Int
ern
at
i
onal
Journ
al of Ele
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
10
,
No.
5
,
Octo
be
r
2020
,
pp.
4770
~
4781
IS
S
N: 20
88
-
8708
,
DOI:
10
.11
591/
ijece
.
v10
i
5
.
pp
4770
-
47
81
4770
Journ
al h
om
e
page
:
http:
//
ij
ece.i
aesc
or
e.c
om/i
nd
ex
.ph
p/IJ
ECE
Des
i
gn a
nd
investigations
of MPP
T strate
gies for
a wind
e
nergy
co
n
version s
yste
m based on d
oubly fed in
du
ctio
n
gene
rator
Bouch
aib
Rac
hed, Mus
t
ap
h
a
El
h
aroussi,
El
ha
ssane
Abdel
mouni
m
La
bora
tor
y
of
Si
gnal
Ana
l
y
s
is a
n
d
Inform
at
ion
Proce
ss
ing,
FS
T
S
et
t
at
,
Hass
an
1
st
Univer
sit
y
,
Morocc
o
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
ul
30
, 2
019
Re
vised
Ma
r
1
2
,
2020
Accepte
d
Ma
r
26
, 202
0
The
purpose
of
thi
s
work
is
to
design
and
to
d
iscuss
var
ious
strategie
s
t
o
opti
m
iz
e
th
e
pr
oduct
ion
of
a
wind
e
ner
g
y
c
onver
sion
chai
n
base
d
on
the
doubl
y
fed
induc
t
ion
g
enerat
or
(DF
IG),
b
y
ca
p
turi
ng
the
m
axi
m
um
power
at
th
e
win
d
turbi
ne
,
using
m
axi
m
um
po
wer
point
tracki
ng
(
MP
PT)
and
pit
ch
cont
ro
l.
The
proposed
cont
rols
al
low
the
g
ene
r
at
or
to
m
onit
or
the
opti
m
al
op
era
t
ing
point
s
of
the
turbi
nes
reg
ard
le
ss
of
wind
spee
d
var
iations,
s
y
s
t
em
par
amete
rs
disturba
nc
e,
and
par
amet
ers
var
iation.
Sim
ula
ti
on
of
W
ECS
b
ase
d
on
a
1.
5
MW
wou
nd
rotor
induc
tion
gene
rat
or
under
MA
TL
AB/S
IMU
LINK
is
ca
rri
ed
out
us
ing
the
PI
cont
r
oll
er
(PIC)
,
RS
T
cont
roller
and
fuz
z
y
log
ic
cont
roller
(FLC)
.
Anal
y
s
is
and
c
om
par
isons
are
m
ade
for
d
i
ffe
ren
t
op
erati
n
g
sce
nar
ios:
R
ef
ere
nc
e
tr
ac
king
,
r
obustness
under
var
i
abl
e
wind
spee
d
condi
ti
ons
a
nd
par
amete
rs
var
iatio
n
.
The
applic
at
io
n
of
FLC
provide
s
a
ver
y
int
ere
st
ing
o
utc
om
e
for
the
robustn
ess and the d
y
namic
cha
l
le
nges.
Ke
yw
or
d
s
:
Dou
bly fe
d
in
duct
ion ge
ne
rator
Fu
zzy
lo
gic c
ontr
oller
MPPT
Pit
ch
a
ngle
c
ontr
ol
RST co
ntr
oller
Copyright
©
202
0
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
Bou
c
haib
R
ach
ed
,
Ph
ysi
cs a
nd E
nginee
rin
g
Scie
nce
Dep
a
rtm
ent,
Hassa
n 1
st
U
ni
ver
sit
y,
Sett
at,
M
orocc
o.
Em
a
il
:
bo
uc
haib.rache
d@gm
ai
l.com
1.
INTROD
U
CTION
In
rece
nt
ye
ar
s,
the
en
vir
onm
ental
and
/or
ecolog
ic
al
co
ntext
has
bec
om
e
an
i
m
po
rtant
or
ev
e
n
pr
e
dom
inant
factor
.
Indee
d,
the
increa
sin
g
consum
ption
of
el
ect
ric
ene
r
gy
in
the
w
or
l
d,
the
lim
it
a
ti
on
an
d
the
pro
gr
essi
ve
dep
le
ti
on
of
fo
ssil
f
uels
(
oil,
gas,
coal.
..)
char
act
e
rized
by
the
em
issi
on
s
of
car
bon
di
ox
id
e
pro
du
ce
d
du
rin
g
their
c
om
bu
s
ti
on
,
on
t
he
on
e
hand,
a
nd
th
e
risk
of
nu
cl
e
ar
powe
r
pla
nt
s
(uran
i
um
rad
ia
ti
on
)
with
the
pro
duct
ion
of
rad
i
oa
ct
ive
residu
e
s
that
are
dif
ficul
t
to
treat
fo
r
the
nu
cl
ear
powe
r,
on
the
oth
e
r
hand,
are
ver
y
real
pro
blem
s
that
ha
ve
push
e
d
resea
rch
e
rs
t
o
e
xp
l
oit
othe
r
s
ources
of
s
us
ta
ina
ble
an
d
m
or
e
env
i
ronm
ental
l
y fr
ie
ndly
en
e
r
gy. T
hese
subs
ti
tute
m
eans are t
he
"
ren
e
wa
bl
e energie
s"
[1
,
2].
