In
te
r
n
ation
a
l Jou
rn
al
o
f Po
we
r
Elec
tron
ic
s an
d
D
r
ive S
y
stem
(IJ
PED
S
)
V
o
l.
10, N
o.
4, D
e
c
e
m
ber
201
9,
pp.
2254~
22
62
ISSN: 2088-
8694,
DOI
:
10.11591
/ijpeds.
v10.
i
4.pp2254-2262
2254
Jou
rn
a
l
h
o
me
pa
ge
:
ht
tp:
//i
a
e
score
.
com
/
j
o
u
r
na
l
s
/
i
n
d
e
x
.
p
hp/IJ
PED
S
Intellige
n
t c
o
ntr
o
l of D
C-D
C
converter based on
PID
-
neural netw
o
rk
Hu
ssain
K
.
Kh
le
af,
A
l
i
K. N
ah
ar,
An
sam S
.
Jab
b
a
r
D
e
partm
e
n
t
o
f El
ectri
ca
l
E
n
g
i
neeri
n
g
,
U
n
i
v
ersity of
Techn
o
l
ogy
, Iraq
Art
i
cl
e In
fo
ABSTRACT
A
r
tic
le hist
o
r
y
:
Re
ce
i
v
e
d
A
pr 18,
2
0
1
9
Re
vise
d Ju
l 1
4
,
201
9
A
c
c
e
pte
d
J
u
l
29,
201
9
Th
is
p
aper
i
nt
ro
d
u
ced
a
“
P
I
D-N
N
”
b
a
sed
o
n
P
a
r
t
i
c
l
e
S
w
a
r
m
O
p
timi
zati
on
con
t
ro
l
t
h
at
w
as
a
pplied
to
a
b
oos
t
co
nv
erter
o
p
erati
n
g
in
l
arg
e-si
gnal
do
m
a
ins.
S
im
u
l
ati
on
result
s
have
s
ho
wn
t
h
a
t
th
e
pro
p
o
s
ed
“
P
I
D-N
N
con
t
ro
ller”
cou
l
d
en
hanc
e
th
e
(b
oos
t
con
v
erter)
s
t
artu
p
respo
n
s
e
w
i
t
h
t
h
e
u
s
e
of
f
ew
er
o
n
-
of
f
s
w
itch
op
eratio
ns
c
om
p
a
red
t
o
t
he
C
o
n
v
e
nt
ion
a
l
“PID
con
t
ro
llers
”
.
This
r
esu
l
t
h
a
s
been
o
f
hi
gh
im
port
a
nce
in
p
rac
t
i
ce
for
reducing
th
e
nu
m
b
er
o
f
on
-off
s
w
it
ches
can
e
f
f
ectiv
el
y
d
ecrease
t
h
e
t
r
an
si
en
t
di
st
urb
a
n
ces
a
n
d
l
oss
e
s
d
u
e
t
o
s
w
i
t
c
hi
ng.
S
imu
l
ati
o
n
s
a
l
s
o
p
r
ov
e
th
at
t
h
e
pro
p
o
s
ed
“
P
I
D
-
NN
co
nt
ro
ll
er”
is
capabl
e
o
f
e
f
f
i
ci
ent
l
y,
i
m
p
ro
v
i
ng
r
ejecti
ng
p
o
te
nt
ia
l
di
sturb
a
nc
e
s
t
h
a
t
c
o
u
l
d
ha
p
p
e
n
i
n
th
e
in
pu
t
vo
lta
g
e
.
Moreover,
it
has
been
n
ot
ice
d
t
hat
th
e
o
u
t
put
v
olt
a
ge
i
s
m
o
re
e
f
f
ici
e
ntly
c
o
ntrol
l
ed
w
hen
app
l
y
i
ng
“
P
I
D-N
N
c
ont
ro
l
l
er”.
T
he
r
esu
l
ts
o
f
th
e
s
i
m
u
l
a
ti
on
sh
o
w
the
ef
fici
ency
o
f
t
h
e
s
ugg
est
e
d
al
g
o
rithm
com
p
are
d
w
ith
other
well
-
kn
own
learn
i
n
g
m
eth
ods.
K
eyw
ord
s
:
D
C
–D
C
bo
ost
con
v
erte
r
P
I
D
base
d on a
ne
ural ne
t
w
o
rk
Con
v
e
n
ti
ona
l P
I
D
contro
ller
Pa
rti
c
le sw
a
rm opt
imiza
tio
n
Co
pyri
gh
t © 2
019 In
stit
u
t
e
of Advanced
En
gi
neeri
n
g
an
d
S
c
ien
ce.
All
rights
res
e
rv
ed.
Corres
pon
d
i
n
g
Au
th
or:
Hu
ssain
K.
Kh
l
e
af
,
D
e
pa
rtme
nt
o
f
El
e
c
t
rica
l
Eng
i
ne
eri
ng,
U
n
i
v
ersi
ty o
f
Tech
no
l
o
g
y
,
Ba
gh
d
a
d,
I
ra
q.
Em
ail:
300
68
@u
o
t
ec
h
n
o
lo
g
y
.e
du.i
q
1.
I
N
TR
OD
U
C
TI
O
N
B
y
w
a
y
o
f
t
h
e
in
dus
t
r
i
a
l
de
v
e
lo
pm
en
t,
c
on
t
r
ol
o
b
j
ects
ke
e
p
g
e
t
t
i
ng
m
o
re
c
om
plica
t
ed,
part
icu
l
arl
y
for
the
u
n
k
n
o
w
n
pa
ram
e
ter
s
o
r
slow
v
ariat
i
o
n
s
i
n
lar
g
e
dela
y,
t
i
m
e-
c
h
angi
ng
,
no
n-l
i
n
e
a
r
c
o
m
p
l
i
c
a
t
ed
syste
m
s,
w
ith
r
and
o
m
i
n
t
e
rfe
r
enc
e
o
r
de
la
y. H
ow
eve
r
,
The P
r
opor
t
i
o
n
a
l
I
n
t
egr
a
l
D
e
riva
t
i
ve
(
P
I
D
)
c
ont
r
o
l
has
si
m
p
le
s
tr
uct
u
re
a
nd
li
nea
r
b
eha
v
io
ur.
Mo
re
over,
i
t
g
i
ve
s
acc
ep
tab
l
e
p
e
r
f
orm
a
nc
e
fo
r
seve
ral
i
n
dustria
l
a
p
pli
c
at
ion
s
[
1].
Th
e
“P
ID
c
o
n
t
r
ol
l
e
r”
i
s
on
e
o
f
t
h
e
t
ra
dit
i
on
a
l
c
ontro
l
l
e
rs
w
hic
h
a
re
u
su
al
l
y
u
se
d
i
n
m
an
y
dri
v
e
sys
t
e
m
s.
H
ow
ever
,
it
i
s
s
l
u
g
g
ish
re
s
ponse
du
e
t
o
s
ud
de
n
c
han
g
e
i
n
p
o
w
e
r
a
nd
t
h
e
se
ns
it
ivi
t
y
t
o
con
t
ro
l
l
er
g
a
i
n
s
[
2]
t
he
c
o
n
v
e
nt
iona
l
“
P
I
D
c
o
n
tro
l
”
para
me
t
e
rs
wil
l
not
c
h
a
ng
e
po
st
c
omp
l
e
tio
n
,
w
hi
c
h
w
il
l
re
su
lt
i
n
pa
ra
me
t
e
r
ch
a
n
g
e
s
of
c
o
n
t
r
oll
e
d
o
b
j
ect
s
th
at
can
’t
b
e
t
ra
ce
d
i
n
r
ea
l
ti
me
,
c
a
n
not
s
ati
s
fy
g
ro
w
i
ng
nee
d
s of
c
o
n
tr
ol
q
ua
li
t
y
i
n
th
e
proc
es
s
of pr
o
d
u
c
t
i
o
n,
t
his is
w
hy,
s
c
h
olar
s
i
m
pro
v
e
d
a
r
a
nge
o
f
e
n
hanc
e
m
e
n
t
s
f
o
r
Co
nv
en
ti
on
al
“
PID
c
o
ntro
ll
er”
,
e
sp
e
c
ia
l
l
y
i
n
2
a
s
p
ect
s:
o
n
e
o
f
the
m
i
s
impr
ovi
n
g
s
truc
ture
,
w
h
ic
h
is,
varia
b
l
e
s
truc
tur
e
c
on
tro
l
[
3]
.
Rece
n
tly,
t
h
e
use
o
f
i
n
t
e
lli
ge
n
t
c
o
n
t
ro
l
li
ke
N
eura
l
N
e
tw
ork,
N
e
u
ro
F
uz
zy
a
n
d
f
u
zzy
c
on
t
r
ol
,
t
h
e
fu
z
z
y
l
ogi
c
c
o
n
t
ro
l
ca
n
a
n
swe
r
t
h
e
u
nce
r
t
a
in
t
y
p
r
o
bl
e
m
a
nd
su
d
d
e
n
d
i
s
t
u
rba
n
c
e
,
on
t
he
ot
her
ha
n
d
its
d
e
s
ig
n
de
pe
n
d
s
o
n
the
e
x
peri
enc
e
w
h
i
c
h
s
o
m
etim
es
i
s
n
o
t
a
va
ila
b
l
e
f
o
r
som
e
s
ys
tem
s
[
1].
