Int
ern
at
i
onal
Journ
al of
P
ower E
le
ctr
on
i
cs a
n
d
Drive
S
ystem
s
(
IJ
PEDS
)
Vo
l.
12
,
No.
2
,
Jun
2021
,
pp.
116
2
~
117
6
IS
S
N:
20
88
-
8694
,
DOI: 10
.11
591/
ij
peds
.
v12.i
2
.
pp
116
2
-
117
6
1162
Journ
al h
om
e
page
:
http:
//
ij
pe
ds
.i
aescore.c
om
Controll
er desi
gn fo
r PV expe
rimental b
ench wit
h ADRC
strategy
super
vised by
Labview
cr
eated i
nterfa
ce
Naoufel
K
ha
l
di
1
, You
sse
f
B
arradi
2
,
K
ha
li
da
Z
azi
3
,
Ma
li
ka
Z
azi
4
1
Ene
rgy and
Sus
ta
in
abl
e
Deve
lop
me
nt
Resea
r
ch T
ea
m
,
High
Scho
ol
of Te
chnol
og
y,
Ibn
Zohr
Un
iv
ersit
y,
Guel
mi
m
,
Morocc
o
2,3,4
El
e
ct
ri
ca
l
En
gine
er
ing
Dep
artme
nt
in
N
at
ion
al School
of
Arts
a
nd
Craf
ts (
ENSA
M),
Mohamme
d
V Unive
rsity
,
Raba
t
,
Morocc
o
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Feb
20
,
2021
Re
vised
A
pr
3
,
20
21
Accepte
d
Apr
15
, 20
21
The
conve
r
te
r
c
ontrol
sch
eme
p
l
ays
an
im
po
rta
n
t
role
in
th
e
per
f
orma
nc
e
of
ma
xim
u
m
powe
r
poin
t
tracki
ng
(MP
PT)
al
gori
t
hms.
In
thi
s
work,
a
mod
el
has
bee
n
an
al
ys
ed,
d
esigne
d
an
d
simul
at
edon
Pow
er
Simul
at
or
software
and
in
Matl
ab
Sim
uli
nk.
A
har
dwa
re
im
pl
emen
tati
on
using
a
microc
ontro
ll
er
(Arduino
Mega
2560
base
d
on
ATme
ga2560)
i
s
provide
d
,
that
oper
at
eson
fee
dba
ck
fro
m
a
PV
pane
l
vol
ta
g
e
and curre
n
t
to
cont
rol the
oper
a
ti
on
of
DC
-
DC
conve
r
te
r
i
n
orde
r
to
dr
a
w
ma
x
im
um
po
wer.
N
ewa
ctive
disturb
anc
e
rej
e
ct
ion
con
tro
l
(AD
RC)
al
gor
it
hm
is
req
u
ired
to
ext
ra
ct
the
ma
x
im
um
power
of
the
s
ola
r
en
erg
y.
Th
is
MP
PT
cont
roll
er
inc
orpor
ates
a
boost
topol
ogy
that
en
suresa
two
con
tinuous
batter
y
in
seri
es
(12V,
5A
h)
ch
arg
ing
in
var
ious
conditions.
The
whol
e
of
the
result
s
show
s
in
one
h
a
nd
that
th
e
conve
rt
er
eff
i
ci
e
ncy
is
ver
y
sati
s
fac
tory
,
and
in
t
he
oth
er
h
and
a
ver
y
good
agr
eeme
nt
b
et
w
ee
n
th
e
result
s
s
im
ulated
and
th
ose
expe
ri
me
nt
a
l
in
te
r
ms
of
per
forma
n
ce.
Th
e
proposed
sys
tem
is
designe
d
in
Proteus,
and
implemente
d
on
har
dwar
e
w
it
h
a
gr
aphica
l
user
int
er
face
built
througho
ut
L
abview
software
.
Ke
yw
or
d
s
:
AD
RC
st
rateg
y
Ardu
i
no
bo
a
rd
Boo
st c
onve
rte
r
Lab
view
M
PP
T
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
:
Kh
al
di N
a
oufe
l
Energ
y
a
nd S
ust
ai
nab
le
Dev
e
lop
me
nt Resea
rch Team
High Sc
hool
of
Tech
nolo
gy
Ibn
Z
ohr
Univ
ersit
y,
8100
0,
Gu
el
mim
, M
orocco
Emai
l:
n.kh
al
di
@u
iz
.ac.
ma
1.
INTROD
U
CTION
In
De
ma
nd
f
or
el
ect
rical
e
ne
rgy
has
re
ma
rk
a
bly
inc
reas
ed
duri
ng
the
recent
yea
rs
w
it
h
gr
ow
i
ng
popula
ti
on
a
nd
industrial
pro
gr
ess
[1].
