Inte
r
n
a
t
io
n
a
l Jour
n
al of
Po
we
r E
l
ect
ron
i
c
s
a
nd Dr
iv
e
System
(
IJ
PED
S
)
V
o
l. 10,
N
o.
4, D
e
c
e
m
ber
201
9,
pp. 213
8~
2147
ISS
N
: 2088-
86
94,
D
O
I
:
10.11
59
1
/
i
j
p
ed
s.v1
0
.
i
4
.
p
p
2
1
3
8
-
2
1
4
7
2
138
Jou
rn
al h
o
me
pa
ge
: h
t
t
p
:/
/
i
ae
sc
o
r
e
.
c
o
m/jo
urn
a
l
s
/i
nd
ex
.p
hp
/
I
JPED
S
Experimental investigation of a
r
tificial inte
lligence applied i
n
MPPT techniques
S
.
D
ella Krac
h
ai
1,4
,
A.
B
o
u
dg
h
e
ne
S
ta
mbo
u
li
2
,
M. Della Krachai
3
, M.
B
e
k
h
ti
4
1,
2
,
3
Dep
a
rtm
e
nt
of
Elect
ron
i
cs, U
n
i
v
e
rsit
y
o
f
S
c
i
e
n
ce and
Tech
no
lo
gi
es –
Mo
h
a
me
d Bo
ud
i
a
f
,
A
l
g
eri
a
4
S
p
at
ial
Dev
e
lopm
ent
Cen
t
er, A
l
geri
a
Ar
ti
cl
e Info
ABSTRACT
A
r
tic
le history
:
Re
ce
ive
d
M
ar
17,
201
9
Re
vi
se
d A
p
r
14,
2019
A
c
c
e
pte
d
Ma
y
28,
2
0
1
9
Nan
o
-sat
ellit
e
s
are
key
f
eatu
r
es
f
o
r
s
h
a
rin
g
t
he
s
p
ace
dat
a
a
nd
s
cientif
i
c
research
es
.
Th
ey
e
m
b
ed
s
u
b
sy
st
ems
th
at
a
re
f
ed
f
ro
m
s
o
l
a
r
p
a
nel
s
and
bat
t
eries.
P
ower
g
enerat
ed
f
ro
m
th
ese
p
a
nel
s
i
s
su
bject
t
o
en
vi
ronmental
con
d
i
t
io
ns
,
m
o
s
t
i
m
p
o
r
tant
o
f
t
h
em
a
re
i
rrad
i
ance
a
n
d
t
e
mp
eratu
re.
Op
ti
mizin
g
t
he
u
s
a
ge
o
f
this
power
v
ersu
s
envi
ron
m
en
tal
v
a
riati
on
s
is
a
pri
m
a
r
y
t
a
s
k
.
Syn
c
hron
ou
s
D
C
-DC
bu
ck
c
o
n
v
e
rter
i
s
u
s
ed
t
o
cont
r
ol
t
h
e
po
wer
tran
sf
erred
f
r
om
P
V
p
a
nel
s
t
o
th
e
s
ubsy
s
t
e
ms
w
hile
m
ai
n
t
a
in
i
n
g
operation
at
m
a
x
ima
l
p
ower.
In
t
his
pa
per,
a
rtifici
a
l
inte
l
l
i
g
e
nce
t
echn
i
qu
es:
neu
r
al
n
et
work
s
and
ad
apt
i
ve
n
e
u
ral
f
u
zzy
i
nf
e
r
ence
sy
st
em
s
(AN
FIS)
a
re
us
ed
t
o
accom
p
l
i
s
h
t
h
e
t
racki
n
g
task
.
S
i
m
u
latio
n
and
ex
peri
men
t
al
r
e
s
u
l
ts
dem
o
nst
r
ate
t
h
eir
effici
ency,
robus
tnes
s
an
d
tracki
n
g
qu
ality.
K
eyw
or
d
s
:
Ad
a
p
t
i
v
e
n
eu
ra
l
fu
zzy
i
nfe
r
en
c
e
system
s
A
r
tificia
l i
n
tel
l
i
ge
nce
Hardw
a
re in the
l
o
op
Neu
r
al
n
et
w
o
rk
s
S
ynchr
on
ous
b
uck c
o
n
v
erter
Cop
y
ri
gh
t © 2
019 In
stitute
of
Ad
va
nced
En
gi
neeri
n
g
an
d
S
c
ien
ce.
A
l
l
rights
res
e
rv
ed.
Corres
pon
d
i
n
g
Au
th
or:
S. Del
la Krach
ai,
De
p
a
rt
men
t
of
El
ect
ro
ni
c
s
, El
ec
t
r
i
c
a
l
Eng
in
ee
ri
ng
Fa
c
u
l
t
y
,
U
n
i
v
e
r
si
t
y
o
f
S
c
ienc
e
a
nd Te
c
h
n
o
l
o
g
ie
s
– M
oham
e
d
Bo
u
d
i
a
f,
El Mnaouar
, BP
1
505,
Bir El
Dj
ir
3
1000, O
r
a
n,
Alger
i
a
.
Ema
il: sa
idia
.d
ella
kra
c
ha
i
@
u
n
i
v
-us
t
o.dz
1.
INT
R
ODUCTION
P
o
w
e
r
ma
nagem
e
nt
i
n
a
sa
te
lli
te
i
s
a
ma
n
d
at
ory
tas
k
.
T
h
e
over
a
l
l
p
o
w
er
i
s
c
o
nsum
ed
f
r
o
m
PV
pane
ls
w
h
i
c
h
e
xh
ibi
t
s
n
o
n
-l
i
n
ea
r
char
ac
ter
i
st
ics
de
pen
d
i
n
g
o
n
t
hee
n
vir
onm
en
ta
l
c
o
n
d
i
tio
ns.
T
h
ese
ha
s
a
con
s
i
d
era
b
le
i
mpa
c
t
on
t
he
m
axima
l
g
e
n
e
r
ated
p
ower.
To
opt
im
i
z
e
t
h
e
e
ffi
c
i
en
cy
o
f
th
e
sup
p
l
y
s
y
s
t
e
m,
extrac
tin
g
t
h
e
m
a
xima
l p
o
w
e
r
is re
a
lize
d
t
ro
ug
h c
ontr
o
l
l
ers
cal
led
Max
i
ma
l
Pow
e
r P
o
in
t
Trac
kers.
S
e
v
era
l
t
e
c
h
ni
que
s
an
d
me
t
h
ods
h
a
v
e
be
e
n
p
rop
o
se
d
for
tr
acki
n
g
t
he
m
a
x
i
m
al
p
ow
e
r
ge
ne
rate
d
b
y
PV
p
a
n
e
l
s.
T
he
t
radi
t
i
on
a
l
s
ol
ut
i
o
n
s
p
ro
po
se
d
fo
r
t
h
ese
t
r
ac
k
e
r
s
are
Hi
ll
c
l
i
mbi
n
g
[1],
P
er
t
u
r
b
a
n
d
O
bserve
[2-5]
a
n
d
Incr
em
enta
lco
n
duc
tanc
e
m
e
t
h
o
d
s
[
6
].
T
he
se
m
eth
o
d
s
pr
e
se
nt
a
n
osc
ill
a
t
i
o
n
p
r
obl
e
m
a
ro
un
d
t
h
e
ma
xi
m
a
l
p
o
w
e
r
point
a
s a
dra
w
back,
how
e
v
er, they
are
li
g
h
t
an
d
e
a
sy t
o
im
plem
en
t.