The
m
ai
n
fo
rm
s
of
ren
e
wa
ble
energies
are
hy
dro
power,
so
la
r
therm
al
,
photo
volt
ai
c
and
w
ind
power,
as
well
as
bi
om
ass
and
ge
oth
e
rm
al
ener
gy.
Am
ong
al
l
ren
e
wab
le
e
nergies
co
ntri
bu
ti
ng
to
el
ec
tric
it
y
pro
du
ct
io
n,
wi
nd
ene
rg
y
was
the
first
t
o
be
e
xp
l
o
it
ed.
It
is
c
le
arly
in
a
go
od
po
sit
io
n,
not
as
a
re
placem
e
nt
f
or
conve
ntion
al
s
ources,
but
as
a
com
ple
m
ent
ary
al
te
rn
at
ive
to
nucl
ear
e
ne
rg
y.
C
urren
tl
y,
the
vast
m
ajorit
y
of
wind
ene
r
gy
conve
rsion
syst
e
m
s
us
e
do
ub
l
y
fed
in
duct
io
n
ge
ne
rato
r
(
D
FI
G
),
w
hich
gi
v
e
the
a
dv
a
nt
age
of
bette
r
e
xp
l
oiti
ng
wi
nd
res
ou
rces
in
di
ff
e
re
nt
wi
nd
co
ndit
ion
s
t
hanks
t
o
it
s
va
riable
-
sp
ee
d
ope
rati
on
a
nd
the use
of
sta
ti
c pow
e
r
c
onve
rters as
so
ci
at
ed
w
it
h
c
ontr
ol s
yst
e
m
s [
3].
A
descr
i
ptio
n
of
the
dif
fer
e
nt
com
ponen
ts
of
the
WECS
ba
sed
on
a
D
FIG
is
il
lust
rated
in
Fig
ur
e
1.
It
can b
e
di
vide
d
into
tw
o
m
a
in
pa
rts.
T
he
m
echan
ic
al
p
art wh
ic
h
is
com
po
se
d
of
the
t
urbine,
t
he
gearb
ox
a
nd
the
tra
ns
m
issio
n
s
haf
t.
The
n,
the
el
ect
rical
pa
rt
w
hich
is
com
po
se
d
of
the
D
FIG
w
hose
sta
tor
is
at
ta
ched
di
rectl
y
to
the
el
ect
rical
gr
id,
as
long
as
the
ro
t
or
is
li
nked
to
the
gri
d
t
hro
ugh
tw
o
sta
ti
c
bid
irect
io
nal
pow
e
r
conve
rters.
T
o
op
ti
m
iz
e
the
el
ect
rical
po
w
er
pro
duct
ion
of
wind
tur
bine
s,
dif
fer
e
nt
strat
egies
an
d
c
on
t
ro
l
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Desig
n a
nd inv
est
iga
ti
ons
of
MPPT str
ategi
es for a wi
nd e
ner
gy
c
on
ver
sion syst
em
...
(
B
ou
c
haib
Rac
he
d)
4771
desig
ns
wer
e
a
pp
li
ed
t
o
wind
tur
bin
e
s
i
n
ea
ch
of
the
opera
ti
ng
zo
nes
su
c
h
as
the
M
PPT
con
t
ro
l
strat
e
gy
and
pitch a
ng
le
con
trol to
ex
tr
act
the m
axi
m
u
m
p
ow
e
r.
W
it
h
the
gro
w
ing
interest
in
wind
powe
r
ge
ner
at
io
n
thr
ough
D
FIG,
se
ve
ral
con
tr
ol
m
et
hods
ha
ve
been
de
velo
pe
d.
In
ge
ner
al
,
a
con
t
ro
l
sy
ste
m
aim
s
at
m
on
it
or
in
g
a
syst
em
to
ob
ta
in
a
correct
f
un
ct
io
ning
of
the
la
tt
er.
T
he
co
ntr
ol
al
lo
w
s
the
syst
e
m
to
trac
k
t
he
de
sired
res
pons
e
f
or
a
giv
e
n
i
nput
a
nd
to
r
eject
the
eff
ect
s
of
disturba
nces.
Unfortu
natel
y,
wh
e
n
syst
e
m
m
od
el
li
ng
is
diff
ic
ult,
the
de
sign
of
the
con
t
ro
l
syst
e
m
can
be
ver
y
com
plex
.
This
di
ff
ic
ult
y
is
caused
by
strong
no
n
-
l
inearit
ie
s,
pa
ra
m
et
ric
var
ia
ti
on
s
or
inaccu
racies
in
the
i
ntern
al
va
riables.