Si
nc
e
long
ti
me
ag
o,
f
os
sil
f
uels
hav
e
se
r
ved
as
the
major
s
ou
rce
of
gen
e
rati
ng
el
ec
tric
al
energy.
Howe
ver
So
la
r
PV
base
d
ap
pl
ic
at
ion
s
are
ge
ner
al
ly
most
pe
rtinent
to
distr
ibu
te
d
gen
e
rati
on
i
n
l
ocati
ons
w
her
e
the
gr
id
is
una
vaila
ble
or
unr
el
ia
ble,
al
so
t
he
gl
ob
al
instal
l
ed
ca
pacit
y
for
s
olar
PV
is
inc
resea
d
es
pecial
ly by
China
, I
ndia
a
nd
U
nited
Stat
es
[
2]
.
H
owe
ve
r,
t
he
ke
y
to
t
he
s
uccess
o
f
a
pp
l
ying
small
PV
e
nergy
s
ys
te
ms
is
a
simple,
lo
w
cost
a
nd
high
-
performa
nce
c
onve
rter.
S
o,
a
bo
os
t
c
onver
t
er
is
desig
ne
d
that
i
t
pro
vid
es
an
ou
t
pu
t
of
24V
DC
,
w
hich
is
t
he
batte
ry
in
put.
T
he
t
ran
s
fe
r
of
e
ne
rgy
res
ulti
ng
from
photov
oltai
c
conversi
on
remains
relat
iv
el
y
wea
k.
T
he
r
efore,
man
y
tra
ckin
g
c
on
tr
ol
s
trat
egies
ha
ve
been
pro
po
se
d
i
n
e
xi
sti
ng
li
te
ratures,
s
uch
as
pe
r
tur
b
a
nd
obse
r
ve
[3],
f
uzzy
log
ic
[
4]
,
par
as
it
e
capaci
ta
nce
,
an
d
oth
e
r
m
et
hods
[
5
]
-
[
6].
Some
of
t
hem
are
e
ve
n
im
plement
ed.
I
n
t
his
paper,
t
he
r
ole
of
M
PP
T
st
rateg
y
is
to
at
ta
in
the
M
P
P
of
t
he
P
V
pa
nel
eve
n
wit
h
va
riat
ion
of
the
at
mo
s
pheri
c
conditi
ons.
The
desig
ned
sy
ste
m
con
t
ro
ls
the
duty
c
ycle
of
t
he
co
nverter
a
nd
e
nsure
the
ope
ning
an
d
the
cl
os
in
g
of
the
M
osfe
t
usi
ng
a
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
Con
tr
oller desi
gn for
PV ex
pe
riment
al b
e
nc
h w
it
h
AD
RC st
r
ategy su
pervis
ed by
…
(
Na
ou
fe
l K
ha
ldi
)
1163
desig
ne
d
dri
ve
r
w
hich
deliv
e
r
a
high
pulse
with
mod
ulati
on
(
PWM
)
frequ
e
nc
y
[
7].
The
cu
rr
e
nt
a
nd
th
e
vo
lt
age
n
e
ede
d as i
np
uts for t
he
c
on
t
ro
ll
er
ar
e obtai
ned th
rough
a
d
e
sig
ned cu
rr
e
nt and
vo
lt
age sen
s
ors.
The
syst
em
w
as
desig
ned
by
Prote
us
,
sim
ul
at
ed
by
Ps
pic
e
sim
ulator
a
nd
M
at
la
b
Sim
ul
ink
us
i
ng
a
new
A
DRC str
at
egy
to e
xtrac
t t
he
maxim
um powe
r
f
rom th
e PV
pan
el
[8
]
-
[
9]. Th
e
pro
po
sed
met
hod
is
a n
ew
te
chnolo
gy
for
est
imat
ing
a
nd
c
ompe
ns
at
in
g
uncertai
nties
an
d
distu
r
bances,
wh
ic
h
ha
ve
bee
n
ex
plore
d
a
nd
us
e
d
recently
a
s
an
al
te
r
native
over
cl
assic
al
te
chn
iq
ues
a
nd
especial
ly
P
ID
co
ntr
oller
[10
]
-
[
11].
U
nlike
man
y
existi
ng
co
ntr
ol
meth
ods,
the
AD
RC
do
es
not
re
quire
the
acc
ur
at
e
mathemat
ic
al
model
of
th
e
plant.
M
ore
ov
e
r,
sel
e
ct
ing
the
plant
order
n
of
t
h
e
AD
RC
is
qu
it
e
flexible
wh
ic
h
make
t
his
co
nt
ro
l
m
ore
co
nve
nient
to
ap
ply
in
ma
ny
co
ntr
ol
s
ys
t
ems
[
10
]
-
[
12].
The
meth
od
i
nvolv
e
s
th
ree
blo
c
ks
:
diff
e
re
ntiat
or
tr
acke
rs
(DT)
,
feedbac
k
c
on
t
r
oller
(
FC),
an
d
exten
de
d
sta
te
obser
ve
r
(
ES
O)
w
hich
is
th
e
main
pa
rt
of
the
c
om
ma
nd
[12
]
-
[
13]
.