Ot
h
e
r
so
l
u
t
i
ons
a
re
b
a
s
e
d
on
a
r
t
i
f
i
ci
al
i
n
t
elli
g
e
n
c
e
t
echn
i
qu
e
s
s
uc
h
a
s
f
u
z
z
y
l
o
g
ic
[
7-9]
,
a
n
t
c
o
lo
ny
op
tim
iza
t
i
o
n
[
10],
a
n
d
ne
ural
n
etw
o
rk
s
[1
1-
14].
These
tec
hni
que
s
im
pro
v
e
co
nsi
d
era
b
ly
t
he
e
ffic
i
enc
y
o
f
the
system
; how
e
v
er, im
p
l
e
me
nt
a
tio
n
of t
hese
al
gor
it
hms
requi
re
m
or
e
c
o
mp
u
t
at
io
ns an
d
c
odi
ng
.
In
t
his
pa
pe
r,
t
he
o
n
line
a
p
pl
ic
ati
o
n
of
a
rti
f
ic
i
a
l
in
tel
l
ige
n
c
e
tec
h
n
i
ques
:
n
eur
a
l
ne
tw
or
k
s
a
nd
fuzz
y
l
ogi
c
i
s
i
nv
e
s
t
i
g
at
e
d
t
h
r
o
ugh
s
i
m
ul
a
tio
n
a
nd
h
a
rd
wa
re
i
n
th
e
l
o
o
p
e
x
p
er
im
entat
i
o
n
usi
n
g
a
low
-
c
o
s
t
s
ol
u
t
i
o
n.
Th
i
s
a
ppr
oach
a
ll
ow
s
not
o
n
l
y
m
a
x
i
m
i
z
i
n
g
t
he
s
ys
t
e
m
e
f
fic
i
e
n
c
y
,
but
a
l
s
o
gi
v
e
s
t
h
e
op
po
rt
u
n
it
y
to
a
dju
s
t
t
h
e
para
me
t
e
rs onl
ine
in
o
rder
t
o m
eet
t
he
de
s
ire
d
re
q
u
i
r
e
m
e
nts
and
s
h
o
rten t
h
e
t
i
m
e
to m
arke
t
del
i
v
ery.
2.
CHARACTE
R
IS
T
I
CS OF
PV
P
ANEL
IN U
S
E
Ea
ch
P
V
pane
l
con
s
is
t
s
o
f
a
se
t
o
f
s
er
i
e
s
c
o
nne
cte
d
c
e
lls
t
o
pr
o
duce
c
o
nsid
er
able
a
m
o
un
t
of
v
ol
t
a
ge.
Th
e
ma
i
n
o
u
t
pu
ts
w
e
a
r
e
i
n
terest
ed
i
n
are
I
–
V
an
d
P–
V
c
u
rv
es
w
h
ich
reve
al
t
hree
i
m
por
t
a
nt
p
o
i
nts.
S
hort-
Evaluation Warning : The document was created with Spire.PDF for Python.
Int J P
o
w
Ele
c
&
D
ri S
yst
IS
S
N
:
2088-
86
94
Expe
rim
e
nta
l
i
n
ve
s
t
i
g
at
ion o
f
artif
ic
ia
l
i
n
t
e
ll
ige
n
c
e
a
p
p
l
ie
d
in MPPT tech
n
i
q
u
es
(
S
.
D
e
lla K
r
ac
hai)
2
139
circuit
current
(
I
sc
)
at
w
h
i
c
h
t
he ou
t
p
u
t
vo
lta
ge
i
s
n
u
l
l
, ope
n-c
i
r
c
u
it vo
l
t
a
g
e
(V
oc
)
a
t
w
h
i
c
h
n
o
o
u
t
p
u
t
c
u
r
r
e
n
t
i
s
pro
duce
d
a
n
d
t
he
P
ma
x
poi
n
t
a
t w
h
ic
h the
pa
n
e
l ope
ra
tes
at i
t
s
m
axima
l
e
ffi
c
ienc
y.
The
PV panel use
d ET-M53675 is
from
ET-
S
o
l
a
r comp
any (Tabl
e
1).
Tab
l
e
1. ET-
M536
75
P
V
p
anel
b
ack-s
he
e
t
d
ata
at
100
0
w
/
m², 25°
C
Pa
r
a
m
e
te
r
V
a
l
u
e
Po
we
r
a
t
MP
-
P
m
(
W)
7
5
O
p
e
n
c
ir
c
u
it
voltage
-
V
o
c
(
V
)
21.
73
Short
-
c
i
rc
uit c
u
r
r
e
n
t
I
s
c
(
A
)
4.
7
Vol
t
a
g
e
at
M
P
P
-
Vm
(
V
)
17.
41
C
u
rre
nt a
t M
pp
-
Im
(
A)
4
.
3
1
The
i
m
pac
t
o
f
i
r
radia
n
c
e
a
n
d
t
e
m
per
a
ture
on
t
h
e
o
u
t
p
u
t
p
ro
duce
d
c
urre
n
t
a
n
d
pow
e
r
i
s
g
i
ve
n
in
F
i
gure
1.
A
n
incr
ease
in
t
e
m
pera
t
u
r
e
i
nv
ol
ves
a
dec
r
ea
se
i
n
o
p
e
n-circ
uit
vo
lta
ge
a
nd
c
o
nse
que
n
t
ly
t
he
ma
xi
m
a
l
p
o
w
e
r.
I
n
contra
ry,
an
i
nc
rea
s
e
i
n
i
rr
adia
nce
i
m
p
lie
s
a
n
i
n
c
r
ease
in
a
s
h
o
r
t
-cir
cu
i
t
c
urr
e
nt
a
nd
con
s
eq
ue
nt
l
y
a
n incre
a
se
i
n
m
a
xima
l
p
o
w
e
r.
(a)
(b)
F
i
g
u
r
e
1
.
(
a
)
Tem
p
era
t
ure
an
d (
b
) Ir
radia
n
c
e
i
nf
lue
n
ce
o
n
P
V pa
n
e
l cha
r
a
c
t
eris
tic
s
F
r
om
t
he
a
bo
v
e
r
em
a
r
ks,
the
pa
ne
l
m
u
s
t
b
e
kep
t
oper
a
t
i
ng
at
i
t
s
m
a
x
im
a
l
p
ow
e
r
e
ve
n
i
f
i
rra
di
a
n
ce
and
tem
p
era
t
u
r
e
chan
ge
.
Th
i
s
i
s
usua
lly
e
n
s
ure
d
by
a
de
vice
c
a
l
l
e
d
M
P
P
T
:
ma
xi
m
a
l
p
o
w
e
r
po
i
n
t
tr
acke
r
.
A
r
itifi
c
ia
l i
n
te
l
l
i
g
e
n
c
e
i
s
use
d
to a
ccom
p
l
i
s
h
t
hi
s task.
3.
NEURAL
N
ET
WORK
M
PPT
N
e
ural
n
e
t
w
o
rks
a
r
e
qua
li
fie
d
a
s
a
un
i
v
e
r
s
a
l
ap
pr
oxim
a
t
o
rs
[
1
5
]
,
they
c
a
n
m
im
i
c
e
ve
r
y
c
omple
x
fu
nc
ti
o
n
.
The
y
a
r
e
u
se
d
to
m
ode
l
u
n
k
n
o
w
n
s
ys
tem
s
b
e
h
a
v
i
our
u
s
i
n
g
a
s
e
t
o
f
i
npu
t
-
out
put
d
a
t
a
and
a
t
r
ai
nin
g
proce
s
s.
N
eura
l
ne
t
w
or
ks
a
re
f
orma
li
z
e
d
us
ing
a
c
a
sca
d
ed
l
ayer
s
i
nterc
o
n
n
ex
i
on.
E
ac
h
l
a
ye
r
co
ns
ists
o
f
a
se
t
of
n
e
u
ron
s
.
