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.
9
, No
.
3
,
J
un
e
201
9,
pp. 164
5~16
55
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v9
i
3
.
pp1645
-
16
55
1645
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
PID cont
ro
ll
er us
ing rapi
d contr
ol p
ro
t
oty
pin
g t
ec
hn
iqu
es
Oda
ir
A. Tr
uji
ll
o
1
, N
ic
ol
ás
Toro
-
G
arcía
2
,
Fred
y
E.
Hoy
os
3
1
,2
Depa
rtment
of
Elec
tr
ical and E
le
c
troni
cs
Engi
n
ee
ring
&
Com
pute
r
Sc
ie
nc
es,
Uni
ver
sidad
Na
ci
on
al
d
e
Colom
bi
a
-
Sede
Maniz
a
le
s,
Colo
m
bia
3
Univer
sidad
Na
ci
ona
l
de
Colom
bia
-
Sede
Mede
l
lí
n
-
Facultad
de
Cie
ncias
-
Escuela
d
e
Fís
ica, Col
om
bia
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
J
un
1
9
, 2
01
8
Re
vised
Dec
18
,
201
8
Accepte
d
Dec
29
, 201
8
To
an
aly
z
e
the
per
form
anc
e
of
the
PID
cont
rol
l
er
in
a
buck
t
y
p
e
conv
erter
implemente
d
in
r
ea
l
ti
m
e.
W
e
beg
in
b
y
design
ing
a
cont
inuous
controlle
r
usin
g
the
an
aly
t
ical
m
et
hod
for
ca
l
culati
ng
PID
s.
Puls
e
width
m
odulat
ion
is
then
used
and
bifurca
ti
on
d
ia
gra
m
s
an
aly
z
ed to
rev
eal
som
e proble
m
s
of s
witc
hin
g
and
sam
pli
ng
t
i
m
e.
The
m
ode
l
conve
rt
er
is
th
e
n
implemente
d
with
a
PID
cont
roller
in
dSP
ACE.
Th
e
expe
ri
m
ent
al
r
esult
s
pr
ovide
d
et
a
il
ed
re
quire
m
ents
of
sam
pli
ng
fre
q
uency
and
sw
it
c
hing
spee
d,
and
show
the
per
for
m
anc
e
of
th
e
PID
cont
roll
e
r.
Convert
ers
are
used
in
power
g
ene
ra
ti
on
sol
ar
s
y
stems
an
d
conmuted
power
source
s
for
fe
ed
te
l
ec
om
m
unic
at
i
on
dev
ic
es
,
sm
ar
t
gri
ds,
an
d
othe
r applica
ti
on
s.
Ke
yw
or
d
s
:
DC
-
DC con
vert
ers
Dynam
ic
p
ow
e
r
c
onver
te
rs
analy
sis
Nu
m
erical
b
if
urcat
ion
PI
D
h
a
r
dw
a
re
PI
D
s
of
twa
re
Ra
pid
c
ontrol
prototypi
ng
Copyright
©
201
9
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
:
Fr
e
dy E. H
oyos
,
Faculta
d de Ci
encias
-
Esc
uel
a d
e
Físi
ca
,
Un
i
ver
si
dad Na
ci
on
al
de
C
olo
m
bia
-
Se
de M
edell
ín
,
Ca
rr
era
65 N
ro. 59
A
-
110, Me
dellí
n,
C
olo
m
bia.
Em
a
il
:
feh
oyosve@
unal
.edu.c
o
1.
INTROD
U
CTION
The bu
c
k
c
onve
rter is a sw
it
c
hing po
wer
s
upply wit
h
t
wo
sem
ic
on
duct
or
dev
ic
es
[1]
, a
t
ran
sist
or
(
)
and
diode
(
),
an
in
du
ct
or
(
),
an
outp
ut
ca
pacit
or
(
),
an
d
it
s
resp
ec
ti
ve
load
as
s
how
n
in
Fi
gure
1.
This
syst
e
m
co
nv
e
rts
a
n
un
re
gu
la
te
d
DC
vo
lt
age
to
an
ot
he
r
regulat
ed
lo
wer
vo
lt
age
.