A
n
im
ple
mentat
io
n
is
e
ns
ure
d
un
der
a
n
el
ect
ronic
platform
bo
a
rd
(
Ardu
i
no
M
e
ga
)
with
a
novel
con
t
ro
l
app
li
cat
io
n
de
velo
ped
in
La
bv
ie
w
e
nviro
nme
nt
w
hich
pro
vid
e
d
a
real
ti
me
ope
rati
on
a
nd
s
up
e
r
vision,
a
pr
act
ic
al
us
er
interface
,
a
nd
more
ot
her
a
bili
ti
es.
Re
su
lt
s
sho
ws
good
pe
rformance
,
high
accu
rac
y
a
nd
rob
us
tness
compa
rin
g
sim
ul
at
ion
a
nd im
plementat
io
n.
The
seq
ue
ntial
w
ork
flo
w
of
this
pa
pe
r
is
a
s
f
ollo
ws:
I
n
s
ect
ion
2,
c
omp
le
te
work
i
ng
proce
dure
of
the
s
ys
te
m
ha
s
bee
n
desc
rib
ed.
Se
ct
ion
3
cov
e
rs
t
he
des
ign
of
boos
t
DC
-
DC
c
onve
rter,
f
ollow
e
d
by
the
desig
n
of
c
urr
ent
se
nsor,
vo
l
ta
ge
se
nsor,
af
te
r
that
t
he
dr
i
ver
an
d
a
disc
us
sio
n
ab
out
a
MPPT
co
mm
and,
i
n
Sect
ion
4.
Sim
ulati
on
res
ults
and
ex
pe
rimen
ta
l
wo
r
ks
are
di
scusse
d
in
Sec
ti
on
5
a
n
d
6
re
sp
ect
ively
.
Las
tl
y,
in
sect
ion
7, a
pr
e
ci
se co
nclusi
on h
as
b
ee
n
a
dde
d
to
f
i
naliz
e th
e wor
k.
2.
COMPL
ET
E
SY
STE
M
O
V
ERVIEW
A
photov
oltai
c
cel
l
is
basical
ly
a
P
N
s
emic
onduct
or
j
un
ct
i
on
di
od
e
w
hic
h
is
c
onve
rting
sun
e
nerg
y
into
el
ect
rical
energ
y
[14].
DC
-
DC
c
onve
rte
r
is
widel
y
us
e
d
as
an
inter
m
ediar
y
betwee
n
the
PV
an
d
l
oa
d
or
batte
ry.
S
olar
pan
el
’s
c
urre
nt
an
d
volt
age
a
r
e
fe
d
t
o
the
mi
cro
c
ontrolle
r
wh
ic
h
deliver
a
su
it
able
duty
cycle
to
the bo
os
t c
onve
rter
on
the
bas
is of an
alg
or
it
hm
.
T
he whole
sy
ste
m
is
giv
e
n
in
Fig
ure
1.
Figure
1.
P
V
S
ys
te
m a
da
pted by Bo
os
t c
onve
rter
2.1.
So
lar
p
an
el
s c
ha
r
act
eri
s
tics
Parameter
s
of
so
la
r
pan
el
a
re
sh
ow
n
in
Ta
bl
e
1.
P
V
m
odul
e
is
made
by
s
olar
c
ompan
y
and
pro
duct
name
is
M
LP
-
020P.
N
um
e
ri
cal
method
cal
le
d
Ne
wton
-
R
aphso
n
is
use
d
to
deter
mine
the
operati
onal
po
i
nt
[14].
T
he
sy
ste
m
is
descr
i
bed
by
(
1)
-
(
3)
belo
w,
meani
ng
of
the
pa
ramete
rs
ex
pr
es
sed
can
be
co
nsult
ed
i
n
[6],
[14].
Table
1.
Para
m
et
ers
of
mlp
-
02
0P
Para
m
eters
Valu
es
Op
en
Circuit
Vo
lta
g
e (
Vo
c)
2
1
.7Vo
lt
Sh
o
rt
Circu
it
Cu
rr
en
t (
Isc)
1
.26
Am
p
Vo
ltag
e at
P
m
ax
(
Vm
p
p
)
1
7
.3Vo
lt
Cu
rr
en
t at
P
m
ax
(
I
m
p
p
)
1
.17
Am
p
Maximu
m
Powe
r
(
Pm
p
p
)
2
0
W
att
=
ℎ
−
[
(
×
)
−
1
]
(1
)
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
.
12
, N
o.
2
,
J
une
2021
:
116
2
–
117
6
1164
=
(
/
)
3
{
(
)
(
1
−
1
)
}
(2)
ℎ
=
[
(
+
(
−
)
]
×
(
)
(3
)
Figure
2.