I
n
cre
a
sin
g
t
he
n
u
m
ber
of
l
a
y
ers
and
ne
ur
ons
l
ea
ds
t
o
bes
t
r
e
p
re
sen
t
at
i
on
o
f
n
o
n
-
l
i
n
ea
ri
t
i
e
s
o
f
t
h
e
system
,
how
eve
r
,
it
e
xh
i
b
i
t
s
c
o
m
p
l
e
x com
p
u
t
at
ion
s
,
and
th
e
r
efor
e,
ha
r
dwa
r
e im
plem
enta
t
i
on
co
ns
train
t
s.
In
t
his
pa
per
t
h
e
y
a
re
u
se
d
t
o
m
imic
t
he
M
P
P
T
funct
i
ona
l
ity.
A
n
e
x
p
er
im
enta
l
da
t
a
bas
e
c
om
pris
in
g
i
r
ra
d
i
an
ce
(
E
s
),
t
em
per
a
t
u
re
(
T)
a
nd
m
a
xim
a
l
ou
t
p
u
t
v
ol
ta
ge
(
V
mp
p
)
a
t
M
PP
h
a
v
e
b
een
c
o
l
l
e
ct
ed
a
nd
p
rep
a
red
for
t
r
ai
ni
ng
an
d
val
i
da
ti
o
n
pr
o
cess.
These
in
p
u
t
s
are
m
easure
d
,
f
ilt
e
r
e
d
a
n
d
nor
ma
lize
d
.
The
n
u
m
b
er
o
f
hid
d
e
n
l
aye
r
s
a
n
d
ne
ur
ons
i
n
s
ide
them
h
a
v
e
be
en
s
u
b
j
e
ct
ed
t
o
a
n
o
p
t
i
m
i
zati
o
n
proce
dure
base
d
o
n
o
b
s
er
vin
g
t
he
m
ea
n
sq
uar
e
e
rror
a
t
t
he
e
n
d
o
f
t
r
ai
n
i
ng
p
ro
c
e
ss.
I
t
h
a
s
b
een
d
e
c
i
d
ed
t
o
ado
p
t
t
he
a
rc
hi
tec
t
ur
e
tha
t
c
omprises
one
h
i
d
den
la
ye
r
w
i
t
h
f
iv
e
(5)
neuro
n
s
(
F
ig
ure
2)
w
hic
h
g
ive
s
sati
sfactory
p
reci
s
i
on.
The
tra
i
ni
n
g
p
r
o
cess
i
s
c
a
rri
ed
o
u
t
u
s
i
ng
L
eve
nbe
rg-
M
ar
quar
d
t
A
lgor
it
hm
[
16]
.
The
data
base
w
a
s
part
it
ione
d
as
f
ol
low
s
:
Tra
i
n
i
n
g
da
t
a :
70%
of
da
tase
t
V
a
l
i
dat
i
on
data
: 15%
o
f da
tas
e
t
Test
in
g da
ta
: 15
%
o
f da
t
a
se
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
nt
J
P
ow
E
le
c
&
D
r
i
S
y
st
V
ol
.
10,
N
o.
4
,
De
c
201
9
:
2
1
3
8
–
2
147
2
140
F
i
gur
e
3
g
i
ves
four
p
l
o
t
s
,
issue
d
a
t
t
h
e
en
d
of
t
he
t
r
a
i
n
i
ng
pr
oc
ess.
T
he
o
ver
a
l
l
p
er
f
o
r
m
ance
i
s
sa
ti
sf
a
c
t
o
ry
, wh
i
c
h
p
rov
e
s t
h
at
t
h
e
n
et
wo
rk
ha
s
b
een
t
rai
n
ed
c
o
rre
ctly a
nd r
e
fl
e
c
t
s w
i
t
h
hi
gh f
i
de
lit
y the
inp
u
t-
o
u
tp
ut
rel
at
io
nshi
p
. T
h
e
resu
l
t
a
nt
n
et
wo
rk
i
s u
s
ed
in
si
mu
l
a
t
i
o
n
a
nd
e
x
per
i
m
e
ntal
b
e
n
c
h
ma
r
k
.
F
i
gur
e
2.
N
eura
l
netw
or
k
pr
o
pos
ed
c
omp
o
s
i
tio
n
F
i
gur
e
3.
N
eura
l
netw
or
k
per
f
o
r
ma
nce
pl
ots
4.
FUZ
Z
Y MPPT-AN
F
IS
APPRO
A
CH
F
u
z
z
y
log
i
c
[
1
7]
i
s
use
d
t
o
d
e
scr
i
be
t
he
o
pe
r
a
t
i
on
o
f
a
pr
o
c
e
s
s
t
h
r
oug
h
a
h
u
m
an
r
ea
sonin
g
l
a
ngu
ag
e
.
I
n
p
u
ts
o
f
t
h
e
pr
oc
ess
a
r
e
f
i
r
s
t
c
o
nver
t
ed
t
o
fuz
z
y
s
pa
ce
u
s
i
ng
a
un
ive
r
se
o
f
di
scourse
a
nd
m
emb
e
rshi
p
fu
nc
t
i
o
n
s
(
f
u
zz
if
ica
tio
n)
.
A
se
t
of
i
f-
t
h
e
n
r
ul
es
a
r
e
i
nfe
r
r
e
d
to
g
i
v
e
a
deci
si
on
ba
sed
on
the
pr
e
s
en
t
inp
u
t
s
.
A
fuz
z
y
o
utp
u
t
is
p
r
o
d
u
ce
d
as
a
c
onse
q
ue
nc
e.
T
hi
s
ou
t
p
u
t
i
s
the
n
c
o
nve
r
t
ed
t
o
r
eal
w
or
ld
v
a
l
ue
t
hr
o
ugh
de
fuzz
ific
a
tio
n
.
I
n
t
h
i
s
pa
pe
r
w
e
h
a
v
e
u
s
ed
a
d
a
pti
v
e
ne
ur
al
f
uzz
y
i
n
f
er
e
n
ce
s
yst
em
s
[
18]
t
o
g
e
ner
a
te
a
f
u
z
zy
s
ys
t
e
m
tha
t
o
pera
t
e
s
a
s
a
n
MPP
t
r
ac
ke
r
usin
g
the
sa
me
t
r
a
inin
g
d
a
ta
s
e
t
fr
om
t
h
e
p
r
e
vi
ou
s
sec
tio
n.
A
NF
IS
(
F
i
gur
e
4)
is
b
ui
l
d
a
r
ound
a
l
a
y
er
e
d
a
r
c
hi
tect
ur
e
,
i
n
w
h
i
c
h
e
a
c
h
l
a
y
er
i
m
p
l
e
me
n
t
s
a
dedica
te
d
fu
nc
ti
on.
L
ayer
1:
E
ach
s
qua
r
e
node
i
n
th
i
s
l
a
y
er
c
om
pu
t
e
s
t
h
e
fu
nc
t
i
o
n
x
O
i
i
A
1
Evaluation Warning : The document was created with Spire.PDF for Python.
Int J P
o
w
Ele
c
&
D
ri S
yst
IS
S
N
:
2088-
86
94
Expe
rim
e
nta
l
i
n
ve
s
t
i
g
at
ion o
f
artif
ic
ia
l
i
n
t
e
ll
ige
n
c
e
a
p
p
l
ie
d
in MPPT tech
n
i
q
u
es
(
S
.
D
e
lla K
r
ac
hai)
2
141
wh
ere
x
i
s
th
e
i
npu
t
to
n
od
e
i
,
and
A
i
i
s
th
e
l
i
ng
ui
sti
c
l
a
b
e
l
a
ssociated
w
ith
t
his
node.
O
1
i
i
s
t
h
e
membership
degre
e
of
A
i.
F
i
gure
4. A
NFIS
a
rchitec
t
ur
e use
d
to pro
d
u
c
e
F
u
zzy
M
P
P
Tra
c
ke
r
Laye
r 2:
Ever
y c
i
r
c
le
node
i
n th
is la
y
er
c
om
pute
s
the
pro
duc
t
of inc
om
i
n
g s
i
gna
l
s
.
y
x
i
i
B
A
i
Th
i
s
r
epr
e
sents
the
firi
n
g
stren
gt
h o
f
the
r
ule
.
Laye
r
3:
E
v
e
r
y
c
i
r
c
le
n
o
d
e
p
r
ocesses
the
rati
o
of
t
he
i
th
r
ule’s
fi
r
i
n
g
st
ren
g
th
t
o
th
e
sum
o
f
a
l
l
r
ul
e
’
s
f
i
ri
ng
st
r
e
ngths
.
2
1
1
i
Laye
r 4:
Ea
c
h squ
a
r
e
node
i
n th
is la
y
er
per
fo
rm
s the
fu
nc
ti
o
n
.
i
i
i
i
i
i
r
y
q
x
p
f
O
i
4
wh
ere {
p
i
,
q
i
,
r
i
} is the
p
ara
m
e
t
e
r
s
e
t
re
f
er
red as c
onse
que
n
t
par
am
eter
s.
Laye
r 5:
The
u
n
i
q
u
e
n
o
d
e
in
this
l
a
yer
com
p
u
t
es t
he
o
vera
l
l
s
um
m
a
tion
o
f
al
l
in
co
mi
ng
s
ig
n
a
l
s
i
i
i
i
i
i
i
i
f
f
O
i
5
The
aver
age
pe
rc
enta
ge
e
rr
or i
s use
d
as
a c
r
iterio
n
to
eva
l
ua
te
b
ot
h
t
e
ch
niqu
e
s
:
%
100
*
)
(
)
(
)
(
'
1
1
p
i
i
T
i
O
i
T
p
APE
F
i
r
s
t
of
a
ll,
a
f
uzz
y
i
n
f
er
ence
s
ystem
wit
h
d
efau
l
t
m
e
m
be
rsh
i
p
fu
nct
i
ons
a
nd
ru
le
s
e
t
i
s
gener
a
te
d.
F
i
gure
5
(t
op)
d
e
p
i
c
t
s
t
h
e
d
e
f
aul
t
m
em
ber
s
h
i
p
fu
nc
ti
o
n
s
fo
r
irr
ad
i
a
n
c
e
and
t
e
mp
erat
u
r
e.
I
t
i
s
o
b
s
erv
e
d
t
h
a
t
t
h
e
s
e fu
nct
i
o
n
s
are uni
form
l
y
di
s
t
r
ib
ute
d
o
ve
r
t
h
e
data
ra
n
g
e
(
unive
rse
of di
s
co
urse
) fo
r
e
ach
i
np
ut.
A
f
t
e
r
tra
i
n
i
n
g
t
he
n
e
t
w
o
r
k
,
p
r
e
m
i
s
e
and
c
onse
que
n
t
p
ara
m
e
t
e
r
s
a
r
e
a
djus
ted
a
c
c
o
rd
in
g
t
o
t
he
b
ac
k-
pro
p
aga
t
io
n
e
rror
.
F
igure
5
(
b
o
ttom
)
s
h
o
w
s
t
ha
t
m
e
m
b
ershi
p
f
u
n
c
t
i
o
n
s
for
t
e
mpe
r
at
ure
ha
ve
no
s
i
g
n
i
f
ic
a
n
t
chan
ge
s,
how
e
v
er,
th
ose
of
i
rradia
n
ce
c
h
a
n
ge
d
co
ns
i
d
era
b
ly,
w
h
i
ch
c
o
n
firm
t
he
h
igh
de
pen
d
e
n
c
y
o
f
the
ma
xi
m
a
l
p
o
w
e
r
on the
i
r
radia
n
c
e
.
F
i
gur
e
6
(
l
eft)
d
em
on
strate
s
t
h
e
per
f
orm
a
nc
e
of
t
r
ack
in
g
o
p
era
t
io
n
based
on
a
n
fis.
T
he
r
igh
t
s
urfa
ce
plo
t
i
n
F
i
gure
6
is
a
s
urface
i
n
d
i
ca
ti
n
g
t
ha
t
the
r
e
lat
i
onsh
i
p
betw
ee
n
te
m
p
e
r
ature
,
i
rra
dia
n
c
e
a
n
d
m
axima
l
vo
ltage
i
s no
n-li
nea
r
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2088-8694
Int
J
Pow
E
lec
& D
r
i
Syst Vol.
10,
N
o.
4
, D
ec
201
9 :
2
1
38
– 2
147
2
142
F
i
gure
5.
A
NF
IS
m
embe
rship
func
tio
ns be
f
o
r
e (top)
a
n
d
a
ft
er (bo
t
t
om
)
t
r
aini
ng
F
i
gure
6.
A
NF
IS
l
ear
ni
n
g
p
ro
ce
ss
r
esult
s
In
o
rde
r
t
o
c
onne
c
t
t
he
P
V
pane
ls
t
o
the
l
o
a
d
,
a
p
o
w
e
r
c
o
n
v
erte
r
is
u
s
e
d
as
a
m
ean
to
t
ransfer
a
regu
late
d
p
o
w
e
r
from
one side
t
o
t
he o
t
h
er
.
5.
SYN
C
H
R
ON
OUS
BUCK
CONVERT
ER M
O
D
ELING
S
ynchr
on
o
u
s
b
u
ck
c
o
nve
rter
i
s
a
s
t
e
p
-d
ow
n
c
o
nv
e
r
ter
w
h
e
r
e
the
d
io
de
i
s
re
pl
a
c
e
d
w
it
h
a
sw
it
c
h
t
o
minim
i
z
e
t
he
l
osse
s
d
u
e
t
o
t
he
c
om
muta
ti
o
n
i
n
the
o
f
f-
tim
e
per
i
o
d
o
f
t
h
e
m
a
i
n
s
w
i
tc
h,
t
hu
s
impro
v
i
n
g
t
he
e
f
fi
ci
en
cy
o
f
th
e
co
nv
ert
e
r.
T
h
e
c
i
r
cui
t
i
s
g
i
v
e
n
b
y
Fi
gu
re
7
,
wher
e
t
h
e
swit
c
h
m
os
fe
t/
2
re
p
l
a
ces
t
he
d
i
o
de
i
n
t
h
e
c
o
nve
n
t
i
o
n
a
l
buc
k c
o
nfi
gur
at
ion.
The
e
l
e
m
e
n
ts
o
f
t
h
e
co
n
v
er
te
r
a
r
e
sized
c
o
r
r
ectly
t
o
me
et
t
he
f
ol
low
i
ng
r
equ
i
rem
e
nts
:
a
s
w
itc
h
i
n
g
fre
que
nc
y
o
f
2
00
k
h
z,
a
nd
a
m
a
xim
a
l
o
u
t
p
u
t
c
urre
nt
o
f
1
0
A
.
A
P
W
M
d
r
i
ver
is
u
se
d
to
d
ri
ve
t
he
m
osf
e
ts
from
t
h
e si
gn
al
g
en
erat
ed
by
a co
n
t
ro
ll
e
r
.
Evaluation Warning : The document was created with Spire.PDF for Python.
Int J P
o
w
Ele
c
&
D
ri S
yst
IS
S
N
:
2088-
86
94
Expe
rim
e
nta
l
i
n
ve
s
t
i
g
at
ion o
f
artif
ic
ia
l
i
n
t
e
ll
ige
n
c
e
a
p
p
l
ie
d
in MPPT tech
n
i
q
u
es
(
S
.
D
e
lla K
r
ac
hai)
2
143
We
h
a
v
e
use
d
a
P
ID
c
ontr
o
l
l
e
r
t
o
ge
nera
t
e
t
he
d
ut
y
c
y
c
l
e
D
nee
d
e
d
t
o
tra
c
k
t
h
e
m
a
x
i
m
u
m
vo
l
t
ag
e
V
m
pp.
T
uni
ng
t
h
i
s
P
I
D
c
o
n
t
r
o
l
l
er
i
s
ea
sy
w
hen
t
h
e
c
o
m
pone
n
t
s
va
l
u
e
s
L
a
n
d
C
a
r
e
p
r
e
c
i
s
e
l
y
k
n
o
w
n
.
H
o
w
e
ve
r,
m
an
ufa
c
t
urer
s
prop
ose
t
o
l
e
ranc
es
on
t
h
eir
com
p
o
n
en
ts
(
t
h
e
i
nduc
t
o
r
use
d
W
E-7
443
631
00
0
i
s
fro
m
Wür
t
h
E
l
ek
tro
n
i
k
w
ith
a
t
o
l
e
r
a
n
ce
of
±
1
5
%
)
.
T
h
e
P
I
D
i
s
t
une
d
t
o
t
a
k
e
in
t
o
a
cc
ou
n
t
c
o
m
pone
n
t
s
t
o
lera
nce
o
f
t
h
e
b
u
c
k
c
on
ve
r
t
e
r
t
o
ob
t
a
i
n
h
igh
effic
i
e
n
cy
w
ithi
n
t
he
r
ang
e
o
f
e
x
p
ect
ed
val
u
e
s f
o
r L an
d C
.
F
i
g
u
r
e
7
.
S
y
n
c
h
r
on
ou
s bu
c
k
co
n
v
e
rt
er t
opol
og
y
6.
S
Y
ST
E
M
S
I
M
UL
A
T
I
O
N
Befor
e
d
ec
i
d
in
g
to
i
m
p
lem
e
nt
t
he
c
on
t
r
ol
l
e
r
a
nd
t
h
e
trac
kers,
w
e
h
a
ve
c
on
du
ct
e
d
a
s
i
m
ul
a
tio
n
con
s
i
d
eri
ng c
h
ange
s in
t
e
m
pe
rature
a
n
d
i
rra
di
a
n
ce.
Ne
ural
n
etw
o
rks
are
f
i
r
s
t
sim
u
la
te
d.
F
rom
Fig
u
re
8
,
t
h
e
syste
m
is
l
ef
t
i
n
i
dle
mo
de
u
n
t
i
l
0
.
05
sec
o
n
d
pas
t
.
The
trac
kin
g
opera
t
i
o
n
is
e
n
g
a
g
e
d
f
o
r
a
n
irra
dia
n
c
e
of
5
0
0W/m²
and
tem
p
era
t
u
r
e
of
2
5°C,
t
h
e
p
a
n
e
l
v
o
l
t
a
g
e
(
pl
ot1
–
re
d
cu
rv
e
)
f
oll
o
ws
p
re
ci
sel
y
t
h
e
m
a
x
i
m
a
l
v
olt
a
ge
c
o
m
pute
d
b
y
the
ne
ural
n
e
t
w
o
r
k
.
A
n
i
r
radia
n
c
e
d
i
s
t
u
rba
n
c
e
(
an
i
nc
rea
s
e
from
5
00
W/m
²
t
o
10
00
W
/
m
²
)
i
s
inj
e
c
t
ed
a
t
th
e
ti
me
0
.15
s
,
th
e
PID
con
t
ro
l
l
e
r
a
d
j
u
s
ts
t
he
d
ut
y
c
y
c
l
e
acc
ordi
ng
ly
t
o
kee
p
t
ra
cki
n
g
t
h
e
m
a
xim
a
l
vo
lta
ge.
A
t
t
ime
0.2
2
s,
a
t
e
m
p
e
r
at
ure
va
riat
io
n
is
i
ntro
duce
d
(
a
n
i
ncr
ease
fr
om
2
5°
C
to
4
0°
C).
A
sm
all
d
e
cr
eas
e
i
n
p
a
n
e
l
v
o
l
t
a
ge
i
s
obse
r
ved
w
i
t
h
no e
f
fec
t
o
n t
h
e
outp
u
t
curre
n
t
.
The
fuzzy
inference
s
y
s
t
em
i
ss
ue
d
from
anfis
training
(F
i
gur
e
9
)
t
r
ac
ks
w
i
t
h
h
i
gh
p
re
cisio
n
t
h
e
ma
xi
m
a
l
p
o
w
e
r
gener
a
ted
fro
m
P
V pa
nel re
gard
less
en
viro
nme
n
ta
l
c
h
a
nge
s.
F
i
gure
8.
N
eura
l netw
orks
m
a
x
i
m
al
p
o
w
er
t
r
acki
n
g
for irra
dia
n
ce
and
tem
p
era
t
ur
e
va
riat
ion
F
i
gure
9.
F
u
z
zy
l
o
g
i
c
ma
xima
l
pow
er
t
rac
k
ing
for
Irradiance an
d
t
em
pe
ratur
e
variation
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
nt
J
P
ow
E
le
c
&
D
r
i
S
y
st
V
ol
.
10,
N
o.
4
,
De
c
201
9
:
2
1
3
8
–
2
147
2
144
7.
EX
PERI
ME
N
T
AL S
ETUP
It
i
s
co
mp
o
s
ed
o
f
t
w
o
ma
jo
r
c
i
rcui
t
s
.
A
me
a
s
u
r
emen
t
c
i
rc
ui
t
(Fi
g
u
r
e
1
1)
a
nd
a
p
o
w
e
r
c
i
r
c
ui
t
(
F
ig
ur
e
10(
a
)
)
.
T
he
p
o
w
e
r
c
i
r
cui
t
c
o
m
pr
ises
t
he
s
y
n
chr
o
n
o
u
s
bu
c
k
c
o
nver
ter
(
t
wo
m
o
s
f
e
ts),
t
h
e
p
wm
d
riv
e
r
IR210
4
an
d
a
l
so
inp
ut/
o
utp
u
t
c
u
r
e
nt
s/v
o
lt
a
g
es
m
ea
su
re
me
n
t
s.
The
me
asur
e
m
e
n
t
cir
c
u
it
i
s
b
uil
t
e
sse
n
t
i
a
lly
a
r
o
u
nd
t
w
o
s
e
ns
or
s
a
t
t
a
c
he
d
t
o
t
he
P
V
pane
l
.
S
P
-
110
i
rrad
i
an
c
e
s
e
n
so
r
fro
m
ap
og
ee
i
nst
r
u
m
e
n
t
s
,
i
s
u
sed
to
cap
tu
re
s
ol
ar
i
r
r
a
di
a
n
ce
w
i
t
h
a
n
ou
t
p
ut
o
f
0.
2mV
/
(
W
/
m
2
)
,
w
hich
i
s
a
m
pli
f
ied
b
y
a
n
ins
t
r
u
me
n
t
a
t
i
o
n
am
pl
if
ier
A
D
620.
T
h
e
se
co
nd
s
ens
o
r
is
t
he
tem
p
er
at
ur
e
se
nsor
L
M
3
5
w
h
ich
pr
od
uce
s
1
0m
V
/
°
C
.
A
r
du
in
o
ba
se
d
m
i
c
r
o
co
n
t
r
o
ller
is
u
se
d
as
a
n
em
b
e
dde
d
c
o
mpu
t
ing
p
l
a
t
for
m
.
The
fina
l
P
C
B
used
i
n
expe
r
i
me
nt
a
tio
n
i
s
g
i
ve
n
i
n
F
i
g
ur
e
10(
b)
.
(a)
(b
)
F
i
gur
e
1
0
.