In
add
i
tio
n,
i
t
c
a
n
a
d
d
c
o
n
t
r
o
l
n
o
n
-li
n
ear
s
ys
tem
s
,
w
h
ich
w
o
uld
be
h
a
r
d
o
r
n
o
t
p
o
s
s
i
b
l
e
t
o
m
o
d
e
l
ma
them
at
i
c
al
l
y
.
A
n
a
rt
ific
ial
ne
ura
l
n
etw
o
rk
c
ou
l
d
b
e
ap
p
lie
d
t
o
a
w
i
de
r
ange
o
f
task
s
l
i
ke
s
ig
na
l
p
r
oce
ssi
ng,
p
a
tt
e
r
n
re
c
ogni
ti
o
n
,
f
u
n
c
ti
o
n
a
p
p
r
oxi
mat
i
on
,
an
d
cl
a
s
si
fi
ca
t
i
on
.
U
s
ual
l
y,
t
her
e
a
r
e
t
w
o
opera
t
i
o
n
s
w
h
en
ut
iliz
i
n
g ne
u
r
al
ne
t
w
o
r
k
s for
con
t
ro
l,
t
h
o
se
o
per
a
t
i
o
n
s
are
[4]
, S
y
st
e
m
i
de
nt
ifi
c
a
t
io
n an
d
Con
t
ro
l de
sig
n
[4].
The
P
S
O
a
l
go
r
i
t
h
m
c
a
n
l
e
a
d
t
o
a
h
i
ghe
r
q
u
a
l
i
t
y
s
o
lu
t
i
o
n
w
i
t
h
t
i
m
e
a
nd
s
ecur
e
c
on
verge
n
ce
i
n
c
o
m
p
a
r
i
so
n
w
ith
o
t
h
er
s
to
chas
tic
m
e
t
ho
ds
[
5].
In
a
dd
i
t
i
on,
M
e
t
a
H
e
ur
i
s
tic
me
t
hods
h
av
e
h
a
d
a
pr
ofo
u
nd
e
ffec
t
o
n
op
tim
iza
t
i
o
n
i
n
m
o
d
i
f
ied
e
n
g
i
nee
r
i
n
g
str
e
a
m
s.
T
he
e
ffic
i
e
nc
y
o
f
t
he
se
a
lgor
it
hms
is
i
m
p
o
r
ta
nt
a
s
t
h
e
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
S
S
N
:
2088-
86
94
I
n
te
ll
i
g
en
t co
nt
ro
l of D
C
-
D
C c
onve
r
te
r
b
a
se
d on
PI
D
-
ne
ural
ne
t
w
ork
(
H
ussa
in K.
K
h
le
a
f
)
2
255
ha
r
d
w
a
r
e
a
p
p
lic
ati
o
n
of
t
hese
a
lg
or
i
t
h
ms
f
or
v
ar
ious
e
n
g
i
ne
er
i
n
g
app
l
i
c
at
i
ons
i
s
to be
ca
rried
o
ut
[
6].
PS
O,
a
s
one
o
f
t
h
e
me
t
a
-
h
eur
i
s
t
ic
m
et
ho
ds
o
f
opt
im
iz
ati
o
n,
r
e
l
i
e
s
o
n
t
h
e
t
h
oro
u
g
h
sea
r
ch
a
re
a's
i
d
eal
s
o
l
u
t
i
o
n
ba
sed
o
n
t
h
e
e
x
c
h
ange
o
f
e
x
p
e
ri
en
c
e
s
a
m
o
n
g
t
h
e
pop
ul
a
t
i
on'
s
p
a
rti
c
l
e
s
.
[
7
]
.
M
or
e
over
,
P
S
O
is a sw
ar
m-
ba
sed m
e
ta-
he
ur
ist
i
c
a
l
gor
i
t
hm
w
i
t
h
s
om
e
i
n
t
r
ins
i
c
i
n
co
n
v
en
ie
nc
es
t
ha
t
w
e
a
k
l
o
cal
s
earch
a
nd
s
lo
w
c
o
nv
e
r
g
e
n
c
e
ra
te
a
nd
tr
a
p
pin
g
in lo
c
a
l o
p
tim
um w
hen so
l
v
i
ng c
o
m
p
lica
t
e
d
m
ul
ti
-
m
odal
is
s
u
es
[
8]
.
A
l
s
o
,
one o
f th
e
m
o
s
t
p
r
o
b
l
em
a
r
a
nge
o
f
D
C
-
v
ol
ta
ge
o
f
t
h
ese
so
ur
c
e
s
ha
ve
l
ow
o
u
t
p
u
t
l
e
v
e
l
s,
u
n
sta
b
le
,
t
h
e
set
tli
n
g
t
i
m
e
is
t
oo
ta
ll
m
or
eover
.
The
r
e
ar
e
als
o
m
a
ny
s
itua
t
i
o
n
s
w
her
e
l
o
ss
m
a
y
be
e
i
t
he
r
u
n
n
o
t
i
c
ea
b
l
e
or
a
cce
pta
b
le
[
9]
c
om
par
e
w
it
h
o
t
he
r
t
e
chni
qu
es
like
(P
I
D
-NN
)
,
t
h
ere
f
or
e
th
e
s
u
g
g
es
te
d
ne
w
co
nf
ig
ur
a
t
i
o
n
s
o
f
t
h
e
i
n
te
ll
ige
n
t
(P
ID-NN)
a
nd
op
t
i
mize
d
by
u
s
in
g
P
S
O
meth
o
d
a
p
p
lie
d
to
t
he
“
Bo
os
t
c
o
nver
t
er
”
due
t
o
sever
a
l
r
e
a
s
ons
a
nd
w
i
l
l
b
e
im
pl
e
m
e
n
te
d
t
o
b
e
t
ter
op
tim
iz
e
b
oos
t
c
o
nve
r
t
e
r
s
tab
i
lit
y
c
o
mpa
r
ed
t
o
t
h
e
co
nve
n
t
i
ona
l
“P
I
D
”
.
H
igh
vo
l
t
a
g
e
incr
ease
s
an
d
d
ecr
ease
d
stead
y-
sta
t
e
er
ror
a
r
e
the
adva
n
t
a
g
e
s
o
f
t
h
e
p
r
o
posed
sc
h
eme [
1
0
]
.
2.
BOO
S
T
CONVE
R
TER
R
e
ce
nt
e
lectr
o
nic
sys
t
em
s
in
vo
lve
h
i
g
h
ly
e
ffic
i
e
n
t
r
esour
c
e
s.
B
y
r
e
plac
in
g
d
i
ode
w
it
h
MO
S
F
ET
[11
]
,
t
h
e
eff
e
c
t
i
v
e
n
e
ss
of
(
DC
t
o
D
C
c
onvert
e
r
s)
s
u
c
h
as
t
h
e
e
n
erg
y
s
o
u
r
ce
of
e
l
e
c
t
ro
nic
e
q
ui
p
m
e
n
t
can
b
e
im
pro
v
ed.
(
D
C
/
DC
c
o
n
v
erte
r
)
can
b
e
u
s
e
d
f
or
d
iffere
nt
a
p
p
lic
a
t
i
o
n
s
suc
h
a
s
vo
lta
ge
o
u
t
pu
t
co
n
t
rol,
e
le
c
t
ric
ve
h
i
c
l
e
e
n
er
g
y
s
t
o
r
a
ge
s
ys
te
m
,
r
enew
a
b
le
pow
er
s
tor
a
ge
s
ys
tem
,
[
12]
a
“
D
C
-
D
C
bo
ost
”
C
on
ver
t
e
r
s
w
i
tc
h
i
n
g
w
a
s
in
te
n
d
e
d
t
o
s
how
t
he
s
c
h
e
m
e
pr
op
ose
d
.
The
b
o
o
s
t
ene
r
gy
pha
s
e
ou
tp
ut
c
ur
r
e
n
t
i
s
e
i
t
h
er
c
ons
tan
t
o
r
n
on-
pu
ls
ing.
[
1
3
]
C
ons
ide
r
i
ng the
“
D
C–D
C
b
oos
t” as
c
o
n
v
er
ter
c
i
r
c
u
i
t
w
h
ic
h is dep
ict
e
d i
n
F
i
g
ur
e
1.
T
hr
ou
gh th
e
in
terva
l
,
w
h
en
t
he
Q
swit
c
h
is
o
ff,
(dio
de)
D
con
d
u
ct
s t
h
e
c
u
rr
en
t
i
of
ind
u
c
tor L on
the way to
t
he ca
p
a
c
it
or
C
a
nd
the
l
o
a
d
R
.
F
o
r
t
h
e
dur
a
t
io
n
of
t
he
i
n
t
e
r
va
l,
w
hen
the
Q
sw
itch
is
o
n
the
d
i
o
d
e
D
o
p
e
n
s
a
n
d
t
h
e
C
di
schar
g
e
s
t
hr
oug
h
t
h
e
R
a
s
sho
w
n
in
F
igur
e
1
.
T
he
c
onver
t
e
r
t
r
a
nsm
i
t
s
t
he
e
ne
r
g
y
betw
e
e
n
t
h
e
i
n
p
u
t
a
n
d
the
o
u
t
p
u
t
w
i
t
h
the
use
of
t
h
e
i
nduc
tor
.
A
"
D
C
-
D
C
B
oos
t"
c
on
ver
t
e
r
de
si
gn
ins
t
a
n
ce
is
a
vai
l
a
b
le
t
o
ac
hi
e
v
e
the
ef
fic
i
e
n
cy
nee
de
d f
o
r
t
h
e imple
m
e
n
ta
tio
n lis
te
d be
l
o
w
.