Tem
per
at
ur
e
va
riat
ion ef
fect i
n PV
Figure
3.
I
rr
a
diati
on
var
ia
ti
on
eff
ect
in
P
V
3.
DC
-
DC
B
OO
ST CON
VER
TE
R
DC
-
DC
co
nver
te
rs
can
be
us
e
d
as
s
witc
hing
mode
re
gula
tor
s
to
c
onver
t
a
n
unre
gu
la
te
d
DC
vo
lt
ag
e
to
a
re
gu
la
te
d
DC
ou
t
pu
t
volt
age
[14
]
-
[
15
].
The
re
gu
la
ti
on
is
normall
y
ac
hieve
d
by
PWM
a
nd
t
he
c
on
trolle
d
switc
h
is
nor
m
al
ly MOS
FET
or IGBT
[16].
Figure
4
belo
w
s
how
s
a
ste
p
up
or
boos
t
conve
rter.
It
c
on
sist
s
of
a
P
V
volt
age
as
i
nput;
boos
t
inducto
r
L,
in
pu
t
ca
pacit
or
,
transisto
r
Mo
sfet
,
Diode
D,
outp
ut
capaci
tor
,
an
d
a
Lo
ad
or
batte
r
y.
M
a
ximum
power
is
reac
hed
w
hen
t
he
M
P
PT
al
gorith
m
changes
a
nd
a
dj
us
ts
t
he
du
t
y
cycle
of
t
he
s
witc
hed
dev
ic
e
[
17].
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
Con
tr
oller desi
gn for
PV ex
pe
riment
al b
e
nc
h w
it
h
AD
RC st
r
ategy su
pervis
ed by
…
(
Na
ou
fe
l K
ha
ldi
)
1165
Figure
4.
Bo
ost
DC
-
DC
c
onve
rter
with P
V
a
s in
pu
t
3.1.
Du
ty cy
cl
e
The
relat
ion
be
tween
in
pu
t
a
nd
ou
t
pu
t
volt
age
i
nclu
ded
a
facto
r
of
e
ff
ic
ie
ncy
w
hich
c
on
t
rib
ute
to
cal
culat
e a m
ore reas
on
a
ble
duty c
ycle
d:
V
out
=
η
1
−
d
V
in
(4)
wh
e
re
cal
le
d
e
ff
ic
ie
nc
y of t
he
conv
e
rter,
is a
bout
80%
,
t
hen
=
0
.
43
.
3.2.
Induc
to
r
Sele
ction
The
gi
ven
paramet
ers
are
re
qu
i
red
to
cal
c
ulate
the
powe
r
sta
ge
(in
duct
ance
an
d
capa
ci
ta
nce)
.
W
e
hav
e
in
pu
t
volt
age
=
17
, d
esi
red
outp
ut volt
age
= 25V
(slig
htly
higher
t
ha
n
the
r
at
ed v
oltage
of
t
he
two batt
eries s
eries w
hich
is
24V), a
nd the
ou
t
pu
t c
urre
nt
I
= 1A.
In
orde
r
to
cal
culat
e
the
est
i
mate
d
in
duct
a
nce
value
L
of
co
nv
e
rter,
we
nee
d
to
cal
cu
la
te
first
the
est
imat
ed
cu
rr
e
nt r
i
pp
le
Δ
I
:
Δ
I
= 0
.
3
×
I
×
(5)
Δ
I
= 0
.
45A
A
small
er
rip
pl
e
reduces
the
mag
netic
hyste
resis
losses
i
n
the
induct
or
c
ompone
nt.
T
he
switc
hing
fr
e
qu
e
nc
y for t
his
project
is
ta
ken
to
be f=
62.
5Khz.
The
i
nducto
r v
al
ue
L,
acco
r
din
g t
o
va
ried
pa
rameters
, is
give
n belo
w:
=
f
.
Δ
I
.
(6)
We take
L=
200
.
3.3.
Inpu
t
c
apacit
or
Sele
c
tion
The
e
quat
ion o
f
the
in
pu
t ca
pa
ci
tor
is
giv
e
n
as:
=
.
f²
.
Δ
V
.
(7)
We take
= 10
0
with
Δ
V
an
in
put v
oltage
rip
ple
Δ
V
<100m
V.
If
t
he
in
put
rip
ple
is
hi
gh
the
n
it
will
re
qu
i
r
e
a
la
rg
e
val
ue
of
capaci
ta
nce
,
an
d
t
hat
res
ul
t
an
inc
rease
in losse
s ca
us
e
d by the
ESR.
3.4.
Out
p
ut ca
pa
ci
to
r
Selecti
on
The o
utput cap
aci
tor
is:
=
f
.
Δ
V
.
(8)
we
ta
ke
=
800
.