S
y
n
c
hr
o
n
o
u
s
b
u
ck
c
on
ver
t
e
r
p
ow
er
e
lec
t
r
onic
s
(a)
(b
)
F
i
gure
1
1
.
(
a
) Irr
a
dianc
e
a
n
d
t
em
pera
ture
m
easure
m
e
n
t
s
c
ir
cui
t
(b
)
A
r
d
u
i
n
o
i
n
t
e
rf
a
ce
8.
HARDWARE-IN-THE
-
LOOP
Th
e
so
ft
wa
re
i
s
i
m
pl
e
m
e
n
t
e
d
in
S
i
m
ul
in
k
usin
g
a
r
d
u
i
n
o
pl
a
t
fo
rm
a
s
a
t
a
r
g
e
t
mi
c
r
oc
ontro
ll
er.
Th
e
in
f
o
r
m
atio
n
ca
pt
ur
e
d
f
r
o
m
se
nsor
s
are
first
co
nd
itio
n
e
d
(amp
lif
i
e
d
an
d
fi
lt
ere
d
)
t
o
p
rod
u
ce
a
hi
gh
-fid
e
lit
y
im
age
o
f
t
he
m
ea
sures.
T
hi
s
is
t
he
r
o
l
e
of
t
h
e
s
i
g
n
a
l
co
n
d
i
tio
n
in
g
b
l
ock
i
n
F
i
gur
e
1
2
.
The
tr
ac
ker
(
n
e
u
r
a
l
ne
tw
or
ks
/f
uzz
y
s
ys
te
m
)
c
om
putes
a
t
i
t
s
o
u
t
p
ut
t
h
e
r
e
f
er
ence
v
o
lta
ge
(
V
mp
p
)
a
t
w
hic
h
t
he
p
r
o
duce
d
pow
e
r
i
s
m
a
xi
m
a
l.
Th
i
s
r
e
f
er
e
n
ce
i
s
t
h
en
c
om
pa
r
e
d
to
t
he
m
ea
sur
e
d
v
o
lta
ge
a
t
P
V
p
ane
l
t
e
r
m
i
nals.
A
P
I
D
contr
o
l
l
er
d
e
si
gn
ed
a
nd
t
u
n
e
d
t
o
el
i
m
ina
t
e
t
h
e
e
r
r
o
r
be
t
w
ee
n
ref
e
r
e
nc
e
a
n
d
m
ea
su
r
e
d
vo
l
t
age
s
,
o
u
tp
ut
s
a
c
o
mm
and
si
gna
l
w
h
ic
h
repr
esen
ts
t
he
v
aria
t
i
o
n
o
n
the
d
u
t
y
c
yc
le.
A
PWM
gen
e
r
a
tor
r
e
c
e
i
v
es
t
he
d
uty
cyc
l
e
a
n
d
pr
o
duc
es
p
u
l
se
s
w
h
ic
h
ar
e
f
e
d
t
o
t
he
I
R21
0
4
c
i
r
cu
it
dr
iver
t
o
s
witch
th
e mo
s
f
et tran
s
i
s
to
rs.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
P
o
w
Elec
&
D
r
i
S
y
st
I
SS
N
:
2088-
86
94
Exp
e
ri
me
nt
al
in
v
e
s
t
i
ga
t
i
o
n
of
a
r
ti
fi
c
i
al
i
n
t
e
lli
g
e
n
c
e ap
pli
e
d
i
n
MPPT
t
e
c
hni
qu
e
s
(S
.
Del
l
a
K
r
a
c
h
a
i
)
214
5
F
i
gur
e
1
2
.
H
a
rdw
a
r
e
i
n
the
L
o
o
p
i
mp
lem
e
n
t
a
t
i
o
n
of
t
he
s
ys
tem
9.
EX
PERI
ME
N
T
AL RES
ULTS
The
P
I
D
out
p
u
t
s
a
du
t
y
c
yc
le
v
ar
i
a
ti
on
c
o
r
r
esp
o
n
d
i
ng
t
o
t
r
a
c
k
in
g
t
h
e
m
a
xima
l
vo
l
t
a
g
e
ge
ner
a
t
e
d
by
PV
p
an
el
i
n
en
vi
ron
m
e
n
t
a
l
co
ndi
tio
ns
g
iv
en
b
y
Figu
re
1
3
,
w
h
i
ch
re
p
r
esent
s
t
h
e
r
eal
-ti
m
e
me
as
u
r
e
m
e
n
ts
o
f
ir
r
a
dia
n
ce
a
n
d
t
em
per
a
tur
e
f
r
o
m
the
se
ns
or
s
for
the
day
0
5
/
19
/
2
01
8
in
O
r
a
n
-
A
l
g
e
r
ia.
F
i
gur
e
1
4
c
on
f
i
r
m
s
t
h
e
tr
ac
k
i
n
g
p
er
f
o
r
m
anc
e
.
I
t
i
s
sh
ow
n
tha
t
t
he
P
V
pane
l
v
o
lta
ge
(
bl
ue
c
ur
ve)
tra
c
ks pre
cise
l
y
Vm
pp (
y
ell
o
w
cur
v
e)
i
n
the
pr
esenc
e
of fl
uct
u
a
t
i
o
n
s
i
n
i
r
r
adia
nce
a
nd
te
m
p
er
a
t
ur
e.
F
i
gur
e
1
5
s
h
o
w
s
that
t
he
p
a
n
el
h
ave
be
en
e
xp
ose
d
t
o
a
v
a
r
i
a
t
i
o
n
o
f
ir
r
a
dia
n
ce
d
ue
t
o
t
h
e
pa
ssa
g
e
o
f
c
l
o
u
d
s
,
w
h
i
l
e
t
h
e
t
e
m
p
e
r
a
t
u
r
e
o
s
c
i
l
l
a
t
e
s
m
o
o
t
h
l
y
a
r
o
u
n
d
3
0
°
C
.
F
or
t
he
se
c
on
d
iti
on
s
the
n
e
ur
a
l
n
e
t
w
o
r
k
M
P
P
c
ontr
o
l
l
er
s
ucc
eeds
to
e
xtr
a
c
t
t
he
m
a
x
i
m
a
l
vol
ta
ge
f
r
o
m
t
h
e
P
V
pa
nel
(
F
igur
e
16)
.
F
i
gur
e
1
3
.
Rea
l
ti
m
e
i
r
r
a
dia
n
c
e
(
b
l
u
e
)
a
nd
tem
p
er
at
ure(
re
d)
m
easure
m
en
ts
Fi
g
u
r
e
1
4
.
A
N
F
I
S
M
PP
t
ra
c
k
er
r
e
s
u
l
t
s
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
nt
J
P
ow
E
le
c
&
D
r
i
S
y
st
V
ol
.
10,
N
o.
4
,
De
c
201
9
:
2
1
3
8
–
2
147
2
146
Figure
1
5
.
R
e
al-t
ime
Ir
radian
c
e
a
nd tem
p
er
ature
mea
s
u
r
emen
t
s
F
i
g
u
r
e
16
.
Tr
a
c
k
i
ng
th
e
M
PP u
sin
g
n
e
u
r
al newo
r
k
s
Ob
t
a
i
n
ed
r
esu
l
t
s
s
ho
w
th
at
a
n
ef
fi
c
i
en
cy
o
f
a
b
out
9
8
%
i
s
e
n
su
re
d
usi
n
g
t
h
e
pr
op
ose
d
t
e
c
hn
i
que
s,
a
s
w
e
l
l
a
s,
t
he
t
r
a
cki
n
g
o
p
er
a
tio
n
is
s
moo
t
h
a
n
d
e
n
sur
e
d.