T
he
t
ran
sist
or
s
witc
hes
the
input
po
w
er
by
pu
lse
w
idth
m
od
ulati
on
(PWM
)
t
o
f
eed
the
filt
er
.
The
diode
co
nducts
wh
e
n
t
he
transisto
r
is
off
if
there
is
posi
ti
ve
current
in
t
he
coil.
T
he
in
du
ct
or
do
e
s
no
t
al
low
rap
i
d
c
hanges
in
c
urre
nt
an
d
the
capaci
to
r
do
e
s
not
sup
port
ab
rupt
c
hanges
o
f
volt
age,
creati
ng
betwee
n
the
m
a
low
pas
s
filt
er.
A
MOS
FET
is
us
ed
f
or
s
witc
hing,
th
us,
the
con
t
ro
l
te
ch
nique
m
us
t
determ
ine
the
OF
F
and
O
N
switc
h
tim
es.
This c
on
t
ro
l
form
is k
now
n
as
pulse
width m
odulati
on (P
WM)
[
2]
-
[
6].
Figure
2
s
how
s
a
blo
c
k
dia
gra
m
o
f
the
syst
e
m
unde
r
st
udy.
This
syst
em
is
div
ide
d
int
o
t
wo
m
ajor
su
bg
rou
ps
:
the
hardw
a
re
that
include
s
the
ph
ysi
cal
and
el
ect
ronic
com
po
ne
nts
an
d
the
s
oft
war
e
im
ple
m
ented
on
a
dS
P
ACE
,
w
hich
pe
rfo
r
m
s
the
sign
al
acqu
isi
ti
on
an
d
im
ple
m
entation
of
P
ID
c
ontr
ol
t
echn
i
que.
Th
e
hard
war
e
co
ns
i
sts
of
a
se
nsors
(
hi
gh
-
im
ped
an
ce
resist
or
s
co
nnect
ed
in
pa
rall
el
)
to
m
easur
e
t
he
sta
te
var
ia
bl
e
(cap
aci
to
r
vo
lt
age)
.
T
he
co
nv
erter’s
switc
h
i
s
ha
nd
le
d
f
ro
m
the
co
ntr
ol
an
d
dev
el
op
m
ent
card
t
hroug
h
the
P
W
M
and
the
pulse
is
opto
-
is
olate
d
th
r
ough
fa
st
optoc
ouplers
.
The
di
gital
pa
rt
is
de
vel
oped
in
the
co
nt
ro
l
a
nd
dev
el
op
m
ent
c
ard
dS
P
ACE
DS110
4
[7
]
,
w
her
e
the
P
ID
c
on
t
ro
l
te
ch
niqu
e
is
im
ple
m
ent
ed.
P
ID
wa
s
use
d
due
to
it
s
easy
avail
abili
ty
and
com
patibil
ity
with
so
m
e
series
of
PL
Cs,
D
CS
so
ft
war
e
,
Adva
ntech
6022,
an
d
popula
r
softw
are
s
uc
h
as
Lab
View
a
nd
MATL
AB
[
8],
[9
]
.
T
he
car
d
is
pr
ogram
m
ed
us
i
ng
the
MATLAB/Si
m
ulink
®
platf
orm
and
ha
s
a
disp
la
y
inter
face
cal
le
d
“
Con
tr
olDe
sk.
”
The
c
on
tr
oll
er
is
i
m
ple
m
ented
in
Sim
ulink
®
an
d
dow
nlo
a
d
e
d
t
o
the
DSP
to
ope
rate
at
a
giv
e
n
f
reque
ncy
[
10]
.
The
sta
te
va
riable
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.
9
, N
o.
3
,
June
2019 :
1645
-
1655
1646
com
es
to
t
he
c
on
t
ro
ll
er
th
rou
gh
th
e
12
-
bit
AD
C
in
put
with
a
sam
pling
r
at
e
set
for
the
exp
e
rim
ent
at
10
kHz
and
t
hen
30
kHz.
T
he
c
on
tr
ol
le
r
par
am
et
ers
are
set
thr
ough
va
riable
gain
blo
c
k
s
be
caus
e
they
wer
e
t
uned
i
n
real
tim
e.
At
ea
ch
sam
pling
pe
rio
d,
the
c
on
t
rol
le
r
cal
culat
es
t
he
duty
cy
cl
e
i
n
real
tim
e
and
it
s
equ
ivale
nt
in
th
e
P
W
M c
ontrol
s
witc
h.
Figure
1. Sc
he
m
at
ic
o
f
buc
k conve
rter m
odel
Figure
2. Bl
oc
k diag
ram
o
f
th
e
pro
posed
sys
tem
This
arti
cl
e
is
intende
d
f
or
th
e
dig
it
al
P
W
M
appro
ac
h,
w
hich
prese
nts
ric
h
no
nlinear
dynam
ic
s
and
has
the
f
ollo
wing
a
dvanta
ges:
im
m
un
it
y
to
a
nalo
g
com
pone
nt
var
ia
ti
ons
;
f
ast
er
process
de
sign
;
lo
w
sens
it
ivit
y
to
c
ha
ng
es
in
par
am
et
ers
;
pr
ogram
m
ab
ilit
y;
al
lows
reduct
ion
or
el
i
m
inati
on
of
pa
ssive
exter
nal
c
om
po
ne
nts
;
al
lows
cal
ib
rat
ion
of p
r
otect
ion
al
gorithm
s
;
and
s
upport
s
di
gital
syst
e
m
interface
s
f
or
im
ple
m
enting
nonl
inear
con
t
ro
l
te
ch
niques
[
11
]
–
[
16]
.
Howe
ver,
t
he
DPW
M
al
s
o
ha
s
disad
va
nt
ag
es
as
it
is
af
fect
ed
by
tw
o l
im
itati
on
s:
qu
a
ntiza
ti
on ef
fects [
12]
, [1
3] an
d delay
s in
t
he
c
on
t
ro
l l
oop [
17
]
, [
18
]
.
2.
BUCK C
ONV
ERTE
R MO
D
EL
Wh
e
re
is
the
inter
nal
re
sist
an
ce
of
the
MOS
FET
(
)
plu
s
the
resist
a
nce
s
our
ce
(
)
a
nd
t
he
re
sist
ance
c
oil
(
)
plu
s
the
s
hunt
re
sist
or’s
resist
ance
use
d
f
or
the
c
urre
nt
m
easur
e
(
)
as
s
how
n
in
Fig
ure
1.
By
su
m
m
at
ion
o
f
curre
nts in
the ca
pacit
or’s node
we o
btain
So
lvi
ng and
sim
pl
ify
ing
giv
e
s
Convertin
g
t
o m
at
rix
form
giv
es
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
PID co
ntr
oller usi
ng rapid
control
prototypi
ng tech
niques
(
Odair A
. T
r
ujil
lo
)
1647
Convertin
g
t
o st
at
e v
aria
bles
and m
aking
=
+
,
=
1
,
=
1
,
=
1
, and
m
ai
ntain
in
g
si
gn
s
g
i
ve
s
(1)
If
the
s
witc
h
st
ay
s
OFF
(
u=0),
the
n
t
he
eq
uation
beco
m
es
X
̇
=AX
because
it
does
no
t
ta
ke
into
acco
unt
the d
i
od
e
’s
volt
age.
In the case
wh
ere the
sw
it
ch
i
s close
d,
we fi
nd the t
ran
s
fer
functi
on
us
in
g
w
hic
h
yi
el
ds
(2)
In
(
2), the
foll
owin
g values
w
i
ll
b
e re
placed
to fin
d
the
syst
em
’s
resp
onse:
=
40
,
=
46
.
27
µ
=
39
.
3
Ω
,
=
0
.
688
Ω
=
2
.
47
,
=
1
.
345
Ω
(3)
The p
oles for
(3) are
The
root
l
ocu
s
is
sho
wn
Fig
ure
3
.
T
he
refor
e
,
the
syst
em
is
sta
ble
with
the
value
s
us
e
d
f
or
c
on
sta
nt
s
and
it
is
un
derd
a
m
ped
.
T
he
r
es
pons
e
t
o
ste
p
pe
d
E
vo
lt
s
can
be
seen
in
Fig
ure
4,
wh
ic
h
s
hows
an
overs
ho
ot
that
m
us
t
be
reduce
d
a
nd
a
ste
ady
-
sta
te
err
or
of
a
bout
2
volt
s,
in
dicat
ing
t
he
sy
stem
is
un
de
r
dam
ped
.
In
lo
w
-
powe
r
bu
c
k
c
onve
rter
s,
su
c
h
as
com
pu
te
r
s
ources,
i
t
is
us
ual
t
o
fi
nd
a
ze
ner
diod
e
an
d
tra
ns
ist
or
-
base
d
re
gula
ti
on
of
the
co
nv
e
rter
o
ut
pu
t.
This
te
chn
i
qu
e
has
th
e
adv
a
ntage
of
avo
i
ding
the
us
e
of
c
on
tr
ol;
howe
ver
,
it
ha
s
the
disad
va
ntage
of
incre
asi
ng
the
pro
bab
il
it
y
of
fail
ur
e
on
point
s
w
her
e
it
ha
nd
le
s
the
powe
r
outp
ut.
T
o
obta
i
n
the
desire
d
volt
ag
e
with
pr
op
e
r
MOSFET
handling
a
nd
wit
hout
ad
ding
el
em
ents
that
m
a
y
fail
in
the
c
onve
rter
ou
t
pu
t i
s
the
goal
of im
ple
m
e
nting a c
on
t
ro
ll
er.
Figure
3
.
Ro
ot
locus
Figure
4
.
Step
respo
ns
e
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.
9
, N
o.
3
,
June
2019 :
1645
-
1655
1648
3.
CONTIN
UOUS P
ID
C
ONT
ROLLE
R
The
m
et
ho
d
use
d
is
know
n
as
the
“
a
naly
ti
cal
m
et
ho
d
.”
We
sta
rt
from
the
desire
d
be
hav
i
or
as
determ
ined
by
the
seco
nd
-
order
can
onic
al
equ
at
io
n,
wh
ic
h
s
houl
d
m
at
c
h
t
he
c
har
act
e
r
ist
ic
equ
at
io
n
of
th
e
cl
os
ed
-
lo
op sy
stem
w
it
h
PID
,
to
fin
d
the
con
sta
nts
Kp
,
Kd
, a
nd
Ki
.
3.1
.
Desired
response
=
2
.
5
,
=
10%
=
0
.
1
Seco
nd
-
order c
anonical
e
qu
at
i
on (
desire
d be
ha
vi
or
):
Po
le
s
on:
3.2.
Clos
e
d l
oop wi
th
unk
no
w
n P
ID
con
stants
Re
placi
ng
the
par
am
et
ers
in
the
t
ran
s
fer
fun
ct
ion
a
nd
m
aking
a
cha
nge
of
va
riables
=
8
7
3
9
229
,
=
1372
,
an
d
=
9
1
9
137
7
,
we
m
us
t
cl
os
e
the
l
oop
with
unknow
n
P
ID.
B
l
ock
dia
gr
am
as
sho
wn
i
n
Figure
5.
Figure
5.
Bl
oc
k diag
ram
This
giv
es
the t
ran
s
fer f
unct
io
n wit
h
the
un
know
n
c
on
sta
nt
s of PID:
Howe
ver,
be
fore
m
a
tc
hin
g
th
e
char
act
erist
ic
equ
at
io
ns
,
we
m
us
t
first
increase
the
de
gr
ee
of
the
se
co
nd
-
orde
r
char
act
e
risti
c
equ
at
io
n
.
This
i
s
achieve
d
by
add
i
ng
a
rem
nan
t
pole
to
the
char
act
erist
ic
equ
at
io
n
s
o
as
to
not
aff
ect
the
desir
ed
be
ha
vior;
in
this
case,
that
is
five
ti
m
es
a
way
from
the
do
m
inant
pole
s.
This
m
od
ifi
cat
ion
us
u
al
ly
aff
ect
s
the am
plit
ud
e,
bu
t t
he beha
vi
or of
the
can
on
ic
al
syst
e
m
is p
reserve
d
:
Th
us
, it i
s
now
possible t
o
m
a
tc
h
the c
oe
ff
ic
i
ents:
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
PID co
ntr
oller usi
ng rapid
control
prototypi
ng tech
niques
(
Odair A
. T
r
ujil
lo
)
1649
Re
placi
ng
the
s
e
co
ns
ta
nts
i
n
the
cl
ose
d
-
lo
op
t
ran
s
fer
func
ti
on
,
the
sim
ul
at
ion
is
the
n
pe
rfor
m
ed
with
a
ste
p
value o
f
E
volt
s.
Figure
6
s
hows
that
the
c
on
t
rol
le
d
syst
e
m
does
not
reac
h
th
e
desire
d
overs
hoot;
ho
wev
e
r,
it
is
norm
al
to
do
s
om
e
tun
in
g,
prefe
ra
b
ly
w
hen
the
c
ontr
oller
is
al
re
ady
m
od
ulate
d
by
the
P
WM
so
a
s
t
o
a
void
double
tun
in
g. T
he
set
tl
ing
ti
m
e and
the stea
dy
-
sta
te
erro
r
are
the
de
sired o
utcom
es.
Figure
6.
Com
par
is
on of s
te
p
respo
ns
e i
n open
-
a
nd close
d
-
loop syst
em
s wi
th PID
4.
DISCRETE
T
IME
S
YS
TE
M
Fo
r
a c
onti
nuous
ti
m
e,
syst
e
m
, th
e g
ene
ral
so
luti
on is
giv
e
n by
Ther
e
f
or
e,
the
so
luti
on i
n discrete
tim
e fo
rm
is g
ive
n by
(4)
The
s
olu
ti
on
of
Equati
on (4
)
i
s obtai
ne
d
as
Ma
king
=
y
=
(
−
)
−
1
, th
e
so
l
ution can
be p
resen
te
d
as
(5)
The
n,
t
he
s
olu
t
ion
of
is
(6)
Fr
om
(6)
a
nd a
pp
ly
in
g
La
plac
e’s
a
nti
-
tran
sf
orm
ation
, we
ob
ta
in the m
at
rix
:
(7)
wh
e
re
1
=
−
(
)
,
2
=
−
−
(
)
−
√
1
−
2
,
and
=
√
1
−
2
;
=
−
1
√
1
−
2
.
At this
point,
we pr
obe that
evaluate
d
at
=
0
is eq
ual to t
he
i
de
ntit
y
m
at
rix.
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.
9
, N
o.
3
,
June
2019 :
1645
-
1655
1650
Kno
wing
,
an
al
gorithm
was
i
m
ple
m
ented
in
MATL
AB
®
to
produce
t
he
discrete
e
quivale
nt
sh
ow
n
i
n
Fig
ur
e
7.
The
c
onti
nuous
ti
m
e
syst
em
has
=
5
.
9
ms
,
so
we
ca
n
c
hoos
e
a
sam
pling
rat
e
=
50
μ
seg
.
I
niti
al
iz
ing
th
e
necess
ary
va
riables,
we
obt
ai
n
the
outp
ut
sh
ow
n
i
n
Fi
gure
7.
Th
us
,
we
can
see
how
t
he discret
e eq
uiv
al
e
nt is
ver
y cl
os
e to
th
e real e
qu
i
valent.
Figure
7.
Disc
r
et
e tim
e equ
iva
le
nt
5.
PID MO
D
UL
ATED B
Y
P
WM
The
f
ound
c
onsta
nts
will
be
use
d
in
this
c
ontrolle
r.
I
niti
al
iz
i
ng
var
ia
bles
an
d
res
pecti
ve
ve
ct
or
s
to
ze
r
o
the
ou
t
put
(
)
(
l
oad
v
olta
ge)
s
how
n
i
n
Fig
ure
8.
We
ob
se
r
ve
t
hat
t
he
ov
ersho
ot
a
ppear
s
e
ven
higher
than
pr
e
viously
est
im
at
ed
;
to
redu
ce
the
ove
rsho
ot
,
we
f
ollo
w
Table
1
ta
ke
n
from
[1
3],
s
ho
wing
a
n
onli
ne
tun
i
ng
form
of
P
ID,
de
pendin
g
on
th
e
outp
ut
ch
ara
ct
erist
ic
s
requi
red
by
t
he
us
er
.
T
his
ta
ble
is
valid
as
lo
ng
a
s
the
const
ants
t
o
t
une
a
re
cal
culat
ed
with
a known
m
et
ho
d.
T
he
refor
e
, f
r
om
the v
al
ue
s
t
hat w
e h
a
ve
cal
c
ul
at
ed,
w
e
reduce
=
2400
to
obt
ai
n
a
n
over
sho
ot
cl
ose
t
o
zer
o
a
s
s
how
n
in
Figure
9,
w
hich
is
ver
y
co
nv
enient
f
or
su
c
h
c
onver
te
r
s.
Figure
8.
PID
with P
WM
Figure
9.
Tu
ne
d ou
t
pu
t
(
)
usi
ng Ta
ble
1
Table
1.
E
ffec
ts o
f
I
nde
pende
nt
P
,
I
,
and
D
T
unin
g
PID
Ris
e T
i
m
e
Ov
ersh
o
o
t
Settlin
g
T
i
m
e
Stead
y
Stage
Er
ro
r
Stab
ility
⇑
Kp
Decr
ease
Increase
S
m
all
increase
Decr
ease
Deg
rade
⇑
K
i
S
m
all
decre
ase
Increase
Increase
Lar
g
e
decrease
Deg
rade
⇑
Kd
S
m
all
decre
ase
Decr
ease
Decr
ease
Mino
r
ch
an
g
e
I
m
p
rov
e
6.
BIFUR
C
ATI
ON DI
AGRA
MS
Bi
fu
rcati
on
dia
gr
am
s
are
def
i
ned
as
qual
it
at
i
ve
a
nd
qu
a
ntit
at
ive
va
riat
ions
of
t
he
syst
em
’s
dynam
ic
s
in
fro
nt
of
the
var
ia
ti
on
of
pa
ram
et
ers
that
c
an
in
flue
nce
th
e
sta
bili
ty
of
the
syst
em
and
char
act
e
risti
cs
of
t
he
ou
t
pu
t
,
am
on
g othe
r
iss
ues.
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
PID co
ntr
oller usi
ng rapid
control
prototypi
ng tech
niques
(
Odair A
. T
r
ujil
lo
)
1651
The
bifurcati
on
dia
gr
am
is
a
t
oo
l
to
analy
ze
the
dy
nam
ic
sys
tem
and
pro
vide
s
inf
or
m
at
ion
on
it
li
m
it
s
of stable
op
e
ra
ti
on
. Fo
r
t
his s
yst
e
m
, o
nly t
he
constant
Kp
ha
s to be
var
ie
d as t
his a
ff
ect
s t
he othe
rs.
This
require
s
the in
dustria
l f
orm
o
f
P
ID
:
6.
1
.
Vcand
D
bifu
rc
at
i
on
diag
r
ams
wi
th
Kp
v
ari
at
i
on
s
Figures
10
an
d
11
sho
w
that
t
he
ope
rati
ng
ra
ng
e
of
Kp
has
an
up
per
li
m
it
of
19
unit
s
ap
x
before
the
syst
e
m
b
ecom
e
s ch
a
otic.
Figure
10.
’s bi
furcati
ons
Figure
11.
D
’
s
bif
ur
cat
io
ns
6.2.
and
bif
urca
tion dia
gr
ams
w
ith
var
yi
ng
T
Figure 12 sh
ows that f
or
sm
al
l values of
T
(sam
pling
p
e
riod),
th
e syst
e
m
h
as a one
per
i
od
ic
wav
e
for
high
fr
e
quenci
es;
howe
ve
r,
by
increasi
ng
th
e
sam
pling
per
i
od
e
qu
al
to
t
he
switc
hing
pe
riod
(a
r
ound
110
μ
seg
)
,
the syst
em
s b
egins
to
t
urn
c
ha
otic.
Figure
12.
’s bi
furcati
on d
ia
gra
m
w
it
h
va
ryi
ng T
Figure
13.
D’s
bif
ur
cat
io
n dia
gr
am
w
it
h vary
ing
T
7.
MA
TL
AB/SI
MU
LI
NK
®
SI
MU
L
ATIO
N
The
ne
xt
sim
ul
at
ion
pe
rfor
m
ed
i
n
sim
ulink
m
ov
es
us
cl
os
e
r
t
o
real
-
ti
m
e
i
m
ple
m
entat
ion
pe
rfo
rm
ed
in
dS
P
ACE.
Fig
ures
14
a
nd
15,
r
especti
vely
,
show
t
he
ci
rcu
it
diag
ram
s
and
el
e
m
ents
us
ed
.
Figure
16
disp
l
ay
s
the
ou
t
pu
t
i
n
ope
n
-
an
d
cl
os
e
d
–
lo
op
syst
em
s
wit
h
P
ID.
Re
duci
ng
the
sam
pling
rate
t
o
hal
f
ca
us
es
t
he
sig
nal
to
lose
sy
m
m
e
try
beca
us
e
it
recei
ves
few
e
r
sam
pling
points
an
d,
t
he
refor
e
, t
he
int
egr
at
io
n
pr
oces
s
str
uggles
as
it
ad
ds
m
or
e
than
n
ee
ded o
r
r
est
s
m
or
e
tha
n
nee
de
d,
f
or
t
he
sam
e
reas
on.
T
his
e
xp
la
in
s
t
he
strong oscil
la
ti
on
s
seen
in
Figure
17.
T
his
eff
ect
is
accen
tuate
d
in
the
im
ple
m
entat
ion
.
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.
9
, N
o.
3
,
June
2019 :
1645
-
1655
1652
Figure
14.
PID
and PWM
sc
hem
a
Figure
15.
B
uc
k
c
onve
rter
Figur
e
16.
out
pu
t i
n op
e
n
-
a
nd clo
sed
-
lo
op
Figure
17.
out
pu
t
with a
m
ino
r
sam
ple
fr
e
qu
e
ncy
s
yst
e
m
s w
it
h
P
ID
8.
E
X
PERI
MEN
TAL RES
UL
TS
Fo
r
real
-
ti
m
e
i
m
ple
m
entat
ion
,
the
c
on
tr
ol
a
nd
de
velo
pm
ent
car
d
dSPACE
was
us
e
d.
T
he
conve
rter’s
PI
D
c
ontr
oller
was
im
ple
m
e
nted
in
sim
ulink
as
diag
ram
m
ed
in
Fig
ur
e
18,
w
her
e
it
con
t
ro
ls
the
c
ontr
oller
i
m
ple
m
ented
on
hard
war
e t
hroug
h
the
P
WM
g
e
ner
at
io
n blo
ck.
Figure
18: R
ap
id contr
ol
prot
otypi
ng to
cont
ro
l t
he bu
c
k
c
onve
rte
r
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
PID co
ntr
oller usi
ng rapid
control
prototypi
ng tech
niques
(
Odair A
. T
r
ujil
lo
)
1653
The
res
ults
pr
e
sented
in
this
s
ect
ion
we
re
obta
ined
with
t
he
exp
e
rim
ental
pr
ot
otype
a
nd
t
he
pa
ram
et
ers
def
i
ned
in
Ta
bl
e
2.
By
cha
nging
the
ref
e
ren
c
e
volt
age
to
1
0
V
,
20
V,
a
nd
30
V,
an
d
var
y
ing
the
sam
pli
ng
a
nd
switc
hing
freq
uen
cy
,
we
s
ho
w
t
hat
the
real
syst
e
m
is
m
uch
m
or
e
sensiti
ve
t
o
a
lo
w
sa
m
pl
ing
rate.
Fi
gure
19
corres
ponds
to
a
sam
pling
frequ
e
ncy
of
10
kH
z
.
In
c
reasin
g
the
sam
pling
fr
e
quency
thr
ee
tim
es
(30
kH
z
)
i
n
Figure
20
r
esu
lt
s
in
a
li
tt
le
im
pr
ov
em
ent
in
the
re
gula
te
d
sig
nal
and,
e
sp
eci
al
ly
,
a
re
du
ct
io
n
i
n
t
he
so
un
d
pro
du
ce
d by th
e inducto
r d
ue t
o
the
switc
hi
ng
fr
e
qu
e
ncy
grow
t
h of t
he
tra
ns
ist
or.
By
increasi
ng
sign
ific
a
ntly
th
e
val
ue
of
t
he
capaci
tor
to
376
μ
F
a
s
s
how
n
i
n
Fig
ur
e
21
,
we
im
p
rove
the
c
onve
rter’
s
filt
erin
g
a
nd
fi
nd
a
bette
r
a
nswer
wit
h
a
lo
w
er
ste
a
dy
-
sta
te
error,
e
ve
n
for
the
desig
n
pa
ra
m
et
ers
sh
ow
n
in
Ta
ble
2.
The
reas
on
is
that
m
or
e
capaci
ty
filt
ers
hi
gher
f
reque
ncies
a
nd,
the
r
efore,
the
regu
la
ti
on
is
bette
r,
th
us
r
e
du
ci
ng
c
onside
rab
ly
the
syst
em
’s
chao
s.
It
c
an
be
see
n
in
Figures
22
a
nd
23
t
hat
if
we
m
ake
instant
loa
d
ch
ang
e
s
go
i
ng
f
r
om
14
.
59
Ω
to
39
.
3
Ω
with
a
capaci
tor
of
376
μ
F
a
nd
s
witc
hing
t
o
10
kHz
,
the
n
the
syst
e
m
reg
ulate
s
too
well
with
er
rors
le
ss
than
2%;
he
nce,
no
t
on
ly
the
good
perf
or
m
ance
of
t
he
PI
D
con
t
ro
ll
er m
us
t be c
onside
red.
Table
2.
E
xper
i
m
ental
Pr
oto
t
ype D
at
a
Para
m
eter
Valu
e
:
Load
r
esis
tan
ce
39
.
3
Ω
:
Cap
acitance
46
.
27
an
d
376
:
Ind
u
ctan
ce
2
.
473
:
Internal resist
an
ce
o
f
the so
u
rce
0
.
688
Ω
:
Internal resist
an
ce
o
f
the in
d
u
cto
r
1
.
345
Ω
:
Inp
u
t vo
ltag
e
40
.
086
:
Switch
in
g
f
requ
en
cy
10
k
Hz and
30
k
Hz
:
Sa
m
p
lin
g
f
requ
en
cy
10
k
Hz and
30
k
Hz
Kp
:
Prop
o
rtion
al gain
2
.
7162
Kd
:
Derivativ
e gain
0
.
0011245
Ki:
Integ
ral
g
ain
2400
Fi
gure
19.
out
pu
t
with
=
46
µF
,
Vc
at
10 kHz
F
igure
20.
out
pu
t
with
=
46
µF
,
Vc
at
30 kHz
Figure
21.
out
pu
t
with
=
376
µF
at
10 kHz
Figure
22.
out
pu
t
with l
oad c
hanges
with
=
376
µF
Figure
23.
I
lo
ad
with l
oad c
hanges
with
=
376
µF
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.
9
, N
o.
3
,
June
2019 :
1645
-
1655
1654
9.
CONCL
US
I
O
NS
Be
cause
t
he
in
creased
value
of
the
ca
pacit
or
produce
d
a
s
ub
sta
ntial
i
m
prov
em
ent
in
t
he
syst
e
m
,
it
is
necessa
ry to
propose
a
d
et
ai
le
d st
udy
of the
bu
c
k co
nverte
r
s c
urren
tl
y
on t
he
m
ark
et
i
n o
r
der to
determ
ine and
com
par
e
wh
ic
h
an
d
w
hat
va
lues
of
their
el
e
m
ents
are
m
os
t
appropri
at
e
to
app
ly
with
dif
fere
nt
con
tr
ol
te
chn
iq
ues
.
It
is
li
kely
that
the
i
m
ple
m
ented
conve
rter
does
no
t
hav
e
t
he
be
st
featur
es
t
o
adequate
ly
respond
t
o
PI
D
c
on
tr
ol.
T
he
P
I
D
c
on
t
ro
l
le
r
in
t
he
c
onti
nuous
case
s
hows
good p
er
f
orm
ance
in
te
rm
s
of
r
eg
ulati
ng o
ut
pu
t
vo
lt
age
(
).
I
n
s
om
e
syst
e
m
s,
a
con
ti
nuous
P
ID
c
ontrