
sure
the
propose
d
pr
oto
ty
pe
is
cost
ef
fe
ct
ive,
this
dri
ve
r
usual
ly
c
ome
s
in
pa
cka
ge
with
28V
YJ
-
48
m
ot
or
as
s
how
n
i
n
Fi
gure
1(b
)
and
the
y
a
re
e
asi
ly
be
obta
in
ed.
The
meas
ureme
nt
of
outpu
t
is
base
d
on
INA
219
cu
rr
e
nt
se
ns
or
as
sho
wn
in
Fig
ure
1(c)
.
T
his
sens
or
i
s
ca
pab
le
of
m
easur
i
ng
bo
t
h
t
he
DC
vo
lt
age
a
nd
c
urre
nt
mea
sured
via
I
2
C
se
rial
prot
oco
l
with
1%
pr
eci
sio
n
[22].
It
on
l
y
ne
eds
a
n
in
pu
t
range
betwee
n
3
–
5
V
to
power
up
an
d
ca
n
mea
su
re
up
to
3.2
A
cu
rr
e
nt
an
d
26
V
ta
r
geted
vo
lt
age
.
A
pre
ci
sion
amplifie
r
mea
s
ur
es
t
he
volt
ag
e
acro
ss
t
he
0.
1
Ω,
1%
resist
or.
Since
the
a
mp
li
fier
ma
xi
mu
m
i
nput
dif
fer
e
nce
is
±
320
m
V,
this
mea
ns
it
can
meas
ur
e
up
to
±3.2A.
INA21
9
has
i
nter
nal
progra
mmable
gain
set
at
the minim
um d
ivisi
on of
8
a
nd the
ma
ximum cu
rr
e
nt
of
±
400 mA
and t
he
r
es
olu
ti
on is
0.1 mA.
Th
is
s
ens
or
is
cheap
a
nd
eas
y
t
o
pro
gr
a
m
s
ince
the
li
brar
y
ca
n
be
dow
nl
oad
e
d
from
th
e
A
rdui
no
li
brary.
T
he
fi
nal
desig
n
structu
re
of pr
opos
e
d protot
ype s
olar
t
rack
e
r
is
pr
ese
nted
in
Fi
gure
2.
Figure
2.
Desi
gn str
uctu
re
of
final
protot
yp
e
sy
ste
m
2.2.
Impro
ved sun
po
si
tion al
go
r
ithm
fo
r
mot
or
contr
ol
The
pro
po
se
d
sy
ste
m
is
ba
sed
on
the
s
ol
ar
pa
nel
it
sel
f
that
works
as
a
sen
sor
t
og
et
her
with
the
INA
219
cu
rr
e
nt
se
nsor
.
T
wo
al
go
rithms
are
pro
po
se
d
that
is
firstly
,
t
he
al
gorithm
wh
ic
h
util
iz
es
global
search
an
d
sec
ondly
the
al
go
rithm
that
e
mpl
oy
s
desc
rete
s
te
ps
searc
h.
T
he
c
on
ce
pt
of
global
searc
h
comes
from
the
c
omp
ass,
the
re
are
f
our
re
gions
fro
m
the
co
mp
a
ss
wh
ic
h
is
East
,
South,
West
a
nd
N
or
th
.
T
he
best
sp
ot
f
or
s
olar
energ
y
ca
pture
m
us
t
be
la
id
i
n
one
of
these
reg
i
on
s
a
nd
th
e
gl
ob
al
searc
h
al
go
rith
m
is
use
d
t
o
search
f
or
t
he
best
spot
a
pproximat
el
y.
Her
e
,
the
c
oord
i
nate
po
i
nts
of
a
ngle
s
are
co
nvert
ed
to
ste
ps
f
or
easi
er
search
by t
he
s
te
pp
e
r
m
otor.
As
s
how
n
i
n
Figure
3,
gl
obal
searc
h
al
go
ritm
firstl
y
set
the
st
ar
tVal
ue
an
d
nextVal
ue
var
ia
bles.
The
st
ar
tVal
ue
is
init
ia
ll
y
assigne
d
as
zer
o,
wh
e
reas
nextV
alu
e
is
set
as
100
ste
ps.
T
he
pro
gr
a
m
will
ge
ner
at
e
rand
om
num
be
r
f
r
om
t
hese
ra
ng
e
an
d
the
nu
mb
e
r
will
be
s
ent
to
the
mo
t
or
to
move
t
o
that
posit
ion.
The
n,
current
meas
ureme
nt
al
gorit
hm
will
run.
T
he
measu
re
d
curre
nt
f
rom
INA
219
se
nsor
will
be
sa
ved
to
a
var
ia
ble
cal
le
d
cu
rrA
mp
.
Her
e
,
if
cu
rrA
mp
is
bigger
t
han
highA
m
p
,
the
highA
m
p
will
be
re
placed
by
currA
mp
an
d
t
he
value
of
the
pr
e
vi
ou
s
hi
ghAm
p
will
be
st
or
e
d
i
ns
ide
a
va
riable
cal
le
d
r
ef
Amp
.
On
t
h
e
oth
e
r
hand,
if
c
ur
rA
mp
is
le
ss
t
han
highA
mp
,
t
he
hi
gh
A
m
p
a
nd
re
fAmp
will
rem
ai
n
uncha
ge
d. The
posit
io
n
is
sto
red
to
posit
ionBe
s
t
accor
ding
to
the
posit
ion
w
he
re
hi
ghAm
p
is
store
d.
If
the
hi
ghA
mp
is
rep
l
aced
b
y
the
cu
rrAm
p
,
the
po
sit
io
n
will
be
ove
rwrit
t
en,
the
n
the
pr
e
vious
posit
ion
Be
st
will
be
st
ored
t
o
the
posit
ion
L
ast
.
Af
te
r
t
he
c
urren
t
meas
ure
ment
is
c
ompl
et
ed,
the
ne
xt
startVal
ue
a
nd
ne
xt
Value
will
be
increase
d
base
d
on
the
pr
e
vi
ou
s
nextVal
ue
.
F
or
exa
mp
le
,
the
fi
rst
sta
rt
V
al
ue
is
0
an
d
ne
xt
Valu
e
is
100,
after
the
cu
rr
e
nt
me
asur
e
ment
is
done
,
the
ne
w
st
ar
tVal
ue
will
be
sta
rt
from
10
1
an
d
nextV
alue
will
be
en
d
a
t
200.
Simply
in
t
his
case,
the
start
Valu
e
incre
ase
s
by
1
from
t
he
previ
ou
s
nex
tValue
a
nd
the
pr
e
sent
nextV
alu
e
is
increase
d
by
100
f
r
om
the
p
re
vious
nex
tValue
.
A
no
t
he
r
two
va
riabl
es,
i.e
fi
na
lV
al
ue
an
d
en
dP
oin
t
is
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
11
, N
o.
3
,
Se
ptembe
r
2020
:
1
30
5
–
13
1
2
1308
dep
e
ndin
g
on
the
us
er
in
pu
t.
If
the
use
r
w
anted
to
do
s
earch
in
only
90
de
grees
w
hich
is
1024
ste
ps,
the
e
nd
P
oin
t
f
or
the
searc
h
will
be
1024
ste
ps
a
nd
t
he
fi
nalVal
ue
will
stop
at
1000.
If
t
he
nextVal
ue
r
eached
1000,
the
next
Valu
e
will
be
r
eplace
d
t
o
10
24
instea
d
of
1000
beca
us
e
it
will
be
easi
er
f
or
t
he
ste
pper
mo
to
r
to ope
rate.
Af
te
r
,
t
he
glob
al
searc
h
al
go
r
it
hm
c
omplet
ed
t
he
run
proce
ss,
the
discrete
ste
p
searc
h
al
gorithm
will
be
e
xec
uted
t
o
fin
d
t
he
real
s
un
posit
io
n.
H
ere,
t
he
s
olar
t
rack
e
r
s
ys
te
m
tu
r
ns
back
to
the
posit
ion
L
ast
then
sta
rted
t
he
dis
crete
ste
p
sear
ch
al
gorithm.
In
this
proce
dure
,
the
br
ig
ht
spot
is
a
ssum
ed
t
o
be
la
id
be
tween
the
posit
ionLa
st
and
posit
ionBe
st
,
meani
ng
that
the
spot
be
tween
the
highest
an
d
the
s
econd
highest.
Eve
n
though
,
the
br
i
gh
t
spot
is
no
t
reall
y
la
id
i
n
po
sit
io
nBest
,
t
he
al
gorithm
will
kee
p
se
ar
chin
g
unti
l
it
r
eached
the
br
i
gh
t
spot
wh
ic
h
is
the
re
al
sun
posit
ion.
I
n
t
his
discret
e
ste
p
al
gorit
hm,
t
he
ste
pper
mo
to
r
is
m
ove
d
with
on
l
y 2 s
te
ps
w
hich
is
0.17
578 de
gr
ee
pe
r
ti
m
e unti
l t
he
th
re
sh
ol
d
is
met.
Figure
3
.
Flo
w
char
t
of the
pro
po
s
ed
alg
or
it
hm
3.
RESU
LT
S
AND DI
SCUS
S
ION
The
performa
nc
e
of
pro
pose
d
al
gorith
m
is
evaluate
d
by
com
par
i
ng
t
he
resu
lt
s
with
t
he
pre
vious
method
as
well
as
the
co
nv
e
nt
ion
al
fix
ed
set
t
ing
s
olar
pa
nel
.
In
t
his
case,
f
ixed
s
olar
pan
e
l,
SPA
dr
i
ven
so
la
r
tracker
an
d
propose
d
al
go
rithm
tr
acke
r
we
re
set
diff
e
ren
t
ly.
Fi
xed
so
la
r
pa
nel
an
d
SP
A
dr
ive
n
so
la
r
tracke
r
wer
e
set
at
th
e
same
directi
on
wh
ic
h
is
303
North
west
and
the
pro
posed
prot
otype
tra
cker
was
se
t
wit
h
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
Sensorless
du
al
a
xi
s s
ola
r
tra
cker usi
ng im
prove
d
s
un posit
ion
algorit
hm
(
Chan Me
n Lo
on
)
1309
diff
e
re
nt
direc
ti
on
.
Fix
ed
s
olar
pa
nel
s
urfa
ce
is
fixe
d
at
0°
zenit
h
a
xis
an
d
didn’t
m
ov
e
d
f
or
t
he
r
est
of
the
e
xperime
nt
.
W
hile,
SP
A
dri
ve
n
so
la
r
t
rack
e
r
is
m
ov
ed
base
d
on
the
cal
c
ulati
on
of
the
s
un
posit
ion
al
gorithm.
On t
he othe
r han
d
,
the pr
opos
e
d p
ro
t
otype trac
ke
r
w
as m
oved
base
d o
n
the
propo
s
ed
alg
or
it
hm.
3.1.
Eva
lu
tatio
n
of t
he
pr
oposed
im
proved s
un positi
on
algori
th
m
The
s
ys
te
m
sta
rts
with
global
search
t
o
est
imat
e
the
best
s
po
t
in
t
he
pred
et
ermine
d
area
.
It
is
base
d
on
the
c
on
ce
pt
of
c
ompass
t
o
ra
ndom
l
y
de
te
rmin
e
the
be
st
spot
f
or
w
hi
ch
the
m
axim
um
cu
rr
e
nt
c
a
n
be
conve
rted
by
the
so
la
r
pa
nel
from
the
s
un
r
adiat
ion
.
Fo
r
e
xam
ple,
i
f
t
he
sp
ot
or
di
recti
on
is
div
ide
d
int
o
f
our
reg
i
on
s of
East
,
South
,
We
st
and
Nor
t
h,
the
re
exists
the
be
st
po
i
nt
within
t
he
re
gions
that
can
pr
oduce high
est
current
from
th
e
so
la
r
pan
el
.
The
global
search
runs
base
d
on
t
his
co
nce
pt
to
ap
pro
xima
te
the
best
c
oor
din
at
e
po
i
nt
fir
st.
Once
the
best
s
pot
is
rou
gh
l
y
est
imat
ed,
th
e
pr
ogra
m
c
on
ti
nue
s
with
t
he
l
oca
l
search
.
In
the
local
search
,
the
rea
l
best
point
in
the
re
gion
is
de
te
rmin
e
d
s
o
that
the
azi
m
uth
axis
mo
t
or
c
an
ope
rate
an
d
locat
e
the
s
olar
p
a
nel
exactl
y
to
the
sun’
s
directi
on.
The
m
ovem
ent
of
so
la
r
tra
cker
axe
s
a
re
base
d
on
ze
nith
a
nd
azi
mu
th
a
ng
le
,
res
pecti
vely.
The
pre
dicti
on
of
ze
nith
ang
le
is
str
ai
gh
tf
orwa
rd
as
it
chan
ge
s
about
15
º
pe
r
hour
[
23].
T
hus,
i
n
this
case,
S
PA
is
us
ed
t
o
cal
c
ulate
the
zenit
h
an
gle.
Howe
ver,
f
or
azi
m
ut
h
an
gle,
as
it
is
var
ie
d
throu
ghout
th
e
year,
the
se
arch
for
the
a
zi
mu
th
a
ng
le
was
ba
sed
on
the
ne
w
al
go
rithm
pro
po
se
d.
T
o
e
valuate
t
he
performa
nce
of
th
e
se
arc
h
al
gorithm
pro
po
se
d
in
this
stu
dy,
a
per
ce
ntage
er
r
or
of
the
act
ual
so
l
ar
azi
mu
t
h
an
gle
an
d
ex
peri
mental
meas
ur
e
d
s
olar
azi
mu
th
a
ngle
are
cal
culat
ed
us
in
g
the for
mu
la
as
(1).
Perce
ntage
e
rror =
|
−
|
×
100%
(1)
W
he
re
V
A
is
th
e
exp
e
rime
ntal
measu
red
val
ue
of
the
s
olar
azi
mu
th
a
ng
le
,
an
d
V
E
is
the
exact
value
(actual
cal
culat
ed)
of
the
s
olar
azi
mu
th
.
Ta
ble
1
s
hows
the
r
esults
of
cal
cul
at
ion
a
nd
meas
ur
e
ment
of
azi
mu
th
ang
le
s
ove
r
one
da
y
of
ex
per
ime
nt.
Fro
m
the
ta
ble
it
cl
early
sho
w
s
that
th
e
al
gorith
m
ca
n
ac
cur
at
el
y
appr
ox
imat
e
t
he
azi
mu
th
an
gl
es
duri
ng
the
da
y
with
er
r
or
of
le
ss
tha
n
3%.
Exce
pt
du
rin
g
the
f
ourt
h
ho
ur
of
measu
reme
nt
(
at
12:5
0
P
M
)
wh
ic
h
sho
ws
a
sig
nificant
de
viati
on
of
ab
out
1210%.
T
his
oc
cu
ran
ce
is
du
e
to
a
s
udde
n
c
ha
nge
i
n
w
eat
her
from
s
unny
t
o
cl
ou
dy
at
a
r
ound
12:5
0
P
M.
T
he
ap
pro
xi
mati
on
of
this
an
gle
dep
e
nds st
ron
gl
y
on
weathe
r
c
onditi
on as t
he a
lgorit
hm
us
es
the curre
nt
me
asur
e
ment
fro
m the sola
r
pan
el
.
Table
1.
Perce
ntage
e
rro
r of
measu
red s
olar
azimuth
a
ng
le
s w
it
h t
he
cal
c
ulate
d values
Exp
erim
en
tal
m
e
a
su
red so
lar
azi
m
u
th
ang
le (
º)
Actu
al so
lar
a
zim
u
th
ang
le (
º)
Percentag
e e
rr
o
r
(
%)
6
5
.39
6
6
.25
1
.30
0
6
0
.17
6
0
.41
0
.48
5
5
1
.91
5
2
.80
1
.67
8
2
7
.25
2
.08
1
2
1
0
.2
1
4
3
5
3
.18
3
5
2
.58
0
.17
0
3
0
3
.76
3
1
0
.49
2
.16
8
2
9
8
.39
3
0
1
.76
1
.11
6
2
9
3
.67
2
9
5
.74
0
.70
6
3.2.