But
i
n
s
pite
of
t
his,
seve
ral
m
et
ho
ds
hav
e
bee
n
f
ound
to
deal
with
the
pro
blem
,
s
uch
a
s
sli
ding
m
od
e
con
tr
ol
(
SMC
)
[
4
-
6],
ba
ckstep
ping
c
ontr
ol
(BC)
[
7
-
10]
,
non
-
li
nea
r
con
t
ro
l
by
sta
ti
c
or
dy
nam
ic
state
feed
bac
k
[
11]
.
Al
l
these
te
chn
i
ques
ha
ve
diff
ic
ulty
in
that
their
desig
n
is
str
ongly
dep
e
ndent
on
the
syst
e
m
m
od
el
.
These
dif
ficult
ie
s
hav
e
le
d
in
recent
ye
ars
to
the
dev
el
opm
ent
of
new
c
on
t
ro
l
te
chn
iq
ues
bas
ed
on
arti
fici
a
l
intel
li
gen
ce
(n
e
ur
al
netw
orks,
fu
zzy
lo
gi
c,
ge
netic
al
go
rithm
s..
.)
t
o
con
t
ro
l
these syst
em
s.
In
t
his
co
ntext
,
this
w
ork
f
oc
us
es
on
the
c
ontr
ol
of
a
D
FIG
-
base
d
wi
nd
energy
co
nver
sion
syst
em
(
W
EC
S)
,
em
ph
asi
zi
ng the
u
s
e o
f
fuzzy
lo
gi
c co
ntr
ol and
unde
rstan
ding it
in
c
onjun
ct
io
n wit
h
c
onve
ntion
al
PI
and
RS
T
-
base
d
vect
or
c
on
t
ro
l.
Test
s
of
the
adap
te
d
con
tr
ol
strat
egies
will
be
carried
ou
t
under
the
Ma
tl
ab/Sim
ul
ink
s
oft
wa
re
e
nv
i
ronm
e
nt.
T
he
res
ults
of
t
he
var
i
ou
s
te
sts
perf
or
m
ed
in
nu
m
erical
si
m
ulati
on
as
well
as the
test
s w
il
l be
prese
nted
a
nd inte
rpreted.
The
pap
e
r
is
st
ru
ct
ur
e
d
in
t
he
fo
ll
owin
g
wa
y:
Sect
ion
2
w
il
l
be
devoted
to
syst
em
descr
ipti
on
an
d
dynam
ic
m
od
el
.
The
c
ontr
o
l
s
trat
egies
are
ca
refull
y
el
abora
te
d
in
sect
io
n
3,
fo
ll
owe
d
by
si
m
ulati
on
res
ults
in
sect
ion
4.
W
e
will
term
inate
this wo
rk w
it
h
a brie
f
c
on
cl
usi
on
.
Figure
1. The
DF
I
G
-
base
d w
ind
c
onve
rsion
ch
ai
n
2.
SY
STE
M DESC
RIPTIO
N AND
D
YNA
MIC
MO
DEL
2.1. Win
d
tur
bine
The
ae
rodyna
m
ic
tor
que,
w
hi
ch
is c
onver
te
d by a
wind tu
r
bin
e,
is e
xpress
ed
as
[12
-
14
]
:
3
1
(
λ
,
β
)
ρ
Ω
2
Ω
a
e
r
o
w
in
d
a
e
r
o
tt
P
PV
T
C
S
(
1)
w
he
re
(
β
,
λ
)
P
C
is
the
powe
r
c
oeffici
ent,
λ
is
def
i
ned
by:
Ω
λ
t
bl
ad
e
w
i
nd
R
V
,
a
nd
is
the
a
ngle
of
blade
or
ie
ntati
on
,
Ω
t
is
the
Rotat
io
nal
sp
ee
d
of
t
he
tur
bin
e
,
w
i
n
d
V
is
the
wind
s
peed,
a
nd
S
is
the
are
a
swe
pt
by
the
pales
of
th
e
tur
bin
e.
T
he
dynam
ic
of
the
wi
nd
t
urbi
ne
is
a
f
unct
io
n
of
th
e
m
echan
ic
al
tor
que
(
m
T
),
the total
ine
rtia
(
J)
, a
nd the
coe
ff
ic
ie
nt
of tot
al
f
rict
io
n
(
K).
It is g
i
ven b
y:
Ω
Ω
m
G
t
N
,
ae
ro
m
G
T
T
N
and
ΩΩ
m
m
e
m
m
J
T
K
T
(
2)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
5
,
Oct
ob
e
r 2
020
:
47
70
-
47
81
4772
2.2.
DFIG m
od
el
The
e
quat
ions
descr
i
bing th
e
functi
onin
g of
DF
I
G
a
re
wr
it
t
en
as
foll
ows [15
-
17
]
:
Ele
ct
rical
eq
ua
ti
on
s
sd
s
sd
sd
s
sq
sq
s
sq
sq
s
sd
rd
r
rd
rd
r
rq
rq
r
rq
rq
r
rd
d
V
R
I
dt
d
V
R
I
dt
d
V
R
I
dt
d
V
R
I
dt
a
nd
sd
s
sd
rd
sq
s
sq
rq
rd
r
rd
sd
rq
r
rq
sq
L
I
MI
L
I
MI
L
I
MI
L
I
MI
(3)
Powers
Exp
res
sion
s
s
s
sq
sq
sd
sd
sq
sq
sd
sd
P
v
I
v
I
Q
v
I
v
I
a
nd
r
rq
rq
rd
rd
r
rq
rq
rd
rd
P
v
I
v
I
Q
v
I
v
I
(4)
e
m
rd
sq
rq
sd
T
pM
I
I
I
I
(5)
3.
CONTR
OL
S
TRATEGIE
S
3.1. Tr
ackin
g of the
m
ax
i
mum p
ower p
oint o
f the
proposed
wi
nd
s
ystem
The
ob
j
ect
ive
of
this
par
t
is
to
de
velo
p
an
MPPT
strat
egy
for
optim
al
ener
gy
tra
ns
fe
r
in
DF
I
G
-
base
d
wind e
nergy sy
stem
s.
3.1.1.
MPPT
b
as
ed
on PI
co
nt
r
oller
T
he
pr
op
os
e
d
s
t
r
a
t
e
gy
c
on
t
r
o
l
s
t
he
s
pe
e
d
r
a
t
i
o
λ
i
n
s
uc
h
a
w
a
y
t
ha
t
i
t
c
an
be
op
t
i
m
a
l
(
opt
λ
)
,
a
n
d
t
he
r
e
f
or
e
t
he
t
or
qu
e
i
s
a
dj
u
s
t
e
d
t
o
a
n
op
t
im
a
l
va
l
ue
opt
T
a
t
w
hi
c
h
t
he
p
ow
e
r
c
oe
f
f
i
c
i
e
nt
i
s
m
a
xi
m
um
(
PMAX
c
)
[
18
]
,
w
hi
c
h
i
s
gi
ve
n
by
(
6
)
:
2
m
op
t
op
t
Tk
w
he
r
e
,
5
3
1
2
opt
b
la
d
e
P
M
A
X
opt
C
KR
(6)
T
he
s
t
ud
i
e
d
P
I
c
on
t
r
ol
i
s
i
l
l
us
tr
a
t
e
d
i
n
F
i
gu
r
e
2
.
Figure
2. Dia
gra
m
o
f
t
he
MPP
T b
ase
d o
n
P
I c
on
t
ro
ll
er
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Desig
n a
nd inv
est
iga
ti
ons
of
MPPT str
ategi
es for a wi
nd e
ner
gy
c
on
ver
sion syst
em
...
(
B
ou
c
haib
Rac
he
d)
4773
3.1.2.
MPPT
b
as
ed
on R
ST c
on
t
rolle
r
T
he
R
S
T
c
on
t
r
ol
l
e
r
i
s
ba
s
e
d
on
t
he
po
l
e
pl
a
c
em
e
nt
t
he
or
y
by
s
pe
c
i
f
y
i
ng
a
n
a
r
b
i
t
r
a
r
y
s
t
a
bi
li
t
y
po
l
y
no
m
i
al
D
(
p
)
a
n
d
c
om
pu
t
i
ng
S
(
p)
a
n
d
R
(
p)
a
c
c
or
di
ng
t
o
B
e
z
ou
t
'
s
e
qu
a
ti
on
[
19
]
.
D
A
S
B
R
w
i
t
h
:
d
e
g
D
p
d
e
g
A
p
d
e
g
S
p
(7)
T
he
s
t
r
uc
t
ur
e
o
f
t
he
M
P
P
T
ba
s
e
d
o
n
R
S
T
c
o
nt
r
ol
l
e
r
i
s
de
pi
c
t
e
d
i
n
F
i
gu
r
e
3
[
20
]
.
F
i
gu
r
e
3.
D
i
a
g
r
am
of
t
he
M
P
P
T
ba
s
e
d
on
R
S
T
c
on
t
r
ol
l
e
r
s
3.1.
3
.
MPPT
b
as
ed
on f
uz
z
y
log
ic
c
ontrolle
r
Fu
zzy
log
ic
c
on
t
ro
l
m
akes
it
po
ssible
to
i
m
it
at
e
hu
m
a
n
reas
on
i
ng
by
har
nes
sin
g
the
div
e
rse
inf
or
m
at
ion
col
le
ct
ed
in li
ng
ui
sti
c fo
rm
s.
It
was
de
velo
pe
d by Zadeh
in 1
965
f
ro
m
the t
heory of
fu
zzy
su
bse
ts
and
has
the
a
dvanta
ge
of
t
ol
erati
ng
the
un
ce
rtai
nty
of
th
e
m
od
el
and
c
om
pen
sat
ing
it
s
eff
ect
.
T
he
m
e
tho
d
pr
ese
nted
bel
ow
c
onsist
s
of
com
bin
ing
the
P&
O
al
go
rith
m
with
a
fu
zz
y
con
t
ro
ll
er
by
ta
kin
g
int
o
account
the d
i
recti
on of
v
a
riat
ion
of
th
e d
ist
urba
nces
.