Wh
e
re
Δ
V
<10
m
V
is
the
ou
t
pu
t
ri
pp
le
volt
age,
a
la
rg
e
value
c
an
ca
us
e
a
fl
uc
tuati
on
up to
t
he
li
mit
s where
the
outpu
t
volt
age c
rosses the
li
mit
s o
f
b
at
te
r
y rati
ngs.
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
.
12
, N
o.
2
,
J
une
2021
:
116
2
–
117
6
1166
4.
ADR
C CO
NTRO
LL
ER
D
E
SIGN
4.1.
ADR
C Alg
orithm
To
imp
r
ov
e
t
he
res
pons
e
s
peed
a
nd
c
ontrol
ada
ptabili
ty
of
t
he
powe
r
s
ys
te
m
[18],
the
a
pp
li
ed
AD
RC
meth
od
desi
gn
is
s
hown
in
Fig
ur
e
6.
T
he
mai
n
ide
a
of
this
te
ch
nolo
gy
is
to
est
imat
e
an
d
c
ompen
sat
e
the unmeas
ure
d
sta
te
o
f
t
he
s
ys
te
m
or
th
e to
ta
l disturbance
,
in
real t
ime, e
ven
with
ou
t a
n exp
li
ci
t mo
del
o
f
t
he
plant
an
d
only
f
rom
the
in
put
-
ou
t
pu
t
in
for
mati
on
[
19
]
-
[
20]
.
To
e
nsure
that,
t
he
co
ntr
oller
c
onta
ins
three
blo
c
ks
:
dif
fer
e
ntiat
or
trac
ker
s
(D
T
),
feedbac
k
co
ntr
oller
(FC
),
an
d
e
xten
de
d
sta
te
obser
ve
r
(ESO
).
It
in
vo
l
ve
s
al
so
a
n
in
ne
r
l
oop
t
o
reject
th
e
total
distu
r
ba
nce
a
nd
a
n
ou
t
er
on
e
to
deliv
er
the
de
sired
s
ign
al
[21
]
-
[
22]
.
T
his
pro
po
se
d
al
gor
it
hm
is
implem
ented
in
an
Ardu
i
no
bo
a
r
d
t
o
co
ntr
ol
t
he
du
ty
c
ycle
of
the
DC
-
DC
c
onve
r
te
r
in
order t
o
t
rack t
he MP
P
e
ven
with cli
mati
c va
riat
ion
s.
Figure
5. A
DR
C co
ntro
ll
er
str
uctu
re
As
sho
wn
i
n
F
igure
6,
t
o
extr
act
the
maxim
um
power
fro
m
the
PV
Sy
st
em,
the
AD
RC
con
tr
oller
is
desig
ne
d
to
set the
fo
ll
owin
g si
gn
al
e
rror at
0.
=
+
(9)
DT
is
us
e
d
to
arra
nge
t
he
tr
ansient
proces
s,
a
nd
to
get
t
he
dif
fer
e
ntial
sig
nals
of
cu
r
ren
t
dI
a
nd
vo
lt
age
dV
with tw
o dif
fer
e
nt
ia
tors’ track
ers
.
Figure
6. MPP
T b
ase
d
i
n ADR
C con
t
ro
ll
er
It
ov
e
rc
om
es
t
he
draw
bac
k
in
the
cl
assic
al
theory
s
uch
a
s
an
am
plific
at
ory
eff
ect
on
the
no
ise
.
M
at
he
mati
cal
form i
s e
xpress
ed by
:
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
Con
tr
oller desi
gn for
PV ex
pe
riment
al b
e
nc
h w
it
h
AD
RC st
r
ategy su
pervis
ed by
…
(
Na
ou
fe
l K
ha
ldi
)
1167
{
̇
1
=
2
̇
2
=
(
1
−
+
2
|
2
|
2
)
(10)
Wh
e
re
m
is
t
he
s
peed
f
act
or,
1
and
2
are
resp
e
c
ti
vely
t
he
t
racki
ng
an
d
diff
e
re
ntial
outp
ut,
X
is
the
input sig
nal a
nd
(
.
)
is a
nonlinea
r
f
unct
io
n.
ESO
is
the
c
ore
par
t
of
A
D
RC
,
it
con
t
rib
utes
to
get
t
he
model
uncert
ai
nty
a
nd
to
de
al
with
t
otal
disturba
nce a
ffec
ti
ng
P
V Po
w
er syste
m.
M
at
hemati
cal
str
uc
ture
is
giv
e
n
by:
{
=
1
−
̇
1
=
2
−
1
̇
2
=
3
−
2
̇
3
=
−
3
(11)
wh
e
re
y
is
t
he
ou
t
pu
t
of
the
s
ys
te
m.
1
,
2
and
3
are
res
pecti
vely
t
he
est
imat
io
n
of
t
he
outp
ut,
t
he
est
imat
io
n
of
the
der
i
vativ
e
of
the
outp
ut
and
the
est
imat
ion
of
the
dist
urba
nce
in
syst
em.