Perfo
r
ma
nce
eva
lu
ati
on
To
e
valuate
th
e
pe
rformance
of
the
propos
ed
im
pro
ved
s
un
posit
io
n
al
gorithm
,
a
co
mp
a
rison
ha
s
been
done
f
or
three
e
xp
e
rime
ntal
set
up
s
of
fixe
d
ty
pe,
S
P
A
an
d
pro
po
se
d
al
gorith
m.
T
ho
s
e
three
dif
f
eren
t
exp
e
rime
ntal
set
up
s
wer
e
ca
r
ried
out
on
4
th
Ju
ly
2019
at
the
hous
i
ng
a
r
ea
in
M
e
ng
le
mbu
I
poh,
Per
ak.
On
the
da
y
of
e
xperime
nt,
it
was
a
bri
ght
day,
no
sh
a
di
ng
on
t
he
so
l
ar
pa
nel
a
nd
the
s
olar
pan
e
l
was
expose
d
t
o
f
ul
l
su
nli
gh
t.
The
res
ults
rec
or
de
d
durin
g
t
he
exp
e
rime
nt
are
as
s
how
n
in
Table
2,
Ta
ble
3
a
nd
Table
4,
res
pec
ti
vely.
Table
2
.
Re
s
ults o
f
fi
xe
d pan
e
l
Measu
red Va
lu
e
Local Ti
m
e
(
h
o
u
rs)
0920
1043
1130
1250
1330
1456
1540
1640
Vo
ltag
e (
V)
1
2
.5
1
2
.93
13
1
3
.32
1
3
.05
1
2
.82
1
2
.82
1
3
.06
Cu
rr
en
t (
m
A)
5
4
.36
1
1
0
.75
1
6
6
.43
2
0
3
.26
2
3
0
.70
2
0
8
.29
2
2
3
.75
1
0
4
.05
Po
wer
(
m
W
)
679
1432
2163
2707
3010
2670
2868
1358
*
Pan
el Dir
ectio
n
(
º
)
3
0
3
NW
3
0
3
NW
3
0
3
NW
3
0
3
NW
3
0
3
NW
3
0
3
NW
3
0
3
NW
3
0
3
NW
Misalig
n
m
en
t (
º)
-
57
-
57
-
57
-
57
-
57
-
57
-
57
-
57
W
eath
er
Su
n
n
y
Clo
u
d
y
Su
n
n
y
Clo
u
d
y
Su
n
n
y
Su
n
n
y
Su
n
n
y
Su
n
n
y
Total Po
wer
(
m
W
)
1
6
8
9
0
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
11
, N
o.
3
,
Se
ptembe
r
2020
:
1
30
5
–
13
1
2
1310
Table
3
.
Re
s
u
lt
s of SPA
drive
n
s
olar
t
rack
e
r
Measu
red Va
lu
e
Local Ti
m
e
(
h
o
u
rs)
0920
1043
1130
1250
1330
1456
1540
1640
Vo
ltag
e (
V)
1
3
.04
1
3
.5
1
2
.95
1
3
.23
1
3
.1
1
2
.89
1
2
.93
1
2
.98
Cu
rr
en
t (
m
A)
6
4
.58
1
2
8
.35
2
0
4
.60
2
2
3
.26
2
3
6
.05
2
2
0
.42
1
9
7
.99
1
5
8
.57
Po
wer
(
m
W
)
842
1732
2649
2953
3092
2841
2559
2058
*
Pan
el Dir
ectio
n
(
º
)
9
.26
3
.38
3
5
5
.82
3
2
3
.57
2
9
5
.58
2
5
3
.49
2
4
4
.76
2
3
8
.74
Misalig
n
m
en
t (
º)
-
57
-
57
-
57
-
57
-
57
-
57
-
57
-
57
W
eath
er
Su
n
n
y
Clo
u
d
y
Su
n
n
y
Clo
u
d
y
Su
n
n
y
Su
n
n
y
Su
n
n
y
Su
n
n
y
Actu
al Azi
m
u
th
A
n
g
le
(º)
6
6
.25
6
0
.41
5
2
.80
2
.08
3
5
2
.58
3
1
0
.49
3
0
1
.76
2
9
5
.74
Total Po
wer
(
m
W
)
1
8
7
2
9
Table
4
.
Re
s
ults o
f p
r
opos
e
d
a
lgorit
hm
s
olar
tracker
Measu
red Va
lu
e
Local Ti
m
e
(ho
u
rs)
0920
1043
1130
1250
1330
1456
1540
1640
Vo
ltag
e (
V)
1
3
.04
1
3
.8
1
2
.97
1
3
.4
1
3
.05
1
2
.8
1
3
.04
1
2
.89
Cu
rr
en
t (
m
A)
6
5
.01
1
4
9
.76
2
2
2
.43
1
8
0
.79
2
5
0
.43
2
3
3
.40
2
3
4
.86
1
8
4
.96
Po
wer
(
m
W
)
847
2066
2884
2422
3268
2987
3062
2384
*
Pan
el Dir
ectio
n
(
º
)
6
5
.39
6
0
.17
5
1
.91
2
7
.24
3
5
3
.18
3
0
3
.76
2
9
8
.39
2
9
3
.66
Misalig
n
m
en
t (
º)
0
-
57
15
-
10
-
20
40
0
0
W
eath
er
Su
n
n
y
Clo
u
d
y
Su
n
n
y
Clo
u
d
y
Su
n
n
y
Su
n
n
y
Su
n
n
y
Su
n
n
y
Actu
al Azi
m
u
th
A
n
g
le
(º)
6
6
.25
6
0
.41
5
2
.80
2
.08
3
5
2
.58
3
1
0
.49
3
0
1
.76
2
9
5
.74
Total Po
wer
(
m
W
)
1
9
9
2
4
*
d
irection
of so
lar
p
an
el f
acin
g
af
te
r
r
u
n
n
in
g
the alg
o
rithm
Fo
r
fixe
d
pa
ne
l
and
SP
A
dri
ve
n
trac
ker,
the
y
we
re
init
ia
ll
y
place
at
ra
ndom
directi
on
du
rin
g
te
sti
ng.
Both
pan
el
s
w
ere
not
fi
xed
f
aci
ng
tr
ue
N
or
t
h
w
hic
h
is
zer
o
de
gree
bu
t
al
ign
e
d
to
t
he
sa
me
directi
on
w
hich
is
303
NW.
H
owever,
f
or
the
propose
d
al
gorit
hm
d
rive
n
s
ola
r
tracke
r,
it
wa
s
set
to
diff
e
re
nt
directi
on
as
sh
ow
n
in
Table
3
a
nd
the
pur
pose
of
doin
g
this
is
to
deter
mine
t
he
po
wer
l
os
s
of
t
he
azi
m
uth
ang
le
misal
ig
nm
e
nt.
The
misal
ig
nme
nt
re
fers
to
the
an
gle
of
incident
of
sun
ra
ys
over
t
he
so
la
r
pa
nel’
s
s
urface,
w
he
re
best
po
sit
io
n
sho
ul
d
be
0º
misal
ingment.
C
omp
arin
g
t
he
t
otal
po
wer
ou
t
pu
t
of
Ta
bles
2,
Table
3
an
d
T
able
4,
cl
early
t
he
ov
erall
po
wer
outpu
t
of
t
he
pr
opos
e
d
al
gorit
hm
dri
ve
n
s
olar
tr
acke
r
pro
du
ce
s
t
he
best
res
ults
com
par
e
d
to
f
ixed
pa
nel
a
nd
SP
A
dr
i
ven
al
gorith
m
s
ol
ar
pa
nel,
res
pe
ct
ively.
H
owever,
due
t
o
t
he
ra
pid
change
in
en
vi
ronme
nt
durin
g
12.
50
P
M
,
t
he
pro
posed
al
gorithm
gav
e
l
ow
e
r
powe
r
outp
ut
wh
ic
h
is
24
22
mW
a
s
c
ompa
red
to
fixe
d
pa
nel
(
2707
m
W)
an
d
t
he
S
PA
dri
ve
n
panel
(29
53
m
W)
res
ults
.
T
he
s
udde
n
change
of
wea
ther
al
s
o
af
fec
ti
ng
t
he
e
xperi
ment
a
ngle
whic
h
gav
e
27.24
º
c
ompare
d
t
o
the
act
ual
azi
mu
th
ang
le
wh
ic
h
is
2.08º.
T
hus,
t
urns
a
big
c
ha
ng
e
i
n
the
pe
r
centage
e
rror
of
1210%.
As
mentio
ne
d
in
[24],
the
fluct
uation
in
so
la
r
i
ntensi
ty
d
ue
to
the
m
ov
i
ng
cl
oud
is unp
red
ic
ta
ble an
d
t
o
ens
ure
the
lo
west
pe
rc
entage
error,
the
sea
rc
h
al
gorith
m
s
houl
d
be
ma
de
i
n
s
hortest
ti
me
possible
to
ta
ckle
the p
r
oble
m.
I
n
t
his
ex
pe
riment,
so
me
delays
wer
e
i
nclu
ded
in
the
al
gorit
hm
to
al
lo
cat
e
ti
me
fo
r
sam
p
li
ng
t
he
data
and
a
s
a
co
nc
equ
e
nc
e
aff
ect
s
t
he
ove
rall
resu
lt
s
ta
ke
n.
Fig
ur
e
4
s
hows
a
summ
ary
of
gra
ph
of
the
e
xperime
ntal
meas
ur
e
d
so
la
r
azi
mu
th
a
ng
le
and
t
he
real
s
olar
azi
mu
t
h
an
gl
e,
resp
ect
ivel
y.
It
cl
early
s
hows
that
the
pro
po
s
ed
al
go
rith
m
can
est
ima
te
the
ac
tual
azi
m
uth
a
ng
le
co
rr
ect
l
y
pro
vid
e
d
t
hat
no
disturba
nce
in
te
rms
of
a
bru
pt
weathe
r
c
hanges
occurs
duri
ng the e
xp
e
rieme
nt.
Figure
4
.
S
olar
azimuth
a
ng
le
measu
reme
nt result
s
as c
omp
ared t
he
cal
c
ulate
d
data
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
P
ow Elec
& Dri S
ys
t
IS
S
N: 20
88
-
8
694
Sensorless
du
al
a
xi
s s
ola
r
tra
cker usi
ng im
prove
d
s
un posit
ion
algorit
hm
(
Chan Me
n Lo
on
)
1311
Her
e
,
to
furth
er
inc
rease
the
accur
ac
y
of
the
meas
ured
a
ng
le
,
the
th
res
ho
l
d
value
s
houl
d
be
se
t
accuratel
y
due
to
the
dif
fer
e
nce
of
c
ur
r
Am
p
a
nd
hi
gh
A
m
p
s
om
et
imes
i
s
relat
ively
s
m
al
l.
Othe
r
tha
n
that,
the
fl
uctuati
on
of
c
urren
t
se
ns
or
will
al
so
aff
ect
t
he
ou
t
put.
On
e
wa
y
t
o
ov
e
rc
om
e
t
he
pro
blem
is
to
ta
ke
the
a
ver
a
ge
va
lue
of
cu
rren
ts
or
al
lo
wing
a
s
mall
dela
y
f
or
se
ns
or
to
proce
ss
t
he
da
ta
.
Fig
ure
5
sh
ows
the
ov
e
rall
po
wer
outp
ut
be
tween
fixe
d
pa
nel,
SP
A
dr
i
ven
s
olar
t
rac
ker
an
d
pro
po
sed
al
gorithm
dr
i
ven
so
la
r
t
rack
e
r.
Figure
5
.
Com
par
is
on of
pow
er
ou
t
pu
t
for di
ff
e
ren
t
meth
ods
Fr
om
Fi
gure
5,
cl
earl
y
ca
n
be
ob
se
r
ved
that
the
pro
po
sed
al
gorit
hm
dr
i
ven
so
la
r
tracker
ga
ve
the
hi
gh
est
po
wer
outp
ut
tha
n
the
oth
e
r
tw
o
beca
us
e
m
os
t
of
t
he
ti
me
the
so
la
r
pa
nel
is
facin
g
ap
pro
xi
mate
ly
to
th
e
s
un
dir
ect
ion
.
I
n
c
on
t
rast,
t
he
misal
ign
me
nt
of
t
he
fi
xed
panel
and
S
PA
dr
i
ve
n
so
la
r
tra
ck
er
wer
e
aff
ect
in
g
their
act
ual
perf
or
m
ance.
F
or
the
pro
po
se
d
met
hod,
the
s
mall
er
the
misal
ignment
will
gi
ve
bette
r
resu
lt
s.
T
he
m
isa
li
gn
me
nt
ca
n
be
at
trib
uted
to
the
direct
powe
r
l
os
s,
w
hich
is
acc
ordi
ng
to
[25],
ab
ou
t
8º
misal
ign
me
nt t
o
the
incid
ence
angle,
will
g
iv
e ab
ou
t l
ess
tha
n 1%
powe
r
lo
ss.
4.
CONCL
US
I
O
N
This
pa
per
has
prese
nted
the
con
ce
pt
of
dua
l
axis
sen
sorle
ss
s
un
trac
ke
r
for
a
m
axim
um
ca
pture
of
so
la
r
e
nerg
y.
The
pro
po
se
d
global
searc
h
a
nd
l
ocal
searc
h
al
gorithm
for
the
micr
ocontr
oller
ha
s
bee
n
fou
nd
eff
ect
ive
with
the
c
omparis
on
of
the
e
xp
e
ri
mental
re
su
lt
s
sh
ow
the
e
ff
ic
i
ency
of
17.
96
%
bette
r
tha
n
t
he
fixe
d
typ
e
sy
ste
m
. C
ompari
ng w
it
h
the r
es
ults
of
previ
ou
sl
y dev
el
op
e
d SP
A
-
base
d
s
ys
te
m als
o
s
hows
i
mpro
ve
ment
of
a
bout
6.3
8%
.
T
he
misal
i
gnment
of
the
azi
mu
th
a
ng
le
s
searc
h
re
su
lt
s
was
al
so
mi
nimal
w
hic
h
is
l
ess
tha
n
3%
duri
ng
cl
e
ar
da
y
of
ope
rati
on.
Furthe
r
more,
the
pr
oto
ty
pe
sy
ste
m
has
be
e
n
buil
t
ba
sed
on
lo
w
co
st
mate
rial
s
that
can
be
ob
ta
ine
d
f
orm
mo
st
e
-
co
mmerc
e
sit
es.
T
he
propo
sed
met
hod
is
scal
able
to
f
ul
l
siz
e
sy
ste
m a
nd
us
e
the sa
me alg
or
it
hm
for t
he
mi
cro
c
ontrolle
r
.
ACKN
OWLE
DGE
MENTS
The
a
uthor
s
w
ou
l
d
li
ke
t
o
ac
knowle
dge
fin
ancia
l su
pport
by Unive
rsiti
M
al
aysia
Tere
ngga
nu.
REFERE
NCE
S
[1]
M.
S.
N.
Sam
sudin,
et
al.
,
“
Pow
er
gen
er
at
i
on
source
s
in
Mala
ysia
:
St
at
u
s
and
p
rospec
t
s
for
sus
ta
in
able
deve
lop
me
nt
,
”
J
ournal
of Adv
an
ce
d
Revie
w
on
S
ci
en
ti
fic
Re
searc
h
,
vol
.
25
,
pp
.
11
-
28,
2016
.
[2]
S.
C.
Chua
and
T.
H.
Oh
,
“Sol
ar
ene
rgy
ou
tl
ook
in
Mala
ysi
a,”
R
e
newabl
e
and
Sustainabl
e
Ene
rgy
Re
v
ie
ws,
vo
l.
16
,
pp.
564
-
574
,
20
12.
[3]
G.
M.
Azha
r
an
d
A.
M.
A.
Rah
ma
n,
“T
h
er
per
f
orma
nc
e
of
thr
e
e
diffe
r
ent
sol
ar
pane
ls
for
sola
r
el
e
ct
ri
ci
ty
appl
y
i
ng
solar
tr
ac
k
ing
de
vic
e
under
the M
al
aysi
an
c
li
m
ate condit
ion
,
”
En
ergy
and
Environm
ent
al Re
search
,
vol
.
2
,
2012
.
[4]
J.
H.
L
ee,
et
a
l.,
“Per
forma
n
ce
of
dual
-
axi
s
solar
t
rac
ker
ver
sus
sta
ti
c
sol
ar
sys
tem
by
segme
nt
ed
clea
rne
ss
ind
ex
in
Mala
ysia
,
”
Inte
r
nati
onal Journal
of Ph
otoe
n
ergy
,
vol. 2013, 2013.