F
r
o
m
exp
e
r
ime
n
ta
l
p
o
in
t
of
v
ie
w
,
t
he
r
esu
lts
a
r
e
a
lso
ve
r
i
f
i
e
d
.
Mor
e
over
,
t
he
p
r
o
p
o
se
d
ap
pr
oac
h
i
s
base
d
o
n
a
l
ow
-
c
os
t
ha
rd
war
e
(
8
b
i
t
m
ic
roc
o
nt
r
o
ller
)
w
hic
h
i
s
pr
o
g
r
a
m
m
e
d
i
n
a
gr
aph
i
ca
l
b
l
oc
k
en
v
i
r
o
nm
ent
instea
d
o
f
h
ar
d
c
o
d
i
n
g.
T
his
gi
ves
the
o
p
p
o
r
t
u
n
i
t
y
t
o
t
une
t
he
pa
r
a
m
t
er
s
in
line
(
w
hen
r
u
n
n
in
g
t
h
e
sy
stem
)
in
o
r
d
er
t
o
a
c
hie
v
e
be
st
p
e
r
for
m
a
n
ces.
The
o
p
t
i
ma
l
o
b
t
aine
d
sys
t
em
,
i
s
t
h
e
n
dow
n
l
oa
de
d
to
t
he
m
icr
o
c
o
n
t
r
o
l
l
e
r
t
o
o
p
e
r
a
t
e
in
a
s
ta
nda
l
o
ne
m
ode
.
10.
CONCLUSION
In
t
h
i
s
pa
p
e
r,
a
rtific
ia
l
i
n
t
e
lli
g
e
nc
e
is
s
uc
ce
ss
fu
l
l
y
a
p
p
l
ied
i
n
t
r
ack
i
ng
the
ma
xima
l
p
o
w
e
r
gene
r
a
ted
by
P
V
pa
nel.
N
eur
a
l
netw
or
ks
a
n
d
a
d
a
p
t
iv
e
f
u
zz
y
i
n
f
e
r
e
n
ce
sys
t
e
m
s
,
g
i
ve
p
r
o
m
i
s
i
ng
r
esu
l
ts
a
nd
s
a
t
is
fa
ct
i
o
n
an
d
can
b
e
both
ap
plied
in
t
r
a
ck
ing
th
e
MPP u
sin
g
a
lo
w-cost h
a
rd
war
e
.
S
i
m
u
l
a
ti
o
n
r
e
s
ults
s
h
o
w
t
h
e
app
l
ica
t
i
o
n
o
f
s
uc
h
te
c
h
n
i
que
s
ex
hi
bi
ts
f
a
s
t
co
nv
e
r
g
e
n
c
e
r
e
g
a
r
d
l
e
ss
t
he
v
a
ri
at
i
o
n
of
e
nv
i
r
o
n
m
en
t
a
l
c
o
nd
i
tio
ns
.
On
t
h
e
o
t
h
er
h
ands,
ex
p
e
r
i
m
e
n
t
a
l
r
esu
lts
p
r
o
ve
t
he
q
ual
i
t
y
of
s
uch
tec
h
n
i
q
u
e
s
i
n t
h
e
effic
i
e
n
c
y
i
m
p
rovem
e
n
t
o
f
the
who
l
e sys
t
em
.
REFERE
NCES
[1]
Tuff
ahaT.H.,
M
.
B
ab
ar,
Y
.
K
han
an
d
N.H.
M
alik,
"Com
p
a
rativ
e
S
t
u
d
y
o
f
Diff
eren
t
Hil
l
C
li
m
b
i
n
g
M
P
PT
t
h
r
ough
S
i
m
u
l
a
t
i
o
n
a
n
d
E
x
p
e
r
imental
Test
B
ed,"
Res
e
ar
ch Jou
r
na
l of
App
l
i
e
d Sci
e
nces,
E
ngi
neer
in
g
an
d T
echno
lo
gy
,
vo
l.
7
(20
)
,
p
p
.
4
2
58-4
2
6
3
,
201
4.
[2]
S
h
ei
k,
S
.;
D
evaraj
,
D
.
;
Imt
h
i
a
s,
T
.,
"
A
no
vel
h
y
b
r
id
M
axim
u
m
P
o
w
e
r
P
o
in
t
Tra
c
k
i
n
g
T
ech
ni
que
u
si
ng
P
e
rt
urb&
Ob
serve
algo
rit
h
m
an
d L
earni
ng
Au
to
m
a
t
a
f
or
s
olar
P
V
system
,
"
En
e
r
g
y
, vo
l
.
11
2
, p
p.
10
9
6
–
1
1
0
6
, 2
01
6.
[3]
Luig
i
P
.
,
Renat
o
R
.,
Iv
an
S
.an
d
d
Pietro
T
.,
"
O
p
t
i
m
ized
A
d
a
pti
v
e
P
ert
u
rb
a
nd
O
b
s
erv
e
M
axi
m
um
P
o
w
er
P
o
i
n
t
Track
ing
Control f
o
r P
hoto
v
o
l
t
a
ic G
enerati
o
n
,
"
En
ergi
es
,
v
o
l
.
8
,
pp.
3
4
18-3
4
3
6
,
2
0
15.
[4]
Ju
baer
A
hm
ed,
Zai
n
a
l
S
a
l
a
m,
"
An
i
mp
rov
e
d
p
e
rtu
r
b
a
n
d
ob
serv
e
(
P
&
O)
m
ax
imum
pow
er
p
oin
t
t
rack
ing
(M
P
P
T)
alg
o
rithm
fo
r
high
er
e
f
f
ici
e
ncy
,
"
App
l
i
e
d
Energ
y
,
vo
l
.
1
50
,
p
p
.
97
-
10
8, 2
01
5.
[5]
Han
a
neYat
i
mi
,
E
l
h
a
ssan
A
rou
d
am
,
"
M
P
P
T
alg
o
ri
thms
b
ased
m
odeli
ng
a
n
d
c
ont
rol
f
o
r
p
hot
ov
o
ltai
c
s
y
s
t
e
m
u
nder
vari
able
c
li
matic
conditions
,
"
P
rocedi
a
Manuf
a
c
t
u
ri
ng,
vol
.
22,
p
p
.
757
-76
4
,
201
8
.
[6]
Chen
di
L
.
,
Yu
anru
i
C.,
D
ong
b
a
o
Z.
,
J
unf
eng
L
.
a
nd
J
un
Z
.,
"
A
H
i
g
h
-P
erform
an
ce
Adap
tive
Increm
en
tal
Con
duct
a
nce M
P
P
T
A
l
g
o
r
ithm
forP
h
o
t
o
v
o
lt
a
i
c Sy
st
ems," E
n
ergi
e, v
o
l
.
9
, p
. 2
88
,
2
0
1
6
.
[7]
Nab
i
po
ur,
M
.
;
Raza
z,
M
.;
S
eifos
s
adat
,
S
.
;
M
o
rt
azavi
,
S
.
,
"
A
n
e
w
M
P
P
T
scheme
b
a
s
ed
o
n
a
n
o
v
e
l
f
u
zzy
app
r
oach,
"
R
en
e
w
.
S
u
s
t
a
in.
E
n
ergy
R
ev,
vol.
74,
pp
.
1
147
–1
16
9,
20
17.
[8]
De
lla
Kra
c
h
a
i
M
.,
Mido
un
A
.,
"
High
E
ffic
ie
nc
y
Ma
x
i
m
u
m
P
o
we
r
Poi
nt
T
racking
Contro
l
inPotovolt
a
ic-Gri
d
Con
n
ected
P
lan
t
," A
cta Elect
ro
t
ech
ni
ca et Inf
ormat
i
ca, v
o
l
. 7,
no
. 1
,
2
007
.
[9]
Ravi
nder
Ku
mar,
H
ari
O
m
B
ans
a
l,
"
Real
‐tim
e
imp
l
em
ent
a
ti
on
o
f
ad
ap
ti
ve
P
V
‐
integ
r
a
t
ed
S
AP
F
t
o
e
nh
anc
e
p
ow
er
quality,
"
I
nternat
i
ona
l
T
ra
nsact
ions On E
l
ectrical
E
nergy Syste
ms
,
20
19
.
[10]
P
r
adeep
V
.
,
D
.,
H
im
a,
B
ind
u
A
.
Div
y
a,
"
A
n
t
Co
lo
ny
Opti
mizat
io
n
based
M
axi
m
um
P
o
w
er
P
oi
n
t
T
rackin
g
(MPP
T)
f
o
r
P
art
i
allySh
aded
Stan
d
a
l
o
n
e
P
V
Sys
t
em,
"
I J C
T A, v
ol
.
9
(1
6),
p
p
.
8125
-813
3,
2
0
1
6
.
[11]
Yi
-HuaLiu,
C
h
u
n
-
Li
ang
L
iu,
Jia-W
e
iH
uan
g
,
Jing
-Hs
i
auCh
ena,
"
Neu
r
a
l
-
net
w
o
r
k
-
bas
e
d
m
a
ximu
m
p
o
w
e
r
p
o
int
track
ing
m
e
th
ods
f
o
r
p
h
o
to
vo
lt
aic
s
y
s
t
ems
operat
i
n
g
u
nd
e
r
f
ast
c
h
a
n
ging
e
nv
ir
on
me
nts,"
So
la
r
En
e
r
gy
,
vo
l.
8
9
,
pp
.
4
2-53
,
2
0
1
3
.
[12]
Rihab
M.
E
.,
M
a
nsou
r
S
.
,
Hsan
H
.
A.,
"
M
axi
m
u
m
P
o
w
er
P
o
i
n
t
T
rack
i
ng
Con
t
ro
l
Usi
n
g
N
e
ural
N
et
wo
rks
f
o
r
S
t
an
d-Al
on
e
P
h
o
t
ov
oltai
c
S
yste
m
s
,
"
I
n
t
ern
a
tio
n
a
l
Jo
urn
a
l
of
M
o
d
ern
No
nli
n
ear
T
heo
r
y
and
Ap
pl
icati
o
n
,
vol.
3,
pp
.
5
3-65
,
2
0
1
4
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t
J
P
o
w
Elec
&
D
r
i
S
y
st
I
SS
N
:
2088-
86
94
Exp
e
ri
me
nt
al
in
v
e
s
t
i
ga
t
i
o
n
of
a
r
ti
fi
c
i
al
i
n
t
e
lli
g
e
n
c
e ap
pli
e
d
i
n
MPPT
t
e
c
hni
qu
e
s
(S
.
Del
l
a
K
r
a
c
h
a
i
)
214
7
[13]
R
.
K
u
m
a
r
,
P
.
C
h
a
t
u
r
v
e
d
i
,
H
.
O
.
B
a
n
s
a
l
a
n
d
P
.
K
.
A
j
m
e
r
a
,
"
A
d
a
p
t
i
ve
a
rtificia
l
neural
n
et
work
-based
c
ont
r
ol
stra
te
g
y
f
or
s
hu
nt
a
c
t
ive
p
o
we
r
fi
l
t
er
,
"
2016
Internati
o
nal
Con
f
e
rence
o
n
E
lect
rical
P
ow
er
a
nd
E
n
e
rgy
S
y
s
t
e
m
s
(ICEPES
),
B
ho
p
a
l,
pp.
1
94-1
99,
2
01
6.
[14]
Ravi
nder
Ku
ma
r,
H
ari
O
m
B
ans
a
l,
"
S
h
unt
activ
e
p
o
wer
fil
t
er:
Cur
ren
t
s
ta
t
u
s
of
c
ontrol
t
echn
i
qu
es
a
nd
i
t
s
in
tegratio
n t
o
r
enewab
le en
e
rgy
so
urces,
"
S
u
s
t
a
in
able
C
iti
e
s
an
d
S
o
ciet
y,
vol.
42,
pp
.
5
74
-5
92
,
2
0
1
8
.
[15]
Ho
rnik
K
.,
Stinchco
m
b
e
M.
,
and
W
h
ite
H
.,
"
M
u
lti
l
a
yer
Fee
d
f
o
rwar
d
N
e
tw
orks
a
re
U
n
i
vers
al
A
p
p
rox
i
m
a
t
o
rs,
"
Neural Network
s, vo
l
. 2
,
p
p
. 35
9
-36
6
, 19
8
9
.
[16]
Marquardt,
D
.
,
"
An
A
lgor
it
h
m
f
or
l
eas
t-s
quares
est
i
m
a
tion
o
f
n
o
n
l
i
n
ear
p
ara
m
et
ers,"
S
I
AM
J
ou
rn
al
o
n
Ap
pl
ied
Math
ematics,
vo
l
. 1
1,
no
.
2
,
pp
.
4
3
1
-4
41
,
1
9
6
3
.
[17]
Chi
a
n-S
o
n
g
C
hiu,
"
T
-
S
F
u
zzy
M
ax
im
u
m
P
o
w
er
P
oi
n
t
T
racking
Con
t
ro
l
of
S
olar
P
ower
G
en
eratio
n
S
y
s
t
e
m
s,
"
IEEE
T
r
ans Ene
r
gy
Conversi
on,
v
ol
. 25,
no.
4,
pp
1123-1132, 201
0.
[18]
Jy
h-S
h
i
n
g
Ro
ge
r
J
a
ng
,
"AN
F
IS
:
Adap
ti
ve-Neural-Bas
e
d
F
u
zzy
I
nf
er
ence
S
yst
e
m,"
IEEE
T
rans
acti
o
n
s
o
n
S
y
s
t
e
m
s,
M
an
,
a
nd
cybern
eti
c
s,
vol
.
2
3
,
no
.
3,
1
9
9
3
.
BIOGRAPHI
E
S
OF A
U
T
HORS
P
hd
st
uden
t
,
Un
iv
ers
i
t
y
of
S
c
ien
ce and
Tech
no
lo
g
i
es
-M
oh
amed
Bou
d
i
af,
Ora
n
, Al
g
eria,
Elect
rical
En
g
i
n
eerin
g
Facu
lt
y
, Depart
m
ent of Elect
roni
cs,
e
-
m
a
il
:
s
a
id
ia
.
d
e
l
l
a
kr
a
c
h
a
i
@u
n
i
v
-
us
to
.
d
z
.
Research D
irect
or, U
n
i
versity of S
c
ien
c
e
and
Techn
o
l
o
g
i
es
-Mo
h
am
ed Bo
u
d
i
a
f
, Or
a
n, A
l
g
eri
a
,
Elect
rical
En
g
i
n
eerin
g
Facu
lt
y
, Depart
m
ent of Elect
roni
cs,
e
-
m
a
il
:
A
b
o
u
d
g
h
e
ne
s
@
gm
a
i
l
.
c
o
m
Dr, U
n
i
v
ersity of
S
c
ien
c
e
and
Te
ch
nol
og
ies
-
Mo
h
a
m
e
d
Bou
d
iaf,
Ora
n
, Algeri
a
, Electri
cal
Eng
i
n
eerin
g
F
acu
lty, D
e
part
men
t
o
f E
l
ectro
n
i
cs,
e-mai
l
:
m
o
h
a
m
e
d.d
e
ll
akrachai
@un
i
v
-
usto.
d
z
Dr,
Spat
i
a
l dev
e
lo
p
e
ment center,
e-mai
l
:
M
_
b
e
kh
ti
@yah
oo.
f
r
Evaluation Warning : The document was created with Spire.PDF for Python.