A
cc
ordin
g
t
o
F
igure
4, we
have:
a.
if a p
os
it
i
ve va
riat
ion
in
S
pee
d
has
ca
us
e
d
a p
os
it
ive
var
ia
ti
on
i
n
Power
, a
po
sit
ive
var
ia
ti
on
i
n
sp
ee
d
m
us
t
be
a
dd
e
d;
b.
if
a
neg
at
ive
va
riat
ion
in
Spe
ed
has
cau
sed
a
po
sit
ive
va
riat
ion
in
Power
,
a
neg
at
ive
va
riat
ion
in
sp
e
e
d
m
us
t be added
;
c.
if
a
neg
at
ive
va
riat
ion
in
S
pe
ed
has
ca
us
e
d
a
neg
at
ive
var
i
at
ion
in
P
ow
e
r
,
a
neg
at
ive
va
riat
ion
in
s
pee
d
m
us
t be added
;
d.
if
a
posit
ive
va
riat
ion
i
n
S
pe
ed
has
ca
us
e
d
a
ne
gative
va
riat
ion
i
n
P
ower
,
a
posit
ive
va
riat
ion
i
n
s
pe
e
d
m
us
t be added.
In su
m
m
ary,
this m
e
tho
d i
s d
e
ta
il
ed
in the
T
a
ble 1
.
Figure
4. MPP
trackin
g
m
et
hod
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
5
,
Oct
ob
e
r 2
020
:
47
70
-
47
81
4774
Table
1.
MP
P t
rack
i
ng
E
DE
NB
NM
NS
AZ
PS
PM
PB
NB
NB
NB
NB
NB
NM
NS
AZ
NM
NB
NB
NB
NM
NS
AZ
PS
NS
NB
NB
NM
NS
AZ
PS
PM
AZ
NB
NM
NS
AZ
PS
PM
PB
PS
NM
NS
AZ
PS
PM
PB
PB
PM
NS
AZ
PS
PM
PB
PB
PB
PB
AZ
PS
PM
PB
PB
PB
PB
To
im
ple
m
ent
this
al
gorit
hm
in
a
f
uzzy
con
t
ro
ll
er,
w
e
us
e
d
a
Ma
m
dan
i
m
od
el
as
presente
d
i
n
F
igure
5
[
1,
15,
21]
.
The
pro
pose
d
co
ntr
oller
has
tw
o
in
pu
ts
:
the
var
ia
ti
on
in
DFIG
powe
r
an
d
the
var
i
at
ion
in
r
otati
on
al
s
peed
,
a
nd
onl
y
on
e
outp
ut,
w
hich
is
t
he
optim
al
sp
eed
var
ia
ti
on
t
hat
ens
ures
m
axim
u
m
aerodynam
ic
e
ff
ic
ie
ncy.
T
hes
e
var
ia
bles
are
def
i
ned
durin
g
one
sam
ple
tim
e.
These
va
r
ia
bles
are
com
m
on
ly
fu
zzi
fie
d
by
s
ymm
et
rical
tri
angular
m
e
m
b
ersh
i
p
functi
ons.
The
f
uzzy
set
s
ar
e
de
fi
ned
as
fo
ll
ow
s:
NB:
Neg
at
ive
Bi
g;
NM:
Ne
gativ
e
Me
diu
m
;
NS
:
Ne
gative
S
m
al
l
;
AZ:
A
bout
Zer
o;
PS:
Po
sit
ive
Sm
a
ll
;
NM:
Po
sit
ive
Me
dium
;
PB:
Po
sit
iv
e
Bi
g.
The
f
uz
zy
ru
le
s,
f
or
de
te
rm
ining
t
he
outp
ut
va
riable
of
the
c
ontr
oller
as
a
functi
on
of
i
nput
va
riables
are
grou
ped in Ta
ble 1.
Figure
5. Bl
oc
k diag
ram
o
f
f
uzzy co
ntr
ol
3.2. Pitc
h
angle co
nt
r
ol s
tr
at
e
gies
Accor
ding
to
(
1
)
,
t
he
wind
powe
r
ca
pture
d
by
a
t
urbine
is
pr
opor
ti
on
al
to
the
cu
be
of
the
wind
sp
ee
d.
T
her
e
i
s
a
wind
s
pee
d
f
ro
m
wh
ic
h
the
m
echan
ic
al
power
t
ran
s
m
itted
by
the
tur
bin
e
is
gr
ea
te
r
tha
n
the
no
m
inal
po
we
r
of
the
el
ect
rical
m
achi
ne.
Va
rio
us
sy
stem
el
e
m
ents
are
siz
ed
accor
ding
to
the
loads
,
ro
ta
ti
on sp
ee
d and
powe
rs
c
orrespo
nd
i
ng
t
o
this win
d
s
pee
d. These elem
ents cannot sup
port v
al
ues
e
xce
edin
g
certai
n
lim
i
ts.