[5]
J.
P.
Nare
ndr
asinh,
et
a
l.
,
“Pass
ive
solar
tracki
n
g
sys
te
m,
”
Inte
r
nati
onal
Journal
of
Eme
rging
Technol
ogy
and
Adv
anc
ed Engi
n
ee
ring
,
vol
.
5
,
p
p.
138
-
145
,
201
5.
[6]
L.
Noe
l,
e
t
al
.
,
“
Semi
-
passive
sol
ar
tr
ac
k
ing
con
c
ent
ra
tor,”
En
ergy
P
roc
edi
a
,
vol
.
57,
pp
.
275
-
284
,
2014.
[7]
A.
Z.
Haf
ez
,
et
al.
,
“Sola
r
tr
ac
k
ing
sys
tems:
Technol
ogi
es
and
tra
ck
ers
dr
ive
ty
pes
-
A
rev
i
ew,”
R
ene
wabl
e
and
Sustainabl
e
Ener
gy
Revi
ews,
vol
.
91
,
pp
.
754
-
78
2,
2018
.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8
694
In
t J
P
ow
Ele
c
&
D
ri
S
ys
t,
V
ol
.
11
, N
o.
3
,
Se
ptembe
r
2020
:
1
30
5
–
13
1
2
1312
[8]
N.
A.
Rousan,
e
t
al
.
,
“Adva
n
ce
s
in
sol
ar
p
hotov
olt
ai
c
tracki
ng
sys
te
ms:
A
rev
i
e
w,”
Re
newab
le
and
Sustainanble
Ene
rgy
Revie
ws,
vol. 82, pp. 254
8
-
2569,
2018
.
[9]
T.
X.
Y.
C
indy,
et
al
.
,
“De
vel
op
me
nt
of
m
ic
ro
co
ntrol
ler
-
base
d
so
la
r
tr
ac
king
sys
t
em
using
LDR
sensor,”
Journal
of
Tele
communic
a
t
ion,
El
e
ct
roni
c
a
nd
Engi
n
ee
ring
,
vol.
9
,
pp
.
2
-
7
,
2
017.
[10]
A
.
A.
Anes
and
A.
S.
W
ae
l
,
“
Design
and
im
p
le
m
ent
a
ti
on
of
an
aut
o
ma
t
ically
al
ign
ed
solar
tra
ck
ing
sys
te
m
,
”
Inte
rnational
Jo
urnal
of Powe
r
El
e
ct
ronics
and
Dr
iv
e
Syst
em
(I
J
PE
DS),
vo
l. 10,
pp.
2055
-
2064
,
2019.
[11]
S.
Purnima
,
et
al.
,
“Dua
l
axi
s
solar
tracki
ng
s
ystem
for
solar
pane
l
,
”
Bul
l
et
i
n
of
Elec
t
ic
al
Engi
ne
ering
an
d
Informatic
s,
vo
l. 5, pp. 403
-
411,
2016.
[12]
O.
Ese
osa
and
U.
Rol
an,
“De
sign
and
simul
at
ion
of
solar
monitori
ng
tracki
ng
sys
te
m,”
IOSR
Jour
nal
of
Elec
tric
al
and
Elec
troni
cs
Engi
ne
ering
(IO
SR
-
JEEE)
,
vol
.
4,
pp
.
35
-
39
,
20
13.
[13]
A.
Z
aka
r
ia
h,
et
al.
,
“Dua
l
-
ax
is
s
ola
r
tracki
ng
sys
te
m
base
d
on
fu
zz
y
logic
cont
ro
l
a
nd
li
ght
dep
e
ndent
resistors
a
s
fee
dba
ck
p
at
h
eleme
n
ts,”
IEEE
S
tude
nt
Confe
ren
ce
on
R
ese
arch
and
Dev
e
lopmen
et
,
Sco
RE
D,
201
5.
[14]
P.
H.
Sneh
al
,
et
al
.
,
“FPG
A
base
d
sun
tr
ac
k
ing
sys
te
m
using
fu
zz
y
logic
,”
Int
ernati
onal
Journal
of
S
ci
en
ti
f
ic
&
Technol
ogy
R
ese
arch
,
vol
.
2
,
pp
.
217
-
220,
2013
.
[15]
A.
Merl
aur
d,
et
al.
,
“
Equa
t
ions f
or
solar tracking
,
”
Sensors
,
vo
l. 1
2,
pp
.
4074
-
409
0,
2012
.
[16]
Y.
R
izal
,
e
t
a
l.,
“Applica
t
ion
o
f
solar
posi
ti
on
al
gor
it
hm
for
s
un
-
tra
ck
ing
sys
t
em
,
”
Ene
rgy
Proce
dia
,
vol
.
32
,
pp.
160
-
165
,
20
13.
[17]
H.
Fatha
bad
i,
“
Novel
high
a
cc
u
rat
e
sensorl
ess
dual
-
ax
is
solar
tr
a
cki
ng
sys
te
m
co
ntrol
le
d
by
ma
x
i
mum
power
poi
nt
tra
ck
ing
uni
t
of
photovol
taic
sys
te
ms,”
App
li
ed
E
nergy
,
vo
l. 173,
pp.
448
-
459
,
20
16.
[18]
A.
A
.
Ri
zvi,
et
a
l.
,
“Sun
posit
ion
a
lgori
th
m
fo
r
su
n
tr
ac
ki
ng
appl
i
c
at
ions,
”
IECON
2014
-
40
th
Annua
l
Conf
ere
nc
e
of
the
I
EE
E
Industr
ial
El
e
ct
ronics
S
oci
e
ty
,
2014.
[19]
K.
Jaz
aye
r
i,
et
a
l.
,
“MATLAB/S
im
uli
nk
base
d
si
mul
ation
of
sola
r
inciden
ce
angle
and
th
e
sun’s
p
ositi
on
in
the
sky
with
respe
ct
to
observa
ti
on
poi
nts
on
th
e
e
art
h
,
”
In
te
rnationa
l
Confe
renc
e
on
Re
newab
le
Ene
r
gy
R
ese
arch
an
d
Apl
licati
ons
,
pp
.
173
-
177,
2012.
[20]
B.
Ja
mi
l
,
e
t
a
l.,
“E
sti
mation
o
f
solar
r
adi
a
ti
o
n
and
op
ti
mu
m
t
il
t
angl
es
for
south
fa
ci
ng
s
urfa
ce
s
in
hum
i
d
subtropic
a
l
cl
i
m
at
i
c
r
egi
on
of
I
ndia
,
”
Engi
n
ee
r
i
ng
Sc
ie
nc
e
and
Techno
logy
,
an
Int
ernati
onal
J
ournal,
vo
l.
19,
pp.
1826
-
1835
,
2016.
[21]
P.
Ram
y
a
and
R
.
Anan
th
ME
,
“
The
im
pl
ementation
of
sola
r
tra
c
ker
using
Ardui
no
with
servomo
tor,
”
Int
ernati
on
al
Re
search Journal
of
Engi
n
ee
ring
and
Technol
og
y (IRJ
ET)
,
vol
.
3
,
pp.
969
-
972
,
20
16.
[22]
Te
xas
Instru
me
n
ts,
“INA219
ze
r
o
-
drift
,
bidirec
tional curre
n
t/
pow
er
mon
it
or
with
I
2
C
interfa
c
e,”
20
08.
[23]
J.
F.
Muthna,
et
a
l.,
“De
sign
and
im
p
lemen
ta
ti
on
of
smar
t
elec
troni
c
sol
ar
t
rac
k
er
base
d
on
Arduino
,
”
TEL
KOMNIKA,
vol.
17
,
pp
.
2486
-
2496,
2019
.
[24]
S.
Dasgupta
,
et
al
.
,
“Dua
l
axis
sun
tracki
ng
sys
te
m
with
P
V
cell
as
the
sensor,
u
ti
l
iz
ing
hybrid
e
le
c
tric
al
cha
ra
cteri
sti
cs
o
f
th
e
cell
to
d
et
er
mi
ne
insol
ation,
”
IE
EE
Int
ernati
onal
Conf
ere
nce
on
Sust
ainabl
e
En
ergy
Technol
ogi
es
(I
CSET),
2010
.
[25]
M.
H.
T
ani
a
and
M.
S.
Alam,
“S
un
tr
ac
king
sch
e
me
s
for
photovo
lt
aic
panels,”
3
rd
Inte
rnationa
l
C
onfe
renc
e
on
the
Dev
el
opmen
ts i
n
Re
n
ewabl
e
En
e
rgy
Technol
og
y (ICDRET),
2014
.
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
Chan
Men
Loo
n
re
ce
iv
ed
his
Bac
he
lor
Degr
e
e
in
Apl
li
ed
Sc
ie
nc
e
(
Elec
tron
i
cs
Phys
ic
s
and
Instrume
nt
at
ion)
from
Univer
sit
i
Mal
aysia
T
erengganu.
He
is
cur
ren
t
ly
pursui
ng
Master
of
Scie
nc
e
(Ren
ewa
ble
Ene
rgy)
in
Univer
siti
Ma
lays
ia
Te
r
engga
nu
.
His
cur
r
ent
r
ese
ar
ch
intere
st
is
solar
en
erg
y.
Muhama
d
Z
alan
i
Daud
co
mpl
e
ted
his
bac
h
el
or’s
degr
ee
in
el
e
ct
ri
ca
l
and
e
lectr
oni
c
engi
n
ee
r
ing,
Rit
sumei
k
an
Un
ive
rsity
,
Kyoto,
Japa
n
in
Mar
ch
2003.
In
Februa
ry
2010
he
co
m
ple
t
ed
his
MS
c
by
rese
arc
h
a
t
School
of
Elec
tr
ical,
Compu
te
r
an
d
T
el
e
com
mun
icati
ons
Engi
n
ee
ri
ng,
Univ
ersit
y
of
Wol
longong
,
Aus
tra
li
a
.
L
at
er
in
2014,
com
pl
e
te
d
his
PhD
in
r
ene
wabl
e
en
erg
y
from
Faculty
of
Engi
ne
eri
ng
a
nd
Buil
t
Envi
ron
me
nt
,
Univer
si
ti
Keba
ngsaa
n
Ma
l
aysia
(UK
M).
He
is
cur
r
ent
ly
a
senior
lectur
er
at
the
Univer
sit
i
Mal
aysia
Te
r
e
ngganu
(UM
T),
Facul
ty
o
f
Oce
a
n
Engi
n
ee
ring
Te
chno
logy
and
Inform
atics
in
which
h
is
r
ese
arc
h
in
te
rests
ar
e
i
n
ren
ewa
ble
ene
rgy,
sm
art
ene
rgy met
er
d
e
vel
opm
ent
and e
ner
gy
ef
ficien
cy
.
Evaluation Warning : The document was created with Spire.PDF for Python.