Re
ferrin
g
to
[
23],
the
obs
erv
e
r
gains
(
=
1
,
2
,
3
)
are
3
0
,
3
0
²
,
0
3
.
=
²
,
=
2
pa
rameters
of
t
he
fee
dbac
k
co
ntr
oller
(
F
C)
are
us
ed
to ge
ne
rate t
he
c
ontr
ol in
pu
t
0
.
0
is
denoted
as
the
ba
ndwidt
h
of
the
ob
se
r
ver
a
nd
the
band
width
of
the
fee
db
ac
k
c
on
t
ro
l.
Finall
y,
t
he
c
ontr
ol law is:
=
0
−
3
(12)
b
is t
he dist
urb
ance c
ompe
ns
a
ti
on
factor
4.2.
Inpu
ts
se
nso
r
s
The
P
V
pa
nel
ou
t
pu
t
is
f
ollo
wed
by
a
ci
rc
ui
try
f
or
se
ns
in
g
t
he
cu
rr
e
nt
a
nd
vo
lt
a
ge
nee
ded
as
i
nputs
for
the
M
P
PT
al
gorithm
as
s
how
n
in
F
ig
ure
7.
A
vo
lt
age
div
ide
r
wh
ic
h
conve
rts
the
volt
age
of
P
V
pa
nel
in
range
t
hat
is
0
-
5V,
so
that
t
he
A
rduin
o
boa
rd
can
i
nter
pre
t
the
vo
lt
age
le
vel.
T
he
co
mbi
nation
of
resis
tor
i
n
vo
lt
age
d
i
vid
e
r
is h
i
gh almost
10K
Ω
in
order t
o mi
nimize
t
he powe
r
loss
es.
Direct
c
urre
nt
sensing
is
base
d
on
O
hm’s
la
w.
A
c
urre
nt
s
ensin
g
m
odule
na
me
d
as
AC
S712
ca
n
be
us
e
d
[24],
but
we
desig
ne
d
o
ur
pr
op
e
r
sens
or
by
placi
ng
a
hi
gh
s
hunt
r
esi
stor
1
in
se
ries
with
th
e
s
yst
em
load,
a
vo
lt
a
ge
is
gen
erate
d
acro
ss
t
he
shun
t
re
sist
or
th
at
is
pr
oport
io
nal
to
the
s
ys
te
m
load
c
urre
nt.
The
vo
lt
age
ac
ross
the
s
hunt
can
be
meas
ur
e
d
by
dif
fer
e
ntial
am
plifie
r
L
M324
a
nd
deliv
er
a
read
a
ble
current
value
t
o
th
e
Ardu
i
no boa
rd.
Figure
7. cu
rr
e
nt and
volt
age
sens
or
s
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
.
12
, N
o.
2
,
J
une
2021
:
116
2
–
117
6
1168
Figure
8
sho
w
s
the
dri
ve
r
de
sign
e
d
t
o
t
ran
s
mit
the
si
gn
al
PWM
t
o
MOS
FET
tra
ns
ist
or
of
the
bo
os
t
conve
rter.
It
ha
s
capa
bili
ty
of
ope
rati
ng
at
high
P
W
M
f
re
qu
e
nc
y
62.5K
Hz
that
the
bo
os
t
co
nverter
ge
ts,
wit
h
a
s
pecific
duty
cycle
f
or
it
s
M
O
SFET
,
f
rom
the
Ard
uino
boa
rd.
T
he
re
is
a
le
vel
s
hift
ing
ci
rc
uitry
usi
ng
a
pu
s
h
-
pull
ampli
fier
as
a
n
outp
ut
sta
ge
betwe
en
A
r
du
i
no
bo
ard
a
nd
co
nver
te
r.
It
inc
rease
s
the
le
v
el
of
PWM
from A
rduin
o bo
a
r
d
acc
ordin
g
to
the
gate t
o so
ur
ce
volt
age
of
M
O
SFET
(fr
om
5V to
12
V)
.
Figure
8. D
rive
r
f
or
boos
t c
onver
te
r
5.
SIMULATI
O
N AND
RES
U
LT
S
In
orde
r
t
o
in
ve
sti
gate
the
performa
nce
of
our
sy
ste
m,
we
hav
e
imple
me
nted
t
he
c
ompl
et
e
ci
rcu
it
,
as
sh
ow
n
by
Fi
g
ure
4
,
i
n
first
ti
me
in
Ps
pice
s
imulat
or
e
nv
ir
onment
an
d
t
he
n
i
n
M
A
TL
A
B/
Simuli
nk
usi
ng
f
or
bo
t
h
sim
ulati
on a
n AD
RC
al
gorithm
.
In
the
first
sim
ulati
on
,
the
P
V
pa
nel
a
nd
the
M
PP
T
al
go
rith
m
we
re
model
ed
resp
ect
i
vely
by
a
so
la
r
pan
el
bl
oc
an
d
micr
oc
on
t
ro
ll
er
blo
c
.
The
P
V
pa
nel
is
co
nnect
ed
to
a
res
ist
ive
loa
d
t
hro
ugh
a
boost
c
onve
rter
whose d
uty cy
cl
e w
as a
dju
ste
d by MPPT
alg
or
it
hm
based o
n
the
v
al
ue of
current a
nd
vo
l
ta
ge
se
ns
ors
(Fi
g
9).
Figure
9.
The
whole s
ys
te
m
In
F
ig
ur
e
10,
t
he
ou
t
pu
t
po
w
er
c
urve
by
us
i
ng
M
PPT
al
go
rithm
w
hich
pr
oduce
a
s
uitabl
e
duty
cycle
us
e
d
to
reac
h
the
M
PP
,
has
an
e
xcell
ent
c
har
act
erist
ic
w
it
h
hi
gh
acc
uracy
a
nd
good
sta
bili
ty.
T
he
resu
lt
s
con
ce
r
ning
the
input
(
ou
t
pu
t)
vo
lt
age
a
nd
in
pu
t
(outp
ut)
c
urre
nt
of
the
c
onve
rter
was
presented
res
pec
ti
vely
in
Fig
ure
11
a
nd
12.
Th
ey
s
how
t
hat
the
e
le
ct
ric
qu
a
ntit
ie
so
sci
ll
at
e
ar
ound
the
opti
mal
value
s
under
the
op
ti
mal c
onditi
on
s
.
Finall
y,
Fi
gure
13
s
how
s
that
the
M
P
PT
al
gorith
m
can
quic
kly
trac
k
the
M
PP
(t=
10ms
)
with
go
od
eff
ic
ie
nc
y
a
nd
low oscil
la
ti
on ev
e
n wit
h
th
e
var
ia
ti
on
of
i
rradiat
ion
.
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
Con
tr
oller desi
gn for
PV ex
pe
riment
al b
e
nc
h w
it
h
AD
RC st
r
ategy su
pervis
ed by
…
(
Na
ou
fe
l K
ha
ldi
)
1169
Figure
10. O
utp
ut
Powe
r wit
h MP
PT al
gorithm
Figure
11.
Inp
ut & O
utput v
ol
ta
ge
of the
conv
e
rter
Figure
12. In
put &
O
utput C
urren
t
of the
co
nv
e
rter
Figure
13. O
utp
ut
Powe
r wit
h va
riat
ion
of ir
r
adiat
ion
In
the
s
eco
nd
simulat
ion,
the
power
s
ys
te
m
was
car
ried
out
in
MATL
A
B
Simuli
nk
as
sh
ow
n
in
Figure
13.
Par
amet
ers
use
d
i
n
A
DRC
al
gor
it
hm
f
or
t
he
bo
os
t
co
nverter
a
re
0
=
440
,
0
=
110
and
b=
600.
Perfo
rma
nces
and
desig
n
re
su
lt
s
of
t
he
c
on
t
ro
l
sy
ste
m
are
il
lustrate
d
thr
ough
ou
t
two
cases.
I
n
the
first
scenari
o,
reali
s
ti
c
ram
p
up
/
down
ra
diati
on
from
10
00w/m²
to
300w/m²
a
r
e
ap
plied
w
hile
the
te
mp
e
rat
ur
e
is
kep
t
c
onsta
nt
at
298K.
I
n
th
e
seco
nd
case
,
the
te
m
per
at
ure
c
hanges
it
s
values
as
fo
ll
ows:
323K
to
298K
a
t
=
1.2
a
nd fr
om
298K
retu
r
ning to
32
3K a
t 1.8s
mean
w
hile t
he
ra
diati
on is maintai
ne
d at
1
00
0w
/m²
.
Figure
14.
Pow
er syste
m sim
ul
at
ion
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
.
12
, N
o.
2
,
J
une
2021
:
116
2
–
117
6
1170
The
res
ulti
ng
c
on
t
ro
l
pe
rform
ances
of
t
he
first
case
is
il
lus
trat
ed
by
Fi
gur
e
14
t
o
Fig
ur
e
17
a
nd
f
or
the
seco
nd
on
e
is
il
lustrate
d
by
Fig
ur
e
18
t
o
Fi
gure
21
.
F
igures
14
a
nd
18
sho
ws
that
the
A
DRC
te
chn
i
qu
e
deliver
a
co
ntr
ol
sig
nal
w
hic
h
dri
ve
the
boos
t
co
nverter
to
trac
k
the
MPP
ve
r
y
quic
kl
y.
A
s
il
lustrat
ed
in
Figure
14,
t
he
PV
outp
ut
power
is
20
w,
6w,
c
orres
pondi
ng
to
t
he
M
P
P
P1
an
d
P3
in
Figure
3
.
Mo
r
eov
e
r,
no
ti
ng
t
hat
th
e
outp
ut
vo
lt
a
ge
is
re
gu
la
te
d
to
it
s
de
sire
d
value
25
with
good
performa
nce.
Finall
y,
t
he
con
t
ro
l
in
pu
ts
of
co
nverter
dis
cl
ea
rly
bo
unde
d.
Se
quen
ti
al
ly
al
l
these
fi
gures
coi
nc
ide
with
the
oret
ic
al
pr
e
dicti
on and
com
pany speci
fied val
ue
whi
c
h
e
nsures
the
validit
y of t
he system.
Accor
ding
to
f
igures,
t
he
A
D
RC
com
man
d
pr
ese
nt
e
xcell
ent
c
har
act
erist
i
cs
an
d
good
pe
rformances
even
withsome
remarka
ble
osc
il
la
ti
on
an
d
hi
gh
ov
e
rs
hoot
a
t
the
first
500ms,
w
hich
due
to
the
ti
me
ta
ken
by
the
ap
proac
h
t
o
t
ra
ns
f
orm
th
e
data
of
the
pro
blem
usi
ng
Simuli
nk
int
o
a
set
of
res
ul
ts.
Othe
rw
ise
,
for
a
n
impleme
ntati
on
,
A
DRC
met
hod
is
ge
ner
al
l
y
c
hea
per
a
nd
le
ss
co
mp
le
x
because
it
re
quires
only
t
wo
tu
ning
par
a
mete
rs
t
o
a
dju
st.
Figure
14.
O
utp
ut
Powe
r wit
h i
rr
a
dia
ti
on ch
a
ng
e
Figure
15.
Inp
ut & o
utput v
ol
ta
ge
co
nverte
r
Figure
16.
Inp
ut &
i
nduct
or
c
urren
t
of the
conv
e
rter
Figure
17.
D
uty
c
ycle of t
he
c
onve
rter
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
Con
tr
oller desi
gn for
PV ex
pe
riment
al b
e
nc
h w
it
h
AD
RC st
r
ategy su
pervis
ed by
…
(
Na
ou
fe
l K
ha
ldi
)
1171
Figure
18.
O
utp
ut
Powe
r wit
h t
empe
ratu
re c
hange
Figure
19.
Inp
ut & o
utput
vol
ta
ge
co
nverte
r
Figure
20.
Inp
ut &
i
nduct
or
c
urren
t
of the
conv
e
rter
Figure
21.
D
uty
c
ycle of t
he
c
onve
rter
6.
EXPERI
MEN
TAL WO
RK
The
M
PP
T
al
gorith
m
wa
s
be
en
im
plemente
d
by
us
in
g
a
n
Ardu
i
no
boa
rd
and
Lab
vie
w
app
li
cat
io
n
wh
ic
h
a
re
desi
gn
e
d
in
s
uch
way
that
can
pe
rform
with
an
y
oth
e
r
resea
rc
h
M
P
PT
al
gori
thm.
F
or
our
M
PP
T
al
gorithm,
t
he
analo
g
sig
nals
and
obta
ine
d
by
the
d
esi
gne
d
sens
or
s
is
a
me
ans
of
tw
o
pins
A0
an
d
A1
of the
Ard
uino
dev
ic
e.
6.1.
System de
sign
a
n
d simul
at
i
on o
n ISIS
Figure
22
pres
ents
the
desi
gn
of
t
he
whole
sy
ste
m
on
IS
I
S
Prote
us.
In
view
of
a
mi
nimiza
ti
on
of
losses,
the
bo
ost
co
nv
e
rter
is
desig
ne
d
a
rou
nd
a
M
osfet
IRFP25
0
wh
ic
h
present
a
s
mall
(
on)=
0.085Ω,
a
nd
arou
nd
a
lo
w
vo
lt
age
diode
su
c
h
as
Schott
ky
diode
w
hos
e
well
ada
pted
f
or
t
his
a
ppli
cat
ion
due
to
it
s
lo
w
vo
lt
age
dr
op a
nd sho
rt sw
it
c
hi
ng
ti
me.
A
vo
lt
age
an
d
current
sens
ors
are
nee
ded.
T
he
fi
rst
one
t
o
detect
volt
age
of
t
he
P
V
pa
ne
l
by
us
i
ng
a
div
ide
r:
=
16
16
+
9
+
5
+
15
(13)
wh
ic
h
m
us
t
be
<
5V
beca
us
e
the
a
nalo
g
i
nput
in
Ardu
i
no
is
li
mit
ed
to
5V.
T
he
seco
nd
one
to
mea
sur
e
the
ou
t
pu
t
pa
nel c
urren
t a
nd the
y t
ran
s
mit
te
d
to
the an
al
og p
i
n A1.
Evaluation Warning : The document was created with Spire.PDF for Python.