The
s
i
m
u
l
a
t
i
on
pac
ka
ge
M
A
T
L
A
B
/
S
i
m
u
lin
k w
a
s
use
d
t
o
che
c
k
t
he
t
he
or
e
tic
a
l
p
r
o
jec
t
ions
[
1
3
]
.
F
i
gur
e
1.
B
o
o
s
t
c
on
ve
r
t
er
.
I
n
F
ig
ur
e
1,
t
r
a
nsfe
r
fu
nc
ti
o
n
t
ha
t
is
d
er
i
v
ed
w
i
t
h
u
se
in
g
of
t
he
a
p
p
r
o
a
c
h
of
s
ta
n
d
ar
d
sta
t
e
s
p
ac
e
a
v
er
a
g
i
n
g,
w
h
e
r
e
t
h
e
c
i
r
cui
t
s
f
or
t
he
t
w
o
a
oper
a
t
i
o
n
al
m
ode
O
N
a
nd
O
FF
m
ode
f
or
t
he
c
o
nver
t
er
a
r
e
m
odel
l
e
d
i
n
t
h
e
fol
l
o
w
i
ng
w
a
y:
(
1
)
w
h
e
r
e:
u
,
y
i
np
u
t
,
ou
t
p
u
t
v
olta
ge
(
V
i
n
,
V
o)
a
nd
x
s
t
ate
var
i
a
b
l
e.
The
p
o
st
i
s
m
odel
l
e
d
,
t
h
e
a
v
er
a
g
e
of
t
h
o
se
t
w
o
m
ode
s
o
v
er
one
s
wi
tc
hin
g
p
erio
d T
is
c
a
l
c
u
l
a
te
d [1
4].
3.
THE
PRO
P
ORTIO
N
AL
I
NTEGRAL
D
ERIVAT
IVE (PID)
C
O
NTROL
PRINCIPLE
“
C
on
t
r
o
l
le
r
PID
"
is
o
n
e
o
f
t
h
e
trad
i
t
i
o
n
a
l
c
ontro
l
l
er
s
of
m
any
d
r
i
ve
s
yste
m
s
.
[2]
.
T
he
C
on
ve
nti
ona
l
“
c
ontro
l
l
e
r
P
ID”
whi
c
h
is
d
e
p
ic
te
d
in
F
ig
u
r
e
2
i
n
c
l
u
d
es
3
d
isti
nc
t
a
c
tio
n
pa
ram
e
te
rs,
whic
h
are
t
h
e
In
tegr
a
l
,
D
e
r
i
vat
i
v
e
an
d
P
r
opor
t
i
o
n
a
l
.
Th
ose
3
pa
r
a
m
e
te
r
s
f
r
o
m
th
e
com
p
u
t
a
t
i
o
n
of
“
P
I
D
”
[
15
]
.
C
on
ve
n
t
i
o
nal
“P
ID
c
o
n
t
r
o
ller
”
i
n
t
h
e
pr
oce
s
s
of
p
r
o
duc
t
i
o
n
i
s
t
h
e
m
o
st
o
f
t
en
u
t
i
l
iz
e
d
c
on
tr
o
l
a
ppr
oac
h
,
c
o
m
m
only
im
p
l
e
m
ented
in
c
hem
i
ca
l
,
m
ac
hi
n
e
r
y
,
m
e
ta
ll
ur
g
y
a
n
d
e
xtr
a
i
ndus
tr
ies
[
1
6]
.
I
n
o
t
her
s
i
d
e
,
the
simu
la
te
d
of
c
o
n
t
r
o
l
s
y
stem
i
s
Co
nv
en
ti
o
n
a
l
“
P
ID
c
o
n
t
r
ol
l
e
r”
b
ased
o
n
t
h
e
d
e
vi
at
io
n
d
i
f
f
eren
t
i
a
l
(
D
)
,
p
r
o
p
o
r
t
i
o
n
(
P
)
a
n
d
i
n
t
e
g
r
a
l
(
I
)
a
r
e
t
h
e
m
o
st
c
om
m
onl
y
i
m
plem
e
n
te
d
contr
o
l
l
e
r
a
ut
om
atic
[
3]
.
F
i
gur
e
2.
C
o
n
v
e
nti
ona
l
co
n
t
r
o
ll
e
r
P
I
D
p
r
i
nci
p
le.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2088-
8694
I
nt
J
P
ow
Elec
& Dr
i
S
y
st V
ol.
10,
N
o.
4
, Dec
201
9 :
2
2
5
4
– 2
262
2
256
The
pr
op
ort
i
on
a
l
v
a
l
ue
r
egu
l
a
t
es
t
he
r
eac
ti
o
n
t
o
the
i
n
sta
n
t
a
ne
ous
error;
t
he
i
nteg
ra
l
va
l
u
e
regu
la
tes
t
h
e
re
a
c
t
i
on
a
c
c
o
rdi
n
g
t
o
t
he
s
u
m
ma
t
i
on
o
f
re
ce
nt
e
rro
rs
a
nd
t
h
e
deri
vat
i
v
e
va
l
u
e
re
gu
l
a
t
e
s
t
h
e
r
eac
tio
n
ac
cord
in
g
t
o
t
h
e
r
ate
of
t
he
e
rror
cha
nge
s.
By
t
hro
u
g
h
t
he
m
,
y
(t)
is
t
he
e
ffec
t
i
ve ou
t
put val
ue
w
h
ile
,
r(t
)
i
s
t
h
e
gi
ve
n va
lue
. T
h
e
n,
e
(t)=r(t)
-
y(t), Conve
n
t
i
o
n
a
l
“P
I
D
c
ont
r
o
l
l
er
” w
h
en the
e
qua
t
i
o
n
is
:
(
2
)
wher
e,
and
are
t
he
in
t
e
g
ra
l
and
der
i
va
t
i
ve
t
i
m
e c
onst
a
nt
s
respec
tive
l
y
[1
5].
3.1.
Di
s
c
ret
e rep
re
se
nt
a
t
ion
of th
e
C
onven
t
ional
PID
con
t
rol
l
e
r
In
t
he
p
as
t
yea
r
s,
a
nalo
g
c
o
nt
rol
l
e
r
s
w
e
r
e
t
y
p
ica
l
ly
r
ep
lac
e
d
w
i
t
h
dig
i
t
a
l
con
t
ro
l
l
ers
t
h
a
t
t
he
o
u
t
pu
t
s
an
d
i
n
put
s
a
r
e
ch
arac
t
e
ri
z
e
d
at
d
i
s
c
r
et
e
ti
me
i
n
s
t
a
n
c
e
s
.
M
o
re
o
v
er
,
t
h
e
d
i
gi
tal
c
o
ntr
o
l
l
ers
a
s
possi
b
l
e
t
o
b
e
in
the
f
o
r
m
o
f
d
i
g
i
tal
c
o
m
p
u
t
e
r
,
dig
i
ta
l
circui
ts
o
r
m
i
c
r
o
p
r
oce
s
so
r
s
.
The
discre
te-
time
C
o
nve
n
t
io
na
l
“PID
co
n
t
roll
e
r
”
me
an
s
di
sc
ret
i
z
ing
t
h
e
co
nti
nuo
u
s
-
t
i
m
e
t
o
d
i
s
cre
t
e
-
tim
e.
T
he
d
iscre
t
e
C
o
n
v
e
n
t
i
ona
l
c
a
n
be
re
prese
n
t
e
d as
f
o
l
l
o
w
s
:
C
ons
ide
r
t
he
c
on
t
i
n
u
o
u
s
t
i
m
e
expr
ess
i
on
of
a
C
o
n
v
e
n
t
i
ona
l
“
P
ID
controller”
i
n
(2),
If
the
sapling
ti
m
e
o
f
d
i
scr
e
tiza
t
i
on
is
Δ
.
The
n
t
he
i
n
t
e
g
ra
l
term
a
bove
c
a
n
b
e
m
e
a
s
ur
ed
“
d
i
sc
rete”
via
a
t
ra
p
ezoi
d
a
l
appr
ox
im
at
ion
:
≅
∑
∆
(
3
)
wher
e,
t
h
e
e
r
r
o
r
o
f
t
h
e
d
i
s
c
r
e
t
e
t
i
m
e
s
y
s
t
e
m
a
t
t
h
e
i-th
sa
mpli
n
g
i
nsta
n
t
.
Similar
l
y,
t
h
e
b
ackw
a
rd-fin
ite
di
ffe
re
nce
a
ppr
ox
i
m
at
io
n o
f
t
he firs
t
-
o
rder
d
eriva
t
i
v
e
:
l
im
∆
→
∆
∆
(
4
)
Ap
pl
ing
th
e
ba
ck
wa
rd
-f
in
i
t
e
d
if
fe
re
n
c
e
app
r
ox
i
m
at
io
n
to
t
h
e
d
i
s
cre
t
e
t
i
me
d
eri
v
at
i
v
e
t
e
rm
i
n
(2
)
bec
o
me
s:
≅
∆
(
5
)
The
r
ef
ore,
the
d
isc
r
ete
tim
e
c
o
n
t
ro
l law,
or
“pos
i
t
i
o
n
a
l
a
lg
o
r
i
thm
”
,
c
o
me
s to be
,
∆
∆
(
6
)
A
n
“
i
n
c
r
em
enta
l
a
l
g
o
ri
t
h
m”
o
r
“
v
eloc
ity
a
lg
ori
t
hm”
m
a
y
be
o
b
t
a
i
ne
d
by
s
u
b
t
ra
c
t
i
n
g
(
−1)
fr
om
(
).
1
∆
∑
∆
(
7
)
∆
(
8
)
1
∆
∑
∆
∆
(
9
)
Th
is c
an be
r
e
w
r
itte
n a
s
:
2
(
10)
wher
e,
∆
,
∆
3.2.
N
eu
ral Ne
twor
k c
o
n
t
r
o
l
l
e
r
of
PID
The
dra
w
bac
k
o
f
t
h
e
Co
n
v
e
n
ti
ona
l
“
c
on
trol
l
e
r
P
I
D
”
i
s
no
t
a
ppr
o
p
riate
for
long
t
ime
dela
y
and
n
onl
in
e
a
r
sy
st
em
c
o
n
t
r
ol
[
17
].
W
h
e
re
D
,
P
an
d
I
p
a
ramet
e
rs
t
un
i
ng
i
s
di
ffi
cu
l
t
w
it
h
t
h
e
way
mod
e
rn
c
omp
u
t
e
r
tech
n
o
l
o
gy
has
deve
l
ope
d,
i
n
a
ddi
ti
o
n
t
o
th
e
con
t
ro
l
t
h
e
o
ries
l
i
k
e
fuz
z
y
a
nd
ne
ura
l
n
e
t
w
o
rks.
T
he
a
d
a
pti
v
e
Con
v
e
n
ti
ona
l
“
c
on
tro
l
ler
P
I
D”
b
a
s
e
d
o
n
Bac
k
P
r
opaga
t
i
o
n
(
BP
)
an
d
neura
l
n
e
t
w
o
rk
a
re
d
esi
gne
d
to
c
om
bi
ne
the
tra
d
it
i
ona
l
stra
t
e
gy
of
t
he
C
onve
n
t
i
o
na
l
“
P
ID
c
ontro
l
l
e
r
”
w
ith
u
s
i
n
g
t
he
n
e
u
ral
ne
tw
ork
ha
s
pr
od
uc
e
d
a
no
ve
l
idea,
a
l
s
o
de
si
g
n
t
he
n
ew
c
o
n
tro
l
s
c
h
e
m
e
.
H
ow
e
v
er,
the
a
b
i
l
ity
o
f
se
lf-
l
ea
rni
n
g
of
n
e
u
ra
l
n
e
tw
o
r
k
an
d
Evaluation Warning : The document was created with Spire.PDF for Python.
Int J
P
o
w
E
l
e
c
&
D
ri S
yst
IS
S
N
:
2088-
86
94
Inte
lli
gen
t
co
nt
ro
l
of D
C
-
D
C conv
e
r
t
e
r
b
a
se
d on
PID
-
neur
al
ne
t
w
ork
(H
ussa
in
K. Khle
a
f
)
2
257
Ba
c
k
P
ropa
g
a
t
i
on
ca
n
a
u
tom
a
tica
l
l
y
t
u
n
e
a
nd
u
p
d
at
e
t
h
e p
a
ram
e
te
rs
o
f
rob
u
s
t
C
o
nve
n
t
io
na
l
“
P
ID
c
ontr
o
l
l
er”
,
b
e
ca
use
it
i
s
co
mb
in
ed
b
e
t
w
e
en
t
h
e
m
coul
d
giv
e
g
ood
r
e
s
ul
t
s
.
Fi
g
u
re
3
s
how
n
of
t
he
“
P
I
D
c
o
n
t
rolle
r
”
b
loc
k
dia
g
ra
m
based
on
n
e
u
ra
l ne
tw
ork
[1
8].
F
i
gure
3.
The
n
eura
l netw
ork
based
o
n
con
trol
ler
of
P
ID.
The
pre
s
en
te
d a
d
ap
t
i
v
e
c
on
t
r
ol
l
a
w
us
i
ng ne
ural ne
t
w
o
r
k
te
c
h
n
i
q
ue
i
s for
exa
m
ple
the ne
xt
e
q
u
a
t
i
o
n
:
1
2
1
2
(
11)
wher
e;
,
1
2
a
r
e
the
in
p
u
t
v
e
ct
ors.
q
i
s
t
he
n
e
u
ral
netw
orks
t
ha
t
c
a
n
be
o
bta
i
ne
d
fr
om
the a
c
t
i
v
a
tio
n
func
t
i
o
n
“
F
”:
,
1
(
12)
The
c
ontro
l
l
a
w
of
t
he c
o
n
t
r
o
l
le
r a
c
t
i
on
:
1
(
13)
The control
param
e
ters of
th
e
“
P
I
D
-NN
cont
rol
l
e
r
”
are
V
1
– V
4
[18
]
.
I
n
s
pi
te
o
f
the
B
P
i
s
a
un
ive
r
sal
com
p
u
tin
g
tec
h
niq
u
e
w
h
i
c
h
i
s
w
i
de
ly
u
s
e
d
t
o
d
e
t
er
min
e
t
he
n
e
u
ral
netw
ork
w
e
i
g
hts.
A
lso
h
a
s
th
eir
sh
orta
ge
t
h
a
t
has
lo
n
g
t
ra
i
n
i
n
g
t
i
m
e.
T
o
i
m
p
r
ov
e
t
h
e
co
nv
e
r
g
e
nt
s
p
e
ed
,
th
e
Pa
rti
c
le
S
w
a
rm
O
ptimiza
t
i
o
n
(PS
O
)
algo
rithm
can
b
e
ado
p
te
d
t
o
e
val
u
a
t
e
t
h
e
“
P
ID”
neura
l
n
etw
o
rk
parameters.
3.3.
PID
-
NN
Based
on
“PS
O
”
The
P
S
O
t
e
ch
ni
q
u
e
i
s
f
irstly
p
re
se
n
t
ed
b
y
Eberha
rt
a
n
d
K
enne
dy
i
n
t
h
e
y
ea
r
of
1
99
5
[1
9].
Thi
s
tech
n
i
q
u
e
use
d
a
m
u
lti-va
r
i
able
f
unc
t
i
ona
l
w
i
t
h
m
u
lti-l
o
c
a
l
o
p
t
imum
p
oi
n
t
s.
T
he
f
u
nda
me
nta
l
o
f
t
h
i
s
a
l
go
rith
m
ut
il
ize
th
e
beh
a
vi
o
u
r
o
f
th
e
bi
rd
’s
f
l
o
c
k
.
In
c
onj
un
c
tio
n
w
i
t
h
o
t
h
er
o
p
tim
iz
a
tio
n
m
e
t
h
o
d
s
t
h
e
P
S
O
has
fe
a
t
ures
o
f
fa
st
c
o
n
v
e
r
gence
a
nd
e
a
sy
i
mplem
e
n
t
at
i
on.
T
his
me
t
h
od
d
e
p
end
s
o
n
moni
to
ri
ng
t
he
mo
v
e
men
t
o
f
fl
o
c
k
s
w
h
e
re
‘
b
i
rd
s’,
c
a
l
l
e
d
p
a
rt
i
c
l
e
s,
c
oope
rat
e
w
i
t
h
e
a
c
h
o
t
h
e
r
a
n
d
i
n
t
e
r
a
c
t
t
o
g
e
t
h
e
r
i
n
t
h
e
‘floc
k
’
w
h
ic
h
i
s
cal
le
d
a
sw
arm
.
T
he
m
ov
ing
s
t
e
p
s
of
e
ac
h
b
i
rd
c
h
a
r
ac
t
e
ri
zed
b
y
i
t
s
ve
l
o
cit
y
a
nd
a
n
o
bj
ecti
v
e
w
h
ic
h
form
ed
t
he
f
itne
ss
o
f
t
he
s
w
a
rm
m
oti
o
n
[15].
The
m
ovin
g
s
te
p
of
t
he
p
a
r
t
i
c
l
e
de
pen
d
s
o
n
t
he
o
pt
i
m
a
l
pos
it
io
n
of
t
he
i
nd
iv
i
d
u
a
l
partic
le
(
p
be
s
t
)
an
d
op
t
i
ma
l
pos
i
t
i
on
i
n
t
he
n
e
i
g
h
bour
h
ood
pa
rt
ic
le
(
g
be
st
)
,
w
h
i
c
h
a
r
e
sa
v
e
d
i
n
t
h
e
b
est
gl
ob
al
p
os
itio
n
of
t
h
e
f
lo
ck.
Th
i
s
g
lob
a
l
b
e
s
t
po
si
ti
on
u
s
ed
t
o
mov
e
p
a
r
t
i
c
l
es
i
n
th
e
di
re
ct
i
o
n
o
f
t
h
e
b
e
s
t
pos
it
ion
,
a
nd
upd
a
t
in
g
i
t
s
v
e
loc
i
ty
a
nd
d
i
r
e
c
t
i
on
a
s
req
u
i
r
ed
.
Thi
s
a
pp
ro
ach
l
ead
s
to
g
ui
d
e
a
l
l
part
icles
i
n
t
h
e
s
w
a
rm
t
o
t
h
e
op
tim
um
(
be
st)
va
lue
(p
o
s
i
t
i
o
n)
[
2
0
].
T
he
f
oll
o
w
i
ng
e
qua
t
i
on
def
i
n
e
s
the
st
a
ndard
f
orm
of t
he
P
S
O
algori
t
hm:
1
(
14)
1
1
i
= 1,2,
…
n
(
15)
a
nd
a
r
e
posi
t
i
o
n
a
nd
ve
l
o
c
ity
o
f
par
t
ic
le
i
,
k
d
e
n
o
t
e
s
t
h
e
n
u
m
b
e
r
o
f
t
h
e
i
t
e
r
a
t
i
o
n
,
w
h
i
l
e
,
R
1
a
nd
R
2
a
r
e
rand
om
v
ariab
l
es
t
ha
t
ha
ve
t
h
e
ir
v
a
l
ue
s
e
v
en
ly
d
i
s
tr
ibu
t
e
d
i
n
t
he
r
a
nge
b
e
t
w
e
e
n
(
0
,
1),
W
i
s
t
h
e
in
erti
a
wei
g
h
t
and
C
1
a
n
d
C
2
r
epr
e
sent
t
he
c
og
nit
i
v
e
c
o
e
f
f
i
c
i
e
n
t
an
d
t
h
e
soc
i
a
l
c
oef
f
ic
i
e
nt
.
,
i
s
the
i
n
div
i
dua
l
o
p
t
i
m
a
l
pos
it
io
n
of
p
ar
tic
l
e
i
, and
g
,
i
s t
h
e
op
tima
l
g
lo
bal
p
o
s
i
t
i
on of all
o
f
the
s
w
a
rm
p
a
r
tic
les.
n
i
s t
h
e
num
ber o
f
bir
d
s
(i.
e
.
par
t
i
c
l
e
s).
If
t
h
e
c
on
dit
i
on
(
16
)
i
s
s
a
t
is
fi
e
d
,
t
h
e
p
o
s
iti
o
n
of
t
h
e
p
arti
cl
e
i
s
c
orr
ected
a
c
c
o
rdi
ng
to
(17):
,
(
16)
,
(
17)
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
n
t
J Po
w El
ec &
D
ri S
y
s
t
Vo
l. 1
0
,
No
. 4
,
D
e
c
2
0
19
:
2
254
–
2
262
2
258
wh
ere;
f
i
s
th
e fit
n
ess
fu
nc
ti
o
n
tha
t sh
o
u
l
d
b
e
minim
i
z
e
d
.
I
n
a
d
d
it
io
n,
t
h
e
P
S
O
a
lg
or
it
hm
i
s
use
d
t
o
eva
l
ua
t
e
o
p
t
im
um
v
al
u
e
s
o
f
th
e
“P
I
D
-N
N
co
n
t
ro
l
l
e
r”
w
e
ight
s
ins
t
ea
d
of
a
p
p
l
i
e
d
t
h
e
t
r
adi
t
i
o
nal
BP
m
etho
d
a
s
s
how
n
i
n
F
i
gur
e
4.
F
i
gur
e
4.
T
he
b
u
i
l
d
i
ng
o
f
t
he
o
p
tim
i
z
e
d
“
P
I
D
-
NN
c
ontr
o
ll
er
”
.
A
M
A
TLA
B
pr
o
g
r
a
m
is
d
e
v
el
o
p
ed
t
o
f
i
n
d
t
he
o
p
tim
al
v
a
l
ue
s
of
t
he
c
on
tro
l
ler
par
a
m
e
ters
V
1
-V
4
.
The
in
te
gr
a
l
t
im
e
squar
e
e
r
r
or
(
ITS
E
)
is
u
sed,
f
or
e
xam
p
l
e
,
an
o
bj
e
c
t
i
v
e
f
itne
ss
fun
c
ti
o
n
o
f
the
s
y
st
e
m
pe
r
f
or
m
a
nc
e,
a
m
i
n
im
i
z
a
t
i
o
n
alg
o
r
i
t
h
m
is
a
ppl
ie
d
in
t
he
s
ea
r
c
h
doma
i
n
of
t
he
p
ar
tic
l
e
s
po
siti
o
n
a
nd
ve
l
o
cit
y
.
(
18)
4.
S
I
MULI
N
K
MODEL
AND RESUL
T
S
F
o
r
the
sa
ke
o
f
ve
r
i
f
y
in
g
t
h
e
per
f
or
m
a
nce
of
t
he
p
r
o
pose
d
d
e
s
ig
n
w
e
h
a
v
e
de
ve
l
o
ped
a
mode
l
o
f
t
he
c
o
n
v
erter
i
n
(
MATLAB
–
Si
m
u
l
i
n
k
),
a
nd
t
he
m
ost
im
p
o
rtan
t
re
s
u
l
t
s
w
i
l
l
b
e
pr
esen
te
d
h
e
r
e
.
F
i
gur
e
5
pr
ese
n
t
the
mode
l for
this
c
on
tro
lle
r for “
B
oos
t c
o
n
v
e
rter
”.
F
i
gur
e
5.
B
o
o
st
c
on
ver
t
e
r
m
odel.
T
o
d
esign
th
e
sui
t
a
bl
e
c
ont
rol
l
e
r
a
n
d
ad
ju
sti
n
g
t
h
e
con
t
roll
er
p
a
rameters,
i
t
’s
r
equ
i
r
e
d
to
e
mp
loy
su
ita
b
l
e
m
ode
l
.
T
he
p
r
opo
se
d
i
n
t
e
l
l
i
g
e
n
t
m
ode
l
c
a
n
be
u
s
e
d
eff
i
c
i
e
n
tly
i
n
t
h
is
w
or
k
t
o
d
es
ig
n
an
d
o
p
t
i
miz
e
the
co
n
t
ro
ll
e
r
p
a
r
am
eter
s.
T
he
p
erform
ance
o
f
the
o
p
tim
iza
t
io
n
a
l
gor
it
hm,
by
us
in
g
the
o
p
t
imizi
n
g
c
o
ef
f
i
c
i
e
n
ts
C
1
=
C
2
=
1
.
3
a
nd
W=0.
9,
g
i
v
e
s
t
he
o
p
t
im
um
v
a
l
ue
s
o
f
t
he
“
PI
D
-
NN
c
ontr
o
l
l
e
r
”
i
n
1
0
iter
a
t
i
ons
a
s
the
fo
l
l
o
w
i
ng
:
V
1
=
1.
69
0
1
e
-
05,
V
2=
2.
9
1
3
9
e-
0
6
,
V
3
=
9.
24
4
0
e-
06,
V
4=
1
.
2
1
5
79e-
0
5
.
T
he
c
on
tr
o
l
s
i
gna
l
sys
t
em
i
n
put
i
s
the
r
a
n
g
e
d
f
r
o
m
0
t
o
4
0V
a
nd
l
i
m
i
t
e
d
w
i
thi
n
t
h
i
s
r
a
n
g
e
.
T
he
o
pt
i
m
iza
t
io
n
per
f
or
m
a
n
c
es
a
r
e
show
n
i
n
F
i
gur
es
6
,
7,
8
a
nd
9
:
F
i
gur
e
6.
T
he
o
p
tim
iza
t
i
o
n
p
e
r
f
or
m
a
nc
e
a
t
V
1.
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
S
S
N
:
2088-
86
94
I
n
te
ll
i
g
en
t co
nt
ro
l of D
C
-
D
C c
onve
r
te
r
b
a
se
d on
PI
D
-
ne
ural
ne
t
w
ork
(
H
ussa
in K.
K
h
le
a
f
)
2
259
F
i
gur
e
7.
T
he
o
p
tim
iza
t
i
o
n
p
e
r
f
or
m
a
nc
e
a
t
V
2.
F
i
gur
e
8.
T
he
o
p
tim
iza
t
i
o
n
p
e
r
f
or
m
a
nc
e
a
t
V
3.
F
i
gur
e
9.
T
he
o
p
tim
iza
t
i
o
n
p
e
r
f
or
m
a
nc
e
a
t
V
4.
F
r
om
t
he
r
esp
onse
i
t
i
s
c
l
ear
t
ha
t
a
f
ter
1
0
m
s
ou
tp
u
t
v
o
lta
ge
i
s
r
e
st
or
ed
t
o
i
t
s
or
ig
i
n
al
v
al
ue
.
T
h
is
se
g
m
e
n
t
i
n
c
l
u
d
es
a
c
om
par
i
s
on
be
tw
ee
n
the
o
u
t
p
ut
r
esu
l
t
s
o
bta
i
n
ed
b
y
a
p
pl
y
i
ng
C
on
ven
t
i
ona
l
“P
I
D
c
o
n
t
r
o
ller
”
a
n
d
“
P
ID-NN”
Based
on
Pa
rt
icle
S
wa
rm
O
ptimiza
t
ion
c
o
n
t
r
o
l
l
e
r
s
t
o
t
he
b
oo
st
c
on
ve
r
t
er
s
o
th
e
sug
g
es
te
d
co
nt
r
o
l
str
a
te
g
y
c
o
u
l
d
b
r
a
nd
the
“
boos
t
c
o
n
v
er
t
e
r
”
f
u
nc
ti
one
d
in
a
s
ta
b
l
e
ma
nner
u
n
d
er
b
i
g
l
oad
tr
a
n
sie
n
t
r
e
spo
n
ses
of
a
l
l
t
he
q
u
i
esce
n
t
i
n
p
u
t
D
C
-
volta
ge
r
a
nge
a
s
sh
ow
n
in
F
igur
e
s
1
0,
1
1
,
12
and
1
3
.
F
i
gur
e
1
0
.
C
o
m
p
ar
i
s
o
n
out
p
u
t
v
o
lta
ge
o
f
th
e
boo
st
c
o
n
v
er
te
r
w
i
t
h
con
v
e
n
ti
o
n
a
l
P
I
D
a
nd P
I
D
-
NN
c
ontr
o
l
l
er
for
10
0
v
-
100
Ω
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
n
t
J Po
w El
ec &
D
ri S
y
s
t
Vo
l. 1
0
,
No
. 4
,
D
e
c
2
0
19
:
2
254
–
2
262
2
260
Fi
g
u
r
e
11
.
C
o
mp
a
r
i
s
on
V
o
o
f
t
h
e
boo
st
c
onv
ert
e
r
t
h
roug
h
c
o
nv
e
n
t
i
ona
l
P
I
D
and
P
I
D-
NN
c
ont
r
o
lle
r
s
f
or
10
0
v
-
200Ω
.
Fi
g
u
r
e 12
.
C
o
mp
a
r
i
s
on
o
ut
pu
t
vol
t
a
g
e
o
f
t
h
e b
o
o
st
c
onv
erte
r
wi
t
h
con
v
e
n
ti
o
n
al
P
I
D
a
nd P
I
D
-
N
N
contr
o
l
l
er
s
for
20
0
v
-
100
Ω
.
Fi
g
u
r
e 13
.
C
o
mp
a
r
i
s
on
o
ut
pu
t
vol
t
a
g
e
o
f
t
h
e b
o
o
st
c
onv
erte
r
wi
t
h
con
v
e
n
ti
o
n
al
P
I
D
a
nd P
I
D
-
N
N
contr
o
l
l
er
s
for
20
0
v
-
200
Ω
.
5.
COMPA
R
I
S
ON ANAL
YSIS AND
D
I
S
C
U
SSIONS
The
pro
b
lem
s
c
a
n
be
ide
n
t
i
f
i
e
d
fr
o
m
the pr
evi
o
us
f
i
g
ures
i
n
the
C
o
n
v
en
t
i
o
n
a
l PI
D c
ont
r
o
lle
r:
-
The
settli
ng t
i
m
e is too
lon
g
.
-
Durin
g
the
tr
a
ns
ien
t
,
the c
o
nver
t
er
oper
a
te
s
i
n
ter
m
it
ten
t
l
y
i
n
di
sc
o
n
t
i
n
uo
us
c
o
nduc
tio
n
mode
.
I
n
o
r
d
e
r
t
o
m
a
k
e
a
b
e
t
t
e
r
c
o
m
p
a
r
i
s
o
n
b
e
t
w
e
e
n
t
h
e
p
e
r
f
o
r
m
a
n
c
e
o
f
t
he
C
on
ve
nti
ona
l
P
I
D
and
P
I
D
-
N
N
c
o
n
t
r
o
lle
r
s
,
base
d
o
n
P
S
O
c
ontr
o
l
l
e
r
s,
c
or
r
e
spon
d
i
n
g
s
te
a
dy-
s
t
a
t
e,
s
ett
l
i
n
g
t
ime
,
over
s
h
o
o
t,
t
r
a
ns
ien
t
r
e
spo
n
se
s
a
n
d
t
h
e
ou
t
p
u
t
vol
t
a
ge
a
r
e
m
or
e
s
t
ab
le
a
n
d
b
e
t
t
e
r
r
e
g
ul
a
t
ed
w
he
n
th
e
c
o
nt
rolle
r
PI
D-NN
i
s
a
p
p
li
e
d
a
s
s
how
n
i
n
T
a
b
l
e
s
1-
4.
I
n
add
i
tio
n,
t
he
opt
im
al
P
I
D
-N
N
contr
o
l
ler
lea
d
s
to
a
f
e
e
dbac
k
s
yste
m
wi
th
a
c
o
n
s
i
d
er
ab
le
f
aster
r
e
sp
on
se,
w
h
i
c
h
a
l
lows
u
s
to
e
nsur
e
c
o
n
t
in
u
ou
s
c
o
n
duc
t
i
o
n
o
p
e
r
a
ti
on
o
f
t
h
e
c
o
n
v
e
r
t
er
a
t
a
l
l
tim
es,
no
o
v
e
r
sh
oots
an
d
shor
ter
sett
l
i
n
g
t
i
m
e
s.
Tab
l
e
1.
P
e
r
for
m
anc
e
a
nal
y
sis
of
P
ID
-
N
N
w
ith
t
he
b
oo
st
c
o
nver
t
e
r
a
t
d
if
fe
r
e
nt
l
oa
d
an
d
V
=
1
0
0
.
V=
100
R=
100
R
=
200
0.
120*
10^-6
0
.
062*
10^
-
6
Se
ttling time (S
e
c
)
0.
093*
10^-6
0
.
125*
10^
-
6
O
v
e
r
shoot
(
%)
2
.
5
7
Tab
l
e
2.
S
how
a
nal
y
s
i
s
o
f
P
ID
-
N
N
w
ith
t
he
boos
t
co
n
v
er
t
e
r
a
t
d
i
ffe
r
e
n
t
l
o
a
d
and
V
=
2
00.
V
=
200
R=
100
R=
200
0.
118*
10^
-
6
0
.
110*
10^
-6
Se
ttli
ng ti
me
(
Sec
)
0.
06*1
0^
-
6
0
.
125*
10^
-6
Ove
r
s
h
o
o
t
(
%
)
3
12.5
Evaluation Warning : The document was created with Spire.PDF for Python.
Int J
P
o
w
E
l
e
c
&
D
ri S
yst
IS
S
N
:
2088-
86
94
Inte
lli
gen
t
co
nt
ro
l
of D
C
-
D
C conv
e
r
t
e
r
b
a
se
d on
PID
-
neur
al
ne
t
w
ork
(H
ussa
in
K. Khle
a
f
)
2
261
Ta
b
l
e 3.
P
e
rforma
n
ce
ana
lysi
s
of
c
on
ve
nt
i
o
n
a
l P
I
D
contro
lle
r w
i
t
h
t
h
e
boost
c
o
n
v
e
rt
er a
t
d
i
f
f
e
ren
t
l
o
a
d
a
n
d
V
=
100.
V=
100
R
=
100
R=
200
100
100
Se
ttling time (S
e
c
)
2.
7*10
^-6
2.
2*10
^
-
6
O
v
e
r
shoot
(
%)
0
0
Ta
b
l
e 4.
P
e
rforma
n
ce
ana
lysi
s
of
c
on
ve
nt
i
o
n
a
l P
I
D
contro
lle
r w
i
t
h
t
h
e
boost
c
o
n
v
e
rt
er a
t
d
i
f
f
e
ren
t
l
o
a
d
an
d
V=20
0
V
=
200
R=
100
R=
200
200
200
S
e
ttl
ing ti
me
(
Sec
)
1.
3*10^-
6
1
.
1*10
^
-
6
O
v
e
r
shoot
(
%
)
0
0
I
n
a
d
d
i
t
i
o
n
,
Ta
ble
5
is
s
h
o
w
n
c
om
par
i
s
o
n
co
ns
i
d
e
r
ed
t
he
f
o
u
rt
h
p
r
e
v
i
o
us
w
orks,
P
I
D
–DO
O
,
H
ybr
id
A
d
a
p
t
i
v
e
N
e
u
r
o
F
u
z
z
y
b
a
s
e
d
s
p
e
e
d
C
o
n
t
r
o
l
l
e
r
,
A
d
a
p
t
i
v
e
N
e
u
r
o
-
F
u
z
z
y
C
o
n
t
r
ol
A
pp
ro
a
c
h
fo
r
a
Si
ngle
Inver
t
e
d
P
end
u
l
u
m
S
y
stem
a
nd
G
A
A
N
F
IS
,
w
h
i
c
h
ha
ve
u
se
d
t
h
e
sa
m
e
d
a
t
a
s
e
t
.
T
h
e
r
e
s
u
l
t
s
s
h
o
w
t
h
a
t
t
h
e
“
P
I
D
-
N
N
w
i
t
h
t
h
e
b
o
o
s
t
c
o
n
v
e
r
t
e
r
”
i
s
a
b
l
e
t
o
p
r
e
s
e
n
t
a
h
i
g
h
e
r
p
erfor
m
a
n
ce
tha
n
o
t
h
ers
i
n
s
o
l
vi
n
g
a
ll
pro
b
lem
s
.
Tab
l
e
5. T
h
e
c
ompa
rison
w
i
th
p
rev
i
o
u
s w
o
r
k
s
R
ef
er
en
c
e
A
r
t
i
c
l
e
(S
e
c
)
S
e
tt
ling ti
me
(S
e
c
)
O
v
e
r
shoot
(
%
)
PID-N
N
w
ith
t
he
boost
c
onve
r
t
e
r
-
0
.
062*
10^
-
6
0
.
125*
10^
-
6
7
PID
–DO
O
(
C
o
m
p
a
rison
of
t
he
v
a
r
ious
buc
k c
onve
rte
r
c
on
trol
m
e
t
hods
i
n
L
E
D
a
p
p
l
i
catio
n
s
)
(S
h
w
eth
a
D
.V
.,
et
a
l.
, 2
0
1
9
)
[21]
-
1
.
07*1
0^
-
5
-
P
I
(
H
y
br
id
a
da
pti
v
e
Ne
uro
Fuz
z
y
sp
e
e
d
c
ontrolle
r
for
B
r
us
hl
e
ss
DC e
ngine
)2017
[22]
0
.
025
0
.
075
1
2
A
N
F
IS
(
Ada
p
tiv
e
Ne
ur
o-Fuz
z
y
P
e
ndulum
Sy
st
e
m
C
ontrol
A
pproa
c
h
)
2018
[23]
-
2
.
2
0
.
095
ra
d
GA ANFIS (I
n
te
llige
n
t s
e
l
f
-tuning G
A
ANFIS Pla
stic
E
xt
rusio
n
Sy
s
t
e
m
Te
m
p
e
r
a
t
ure
C
ontro
l
l
e
r
d
e
s
i
g
n
)
2011
[24]
6
0
1
650
0
6.
CONCL
U
S
ION
Th
is
a
rt
i
c
le
s
u
gge
sts
a
n
i
nt
e
l
l
i
ge
nt
P
ID
c
ontr
o
l
l
e
r
b
a
s
e
d
on
A
N
N
a
n
d
op
t
i
mize
d
b
y
u
s
i
n
g
P
S
O
met
h
od
a
ppl
i
e
d
t
o
t
h
e
c
onv
ert
e
r
Bo
ost
.
T
he
i
nt
e
l
li
g
e
nt
m
o
d
e
l
i
s
u
se
d
e
f
f
i
ci
en
tl
y
to
a
da
p
t
a
n
d
opt
i
m
iz
e
th
e
P
I
D
-
NN
c
ontr
o
lle
r,
S
imula
t
i
o
n
fin
d
i
ngs
h
a
v
e
s
how
n
tha
t
t
he
P
ID
-
NN
c
o
nt
roll
e
r
p
rovi
d
e
s
a
smoot
h
re
a
c
t
i
on
t
o
t
h
e
r
ef
e
r
en
ce
mo
n
i
to
ri
ng
a
n
d
ret
a
in
s
th
e
b
o
o
s
t
co
nv
ert
e
r
o
u
t
pu
t
vo
l
t
a
g
e
a
cc
o
r
d
i
ng
t
o
t
h
e
requi
re
d
vo
lt
a
g
e
com
p
ara
t
i
v
e
wit
h
t
he
C
o
n
v
en
ti
o
n
al
P
ID cont
rol
l
e
r
. This co
n
t
ro
ll
er ca
n
p
ro
vi
de a
m
uch
be
tter rea
c
t
i
on
t
o start-
up
t
ha
n t
h
e
P
I
D
con
t
ro
l
l
er
f
o
r
t
he
w
h
o
le
o
rd
e
r
.
The
P
I
D
-
NN
c
on
t
roll
e
r
a
lso
h
a
s
a
g
ood
d
yn
ami
c
r
es
pon
se
a
nd
an
o
ut
s
t
a
n
d
i
n
g
start-up
rea
c
t
i
o
n
as
s
h
o
w
n
.
F
i
na
l
l
y,
t
he
r
esu
l
t
i
ng
de
sig
n
w
as
b
ased
on
the
s
i
m
p
lic
i
t
y
of
t
h
i
s
con
v
er
t
e
r,
i
t
s
rob
us
t
n
ess
a
n
d
i
t
s l
o
w
pa
rt
c
o
u
n
t, ap
p
lica
t
io
n i
n hig
h
p
ow
er hi
gh re
l
i
ab
i
l
i
t
y
app
l
icat
i
o
n
s
.
Th
is i
s
espec
i
al
ly
t
he
case
i
f
f
a
s
t
a
n
d
com
p
ac
t
co
n
t
rol
te
c
h
niq
u
e
s
,
lik
e
t
h
e
o
n
e
p
r
e
s
e
n
t
e
d
h
e
r
e
,
a
r
e
u
s
e
d
t
h
a
t
a
l
l
o
w
f
o
r
ine
x
pen
s
i
v
e a
n
d rob
u
st c
o
n
tr
oller
s
u
se.
REFE
RENCES
[1]
Mo
ham
e
d
.
A
.
Sham
sel
d
i
n
,
M
oham
e
d
S
a
llam,
A
.
M.
B
as
siu
n
y
,
A
.
M
.
A
bd
e
l
G
ha
n
y
,
"A
n
o
v
el
s
elf
-
tu
ni
ng
f
r
acti
onal
order
P
I
D
co
nt
rol
b
a
sed
on
o
p
t
i
m
a
l
m
o
de
l
ref
e
ren
c
e
ad
apt
i
ve
s
ys
tem,"
Internat
ional Journal
o
f
Po
wer
El
ectro
n
i
cs
and
Drive
Syst
e
m
(
I
JPEDS)
,
vo
l. 10
,
no
. 1
,
2
0
1
9
.
[2]
Deept
i
Y
adav,
Arun
im
a
V
e
rma,
"Co
m
perati
ve
P
erf
o
rman
ce
An
alys
is
o
f
P
M
SM
D
ri
ve
Usin
g
M
P
S
O
a
n
d
AC
O
Techn
i
q
u
es
,"
In
ter
n
a
t
i
onal
J
o
ur
n
a
l of Power
El
ectr
o
n
i
cs a
nd Driv
e
Syst
em
(
I
JPEDS
)
,
v
o
l
.
9
,
no
.
4
, 2
01
8.
[3]
Liu
L
u
o
ren
an
d
L
uo
Jin
l
i
ng,
"
Res
earch
o
f
P
I
D
Cont
rol
Alg
o
rithm
B
a
se
d
on
N
e
u
r
a
l
Ne
tw
or
k
,
"
J
o
urna
l o
f
En
e
r
g
y
Pro
cedi
a
,
vol.
13,
pp.
6
9
8
8
–
69
9
3
,
20
11.
[4]
Leil
a
F
a
ll
ah
A
ra
gh
i,
M
.
H
a
b
i
b
n
ej
ad
K
oray
em,
A
m
i
n
N
ik
oo
b
i
n
a
n
d
F
arb
o
d
S
e
toud
eh,"Neu
r
al
N
etw
o
rk
C
ont
ro
lle
r
Ba
s
e
d
on
P
ID
C
o
n
t
r
oller
f
o
r
Tw
o
li
nks
-
Ro
boti
c
M
an
ipulat
or
C
on
t
rol
,
"
Jo
urn
a
l
of En
g
i
neerin
g and
Comp
u
t
er
Scien
ce W
C
E
C
S
,
S
an
F
ranci
s
co,
US
A,
2
008.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SSN: 2088-
8694
I
nt
J
P
ow
Elec
& Dr
i
S
y
st V
ol.
10,
N
o.
4
, Dec
201
9 :
2
2
5
4
– 2
262
2
262
[5]
Mo
hd
R
u
d
d
i
n
A
b
G
han
i
,
S
a
if
T
ah
seen
H
us
sein
,
Zanari
ah
J
an
o,
T
ole
S
utikno,
"
P
a
rti
c
l
e
S
warm
O
ptimizatio
n
P
e
rf
o
r
m
a
nce:
C
om
p
a
riso
n
o
f
D
y
n
am
i
c
E
conomic
D
isp
a
tch
with
D
ant
z
ig-
W
olfe
D
ecompositi
o
n
,
"
TELK
OMNI
KA
(T
e
l
ecom
m
un
icati
on,
Co
m
p
u
t
i
n
g
,
El
ectr
o
n
i
cs a
nd
Con
t
r
o
l)
,
vo
l
.
1
4
, no
.
3,
20
16
.
[6]
T.
G.
M
an
j
u
n
a
t
h
,
Ash
ok
K
usag
ur,
"An
a
ly
sis
of
D
iff
e
ren
t
M
eta
Heur
i
s
tics
Me
t
h
od
i
n
Intel
l
igent
F
a
ul
t
De
t
ecti
o
n
o
f
Mu
lt
il
evel
I
nv
erter
w
i
th
P
hoto
v
o
l
t
a
ic
P
o
w
er
G
enerati
o
n
S
o
u
r
ce,"
Inter
n
a
t
i
onal Jou
r
na
l
of Po
we
r
E
l
ectr
onics
a
n
d
Dri
ve System
(
I
JPEDS
)
, v
ol
.
9
, n
o
.
3
,
2
0
1
8
.
[7]
Yass
er
A
hm
ed,
A
y
m
a
n
H
oballah
,
"
A
dap
t
i
v
e
filter-F
LC
i
ntegrat
i
o
n
f
or
t
o
r
qu
e
ri
p
p
les
minimi
zatio
n
in
P
MS
M
using
PS
O,"
Int
e
rn
a
t
i
o
n
a
l
Jou
r
n
a
l
o
f
Po
wer Electr
onics an
d Dri
ve System
(
I
JPED
S
)
,
v
o
l
. 10
, n
o.
1
, 2
01
9.
[8]
Zahra
Behesht
i
,
Siti
Mariyam
Hj
.
Shamsuddi
n
,
"A
R
eview
of
P
o
p
ul
a
tion-based
Met
a
-Heuri
s
t
ic
A
l
g
or
it
h
m
,"
Int.
J.
Adva
nce.
Soft Comp
ut
.
A
ppl.
,
vo
l
.
5
,
n
o
.
1,
2
01
3.
[9]
Ali
K.
N
ah
ar,
A
n
s
a
m
S
.
J
abb
a
r,
M
o
h
mm
ed
J
.
Mo
rtada,
"
A
N
ov
el
I
m
p
r
ov
e
and
C
o
m
p
ression
f
o
r
t
h
e
M
edi
cal
I
m
a
ge
Techn
i
q
u
e
Bas
e
d
O
n
t
he
D
ouble
Dens
it
y
W
a
vel
e
t
,"
W
o
r
l
d
W
i
de J
o
u
r
na
l of Engin
e
eri
n
g
an
d
T
ech
nol
og
y
,
2
018.
[10]
P
.
B
h
a
skara
P
r
asad,
M
.
P
adm
a
L
ali
t
h
a
,
B.
S
arv
e
s
h
,
"Fract
io
nal
Ord
e
r
P
I
D
Contro
ll
ed
C
as
cade
d
R
e-bo
os
t
Seve
n
Level
Invert
er
F
ed
I
n
d
u
c
ti
on
M
o
to
r
S
y
st
e
m
w
ith
En
hanced
R
es
po
ns
e,"
Internat
ion
a
l
Journal of Powe
r El
e
c
t
r
on
i
c
s
and
Drive
Syst
e
m
(
I
JPEDS)
,
v
o
l
.
9
,
no
.
4
,
2
018
.
[11]
Mo
ham
a
d
Isn
aeni Ro
m
a
dh
on
, etc…,
"
A
C
omp
a
r
i
sson
o
f
Syn
c
h
r
on
ou
s
a
n
d
Non
s
y
n
c
h
ro
no
us
B
oo
st
C
on
ve
rte
r,"
IAES
Inter
n
a
t
i
onal
Co
nf
eren
ce on
El
ectrical
En
g
i
neer
in
g,
Co
mp
u
t
er
Sc
i
e
nce a
n
d
Inf
o
rma
tics
,
20
17
.
[12]
Chand
r
a
S
h
ekhe
r
P
u
ro
hit,
e
tc.……
.,"P
e
rf
orm
a
n
c
e
anal
ys
is
o
f
DC/
D
C
b
i
d
irectio
nal
con
v
ert
e
r
w
ith
s
li
di
ng
m
od
e
an
d
pi
c
ont
ro
ller,
"
In
tern
atio
nal Jo
urnal
of
P
o
wer El
e
c
t
r
o
n
i
c
s a
n
d
Drive S
y
st
em
(
I
JPEDS
)
,
vol.
1
0
,
n
o
.
1
,
2
019.
[13]
Mu
ham
m
a
d
W
a
s
i
f
Um
ar,
Norzai
har
B
Y
a
h
a
ya,
Zu
hai
r
i
B
Baharu
dd
in"P
W
M
D
immi
ng
C
o
n
t
r
ol
f
o
r
H
i
g
h
Brig
h
t
ness
LE
D
B
a
s
e
d
Au
to
mot
i
ve
L
ig
ht
ing
Applications,"
In
ter
n
a
t
i
ona
l J
o
ur
nal
of
El
e
c
tri
c
al
an
d Com
puter E
ngin
eerin
g
(IJE
C
E
)
v
o
l
.
7
, no
.
5
,
2
017
.
[14]
P
.
S
iv
a
S
u
b
r
a
m
ani
a
n
an
d
R.
K
ay
a
l
vi
zhi
.
"
An
O
p
t
i
m
um
S
ett
i
ng
of
P
ID
C
ontrol
l
er
f
o
r
B
oost
Co
nvert
er
U
s
i
ng
Ba
ct
erial
F
o
rag
i
ng
Op
t
imizat
ion
Tech
ni
qu
e,"
Jour
na
l o
f
Sprin
ger
Ind
i
a
,
20
15
.
[15]
P
r
o
f
.
Dr.
A.
K
.
Al
-S
ha
ik
hli,
A
s
s
t
.
P
rof.
D
r.
A
bdul
-Ra
h
im
T
.
Hum
od,
F
adhil
A.
H
asan
(
M
S
c),
“Travel
i
n
g
W
ave
Indu
cti
o
n
H
eating
Co
nt
rol
Based
on
R
o
b
u
s
t
In
telli
gen
t
C
on
t
r
ol
le
r,"
Ad
van
c
es
in
Na
tura
l
a
n
d
App
l
ie
d
Sc
ie
nc
e
s
,
vol
.
1
0
,
no
.
17
,
pp.
125
-13
4
.
[16]
Jam
e
s
Carv
ajar,
G
u
ang
r
on
g
Ch
en
,
"F
uzzy
P
ID
C
on
tro
l
ler:
D
esig
n,
p
e
rf
orm
a
nce
ev
aluat
i
o
n
a
n
d
s
tabili
ty
a
naly
si
s
,"
Informa
tion Scie
nces
, vo
l
. 1
23
,
n
o
. 3
, p
p.
24
9
-
27
0, 20
0
0
.
[17]
B
e
i
t
a
o
G
u
o
,
H
.
L
i
u
,
Z
.
L
u
o
,
F
.
W
a
n
g
,
“
A
d
a
p
t
i
v
e
P
I
D
C
o
n
t
r
o
l
l
e
r
B
ased
o
n
Neu
r
al
N
et
wo
rk
,"
IE
E
E
,
Interna
t
ional
Joint Con
f
erence
on A
r
t
ific
i
a
l In
t
e
l
l
igence,
20
09
.
[18]
Ahm
e
d
S
.
A
l-A
r
aji
,
A
h
m
ed
I
.
A
b
d
u
l
Kareem
,
"
A
Non
l
i
n
ear
N
eural
Con
t
ro
ll
er
D
es
ig
n
f
o
r
t
h
e
S
i
ngle
Axi
s
M
ag
net
i
c
Ba
l
l
Levitatio
n S
y
stem
B
as
ed
on
S
lice
Genetic
A
lgor
i
t
h
m
,"
E
n
g
.
An
d
T
ech
. Jou
r
n
a
l
,
34
Part(
A
):
1
, 20
1
2
.
[19]
Ishaq
u
e,
K
.
and
Z.
S
al
am
,
"
A
d
eterm
i
n
i
s
tic
p
art
i
cl
e
s
w
arm
o
p
tim
ization
maximum
powe
r
point
t
r
ac
ker
f
o
r
pho
to
vo
lt
aic s
y
stem
un
d
er p
arti
al s
had
i
ng
co
nditi
o
n
,"
IEEE T
r
an
s
.
In
d.
Elect
ro
n
., v
ol.
6
0
(
8),
pp
.
3
1
95
-32
0
6
, 20
1
3
.
[20]
Lian
,
K
.L
.,
J
.
H
.
J
h
ang
a
n
d
I.
S
.
T
i
a
n,
"
A
m
a
xim
u
m
power
p
o
i
nt
t
r
a
c
k
in
g
me
th
od
b
a
s
e
d
o
n
pe
rturb
-
a
n
d
o
b
s
e
r
v
e
com
b
i
n
ed
w
ith
p
art
i
cle swarm o
p
t
imi
zatio
n
,
"
I
E
EE
J.
P
h
ot
ovo
l
t
a
ic
,
vo
l.
4
(2),
pp.
6
26-6
3
3
,
2
014
.
[21]
S
h
wet
h
a
D.V.,
L
a
ks
hm
an
R
ao
S
.
P
a
rag
o
n
d
,"Com
p
ari
s
o
n
o
f
t
h
e
d
i
ff
e
rent
c
ont
rolli
ng
t
echn
i
qu
es
f
o
r
b
uck
conv
erter
in LED appl
i
cat
ion,"
Int
e
rna
t
i
onal
Jou
r
nal
o
f
Powe
r
Electr
onics
an
d D
r
i
v
e
Sys
t
e
m
(IJPED
S
)
,
v
o
l
. 10
, n
o.
1
, 2
01
9.
[22]
K.
H
arshav
ardh
ana
RE
DDY
,
etc….
,
"
Hyb
r
id
A
d
a
pti
v
e
N
e
uro
F
u
zzy
b
a
se
d
spe
e
d
C
on
trolle
r
fo
r
Bru
s
h
l
e
s
s
DC
Mo
t
o
r,
"
Jou
r
nal of
Sci
e
nce
, 20
1
7
.
[23]
Mo
ham
m
e
d
A
.
A
.
Al
-M
ekh
l
a
f
i1,
etc….,"Ad
apt
i
v
e
N
euro
-Fu
zzy
C
on
tro
l
Ap
proach
f
o
r
a
S
i
n
g
l
e
Inv
e
rted
P
end
u
l
u
m
Sy
st
e
m
,"
In
t
e
rna
t
i
o
n
a
l Jou
r
n
a
l
of El
ectr
i
cal
an
d
Co
mp
ut
er
E
ngin
eer
in
g
(
I
JE
CE)
,
vo
l
.
8,
n
o
.
5,
20
1
8.
[24]
S
.
R
avi,
M
.
S
udha,
and
P
.
A
.
Ba
l
a
kri
s
h
n
an,
"
D
e
sig
n
o
f
Inte
l
l
i
g
en
t
Sel
f
-
T
uning
G
A
ANFI
S
T
e
m
p
e
rature
C
ont
rol
l
er
fo
r P
l
asti
c Extrusi
on S
y
stem,"
Modelling and Sim
u
lat
i
on
in
Eng
i
neeri
n
g
, 2
01
1
.
Evaluation Warning : The document was created with Spire.PDF for Python.