It
is,
therefore
,
necessary
to
lim
i
t
the
po
we
r
captur
e
d
by
the
tur
bin
e
wh
e
n
the
wind
bec
om
es
too
stron
g [22].
3.2.1.
Pitch
angle co
nt
r
ol b
ase
d on PI
c
on
t
rolle
r
The
value
of
t
he
pitch
an
gle
β
var
ie
s
betwe
en
0
°
an
d
90°
to
lim
it
the
m
echan
ic
al
pow
er
at
it
s
rated
value.
We
will
us
e
a
m
et
ho
d
base
d
on
the
c
on
t
ro
l
of
the
ge
ner
at
or
po
wer.
A
s
descr
i
bed
in
Fig
ur
e
6,
t
wo
P
I
con
t
ro
ll
ers
we
re
us
e
d.
T
he
f
irst
PI
to
ge
ne
rate
the
ref
e
re
nce
pitch
a
ng
l
e
βref
an
d
the
second
re
gu
la
tor
to
con
t
ro
l
t
he
er
r
or
betwee
n
βr
ef
an
d
β.
T
he
sam
e
app
ro
ac
h
is
fo
ll
ow
e
d
to
co
ntro
l
t
he
Pi
tc
h
an
gle
us
in
g
RS
T
con
t
ro
ll
ers
.
Figure
6. Pit
ch
an
gle contr
ol
us
in
g
a
PI co
nt
ro
ll
er
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Desig
n a
nd inv
est
iga
ti
ons
of
MPPT str
ategi
es for a wi
nd e
ner
gy
c
on
ver
sion syst
em
...
(
B
ou
c
haib
Rac
he
d)
4775
3.2.2.
Pitch
angle co
nt
r
ol b
ase
d on
fu
z
z
y
log
ic
c
ontr
oller
In
this
par
t,
t
he
fu
zzy
lo
gic
m
et
ho
d
without
the
wi
nd
s
pe
ed
m
easur
em
ent,
as
s
how
n
in
F
ig
ur
e
7,
is
us
ed
.
T
he
r
ot
at
ion
sp
ee
d
of
the
ge
ner
at
or,
the
act
ive
po
w
er
t
rack
i
ng
er
r
or
a
nd
the
var
i
at
ion
of
the
t
ra
ckin
g
error are t
he
i
nputs
of the
con
trol.
T
he o
utpu
t of the c
ontr
ol is the
ref
e
renc
e p
it
ch
a
ngle
β
ref
.
Figure
7. Pit
ch
angle co
ntr
ol
us
in
g
F
LC
3.3.
Rotor
si
de
co
nv
er
ter c
on
t
rol
The p
urpose
of co
ntr
olli
ng
t
his con
ver
te
r
is to c
on
t
ro
l t
he
st
at
or
'
s pow
e
rs.
3.3.1.
Vec
to
r
c
on
t
rol
of DFI
G
The
t
wo
phase
s
m
od
el
li
ng
of
the
m
achine
i
s
us
e
d.
T
he
re
fer
e
nce
fr
am
e
(d
-
q)
is
ori
ent
ed
to
al
ig
n
the ax
is
d with
the stat
or f
l
ux [15,
16, 23,
24
]
.
Th
e
(3) bec
ome
s:
0
S
q
S
S
d
a
nd
0
sd
sq
s
sd
s
V
VV
(8)
So
,
w
e
can
wri
te
:
0
Sd
S
Sd
rd
Sq
S
Sq
rq
L
I
M
I
L
I
M
I
(9)
The
a
dju
stm
e
nt
of
the
eq
ua
ti
on
s
(4)
a
nd
(9)
t
o
the
c
ho
s
en
a
xis
sy
stem
and
th
e
si
m
plifyi
ng
assum
ption
c
onside
red in t
his
case
giv
es:
S
S
r
q
S
S
S
S
S
r
d
SS
M
P
V
I
L
V
V
M
QI
LL
an
d
r
S
r
q
S
S
r
r
d
S
M
P
g
V
I
L
VM
Q
g
I
L
(
10)
Th
us
, p
ow
e
r
c
on
t
ro
l i
s
obta
in
ed by ad
justi
ng
the
ro
t
or
c
urre
nts
d
a
nd q
:
rd
r
rq
r
r
I
r
r
r
q
r
r
d
r
d
r
q
r
r
q
r
r
r
d
s
d
s
dI
v
R
I
L
I
L
dt
dI
M
v
R
L
I
L
L
d
t
(
11)
If
F is t
he
m
odel
o
f
the
proces
s,
the
curre
nt c
on
t
ro
l l
oop
is
s
how
n
in
(1
2)
a
nd (1
3)
.
()
(
)
)
()
()
(
S
rr
r
s
s
r
r
d
q
rq
r
rd
Is
Is
VM
R
L
sL
L
V
s
V
Fs
s
(12)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
5
,
Oct
ob
e
r 2
020
:
47
70
-
47
81
4776
a
nd
d
s
s
S
Sr
q
s
M
eg
L
VR
e
M
(13)
The
st
ru
ct
ur
e
of
vector co
ntr
ol
f
or RSC i
s
il
lustrate
d i
n Fi
gure
8
.
Figure
8. Bl
oc
k diag
ram
o
f
th
e ind
i
rect vect
or co
ntr
ol for
RSC
3.3.2.
PI
c
ontr
oller
synthe
sis
f
or RSC
This
ty
pe
of
re
gu
la
to
rs
is
a
c
om
bin
at
ion
of
both
pro
portio
na
l
an
d
inte
gr
al
act
ion
s.
It
has
the
e
ff
ect
of
i
m
pr
ovin
g
sim
ultaneo
us
ly
the
ste
ady
an
d
the
transie
nt
sta
te
s
[25].
T
o
c
on
t
ro
l
the
m
achin
e
side
co
nverte
r
by
a
PI
c
ontrolle
r
, we
hav
e
chose
n
the
str
uctu
re
sh
ow
n
in
Fi
gur
e 9
.
Figure
9. A
ty
pi
cal
PI
c
on
tr
oll
er s
tr
uctu
re
()
S
r
s
s
r
VM
R
L
L
F
sL
s
(14)
The
tra
nsfer
fu
nction o
f
a
PI r
egu
la
to
r
is
gi
ve
n by the
f
ollo
wing e
qu
at
io
n:
()
i
p
k
C
s
k
s
(
15)
The o
pen lo
op
trans
fer
f
un
ct
io
n
is
giv
e
n by:
)
(
)
(
(
)
.
(
)
)
(
i
p
s
sr
r
r
p
VM
s
LL
k
k
G
s
C
s
F
s
R
s
k
s
L
(
16)
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Desig
n a
nd inv
est
iga
ti
ons
of
MPPT str
ategi
es for a wi
nd e
ner
gy
c
on
ver
sion syst
em
...
(
B
ou
c
haib
Rac
he
d)
4777
To
c
om
pen
sat
e
the ze
ro of th
e
tran
s
fer f
unct
ion we
cho
os
e:
i
r
p
r
L
k
R
k
1
()
r
Gs
s
and
sr
p
s
r
LL
VM
k
(
17)
The
cl
os
ed
lo
op tra
nsfer
f
un
ct
ion
ca
n be e
xpr
essed
a
s:
1
()
1
r
Hs
s
(
18)
r
desi
gn
at
es t
he t
i
m
e con
sta
nt
of the
syst
em
.
So
the
term
s
p
k
an
d
i
k
are
g
i
ven by
:
p
r
i
r
sr
S
sr
S
LL
VM
L
V
k
k
R
M
(19)
3.3.3.
RS
T
co
nt
r
oller
sy
n
the
sis
f
or RSC
T
he
c
hosen
contr
oller m
od
ul
e is as
fo
ll
ows
[20]:
10
0
32
3
2
1
0
10
2
2
1
0
()
()
()
()
()
A
p
a
p
a
B
p
b
D
p
d
p
d
p
d
p
d
R
p
r
p
r
S
p
s
p
s
p
s
(
20)
In
acc
orda
nce
with
the
strat
e
gy
of
placi
ng
r
obus
t
po
le
s
,
th
e
de
gr
ee
of
the
poly
no
m
ia
l
D
requires
us
to p
la
ce t
hr
ee
poles. T
his
poly
no
m
ia
l i
s f
irst
deco
m
po
se
d
i
nt
o
a
pro
du
ct
of
two p
olyno
m
ials C and F
.
2
11
.
(
)
(
)
cf
D
C
F
p
p
TT
(
21)
Let
'
s p
ut
1
c
c
P
T
and
1
f
f
P
T
2
5
1
3
5
1
c
c
sr
ca
c
s
f
r
LL
P
T
P
LM
T
P
T
L
(
22)
To
s
pee
d up th
e syst
e
m
, w
e a
re taki
ng t
he f
ol
lowing m
easur
es
:
2
(
)
(
1
5
)
5
aa
D
p
p
PP
(
23)
Fr
om
equ
at
io
ns
(20)
an
d
(
23)
,
we
deduce
th
e
coeffic
ie
nts
of
t
he
poly
no
m
ia
l
D
wh
ic
h
are
relat
ed
t
o
the
coeffic
ie
nt
s
of
R
an
d
S
by
Syl
vester'
s
m
at
rix.
In
this
way,
we
ca
n
der
i
ve
the
pa
r
a
m
et
ers
of
the
RST
con
t
ro
ll
er i
n
t
he
foll
ow
i
ng w
a
y:
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
2088
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
10
, No
.
5
,
Oct
ob
e
r 2
020
:
47
70
-
47
81
4778
3
3
1
2
2
1
2
2
1
1
1
1
1
0
1
1
0
1
0
1
1
0
1
0
0
0
0
0
1
d
d
a
s
s
a
d
d
a
s
s
a
d
a
s
d
a
s
b
r
r
b
d
d
b
r
r
b
Tr
(24)
3.3.4. F
uz
z
y
log
ic
contr
oller
sy
n
th
esi
s
fo
r
RSC
By
u
sin
g
e
qu
at
ion
10, t
he rot
or c
urren
t
ref
e
re
nces ca
n be
deri
ved
a
s foll
ows
:
s
sd
I
s
sd
I
s
M
L
re
f
re
f
s
P
rq
M
L
re
f
re
f
s
Q
rd
MV
s
a
nd
r
e
f
r
e
f
s
e
m
s
PT
ω
(
25)
w
he
re
r
ef
em
T
is t
he
el
ect
ro
m
agn
et
ic
r
efe
ren
ce
to
rqu
e d
e
du
ce
d fro
m
the MPPT
strat
egy.
T
he FL
C co
ntro
l
structu
re
for
R
SC is de
picte
d i
n
Fig
ure
10
.
Figure
10. FL
C C
on
t
ro
ll
er
struct
ur
e
f
or
R
S
C
4.
RESU
LT
S
AND A
N
ALYSIS
In
t
his
sect
io
n,
sim
ulatio
n
r
esults
are
pr
e
s
ented
for
the
stud
ie
d
syst
em
.
The
pro
pose
d
m
od
el
is
i
m
ple
m
ented
in
MAT
LAB/Si
m
ul
ink
.
T
o
s
how
t
he
pe
rfo
r
m
ance
of
t
he
c
on
t
ro
l
strat
e
gies
stud
ie
d
in
thi
s
wor
k,
nam
ely
FLC,
RST
co
ntr
oller
,
an
d
PI
co
ntr
ol
le
r,
we
ha
ve
pe
rfor
m
ed
se
veral
te
sts.
The
se
te
sts
are
ca
rr
ie
d
ou
t
unde
r
the
f
ollo
wing
co
ndit
ions:
Po
w
er
set
point
var
ia
ti
on;
S
peed
va
riat
ion
:
at
t
=
8
[s],
→
2
:1
50
[rd/s]
→
17
0
[rd/s]
;
DF
I
G
pa
ram
et
er
var
ia
ti
on
:
t
=
12
[s
]
,
Rr
→
2*
Rr.
The
f
ollow
i
ng
WECS
pa
ra
m
et
ers
are
us
e
d
f
or
a
si
m
ulati
on
[
21,
6]:
Rs=
0.
012Ω;
Rr=
0.0
21
Ω;
Ls=
0.0
137H;
Lr=
0.0
136H;
M=
0.0
135H;
J=0
.17
5
kg.m
²
;
K=0.0
024 N/
rd/
s; P=2.
4.1.
Ref
ere
nc
e tr
ackin
g
To
bette
r
ap
pr
eci
at
e
the
diffe
ren
ces
bet
wee
n
the
pro
posed
con
t
ro
ls
,
th
e
first
series
of
sim
ula
ti
on
s
are
carried
out
to
te
st
the
perfor
m
ance
in
ref
e
r
ence
tracki
ng.
The
P
ow
e
r
co
eff
ic
ie
nt
with
MPPT
is
il
lustrate
d
in
F
igures
11
a
nd
12.
T
he
sp
ee
d
of
the
ge
ner
at
or
is
sho
wn
in
Figure
s
13
-
15.
Th
e
sta
tor
po
wer
s
are
de
pic
te
d
in
F
igure
s
16
-
18
.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N: 20
88
-
8708
Desig
n a
nd inv
est
iga
ti
ons
of
MPPT str
ategi
es for a wi
nd e
ner
gy
c
on
ver
sion syst
em
...
(
B
ou
c
haib
Rac
he
d)
4779
Figure
11. Po
w
er c
oeffici
ent
with MPP
T
(P
I
)
Figure
12. Po
w
er c
oeffici
ent
with MPP
T
(FLC
)
Figure
13.S
pee
d respo
ns
e
(PI)
Figure
14. Spe
ed resp
onse
(R
ST)
Figure
15. Spe
ed resp
onse
(FLC
)
Figure
16. Po
w
er r
e
spo
ns
e
(P
I
)
Figure
17. Po
w
er r
e
spo
ns
e
(RST)
Figure
18. Po
w
er r
e
spo
ns
e
(FLC
)
MPPT
co
ntr
ol
le
rs
track
the
op
ti
m
u
m
power
point
at
var
yi
ng
wind
sp
ee
ds
,
kee
ping
the
powe
r
coeffic
ie
nt
at
i
ts
m
axi
m
u
m
value
Cpm
ax
=
0.476.
Ba
se
d
on
the
sim
ulatio
n
res
ults,
the
F
LC
m
et
ho
d
rea
ches
a
ste
ady
sta
te
with
a
ne
glect
ed
er
ror.
T
he
PI
co
ntr
ollers
sh
ow
rem
ark
able
os
ci
ll
at
ions
in
the
ste
ady
sta
te
.
Accor
ding
to
Figure
13
to
15
,
the
resp
onse
ti
m
e
of
the
PI
and
FLC
co
ntr
ol
le
rs
is
bette
r
t
han
the
RST
m
et
hod,
Evaluation Warning : The document was created with Spire.PDF for Python.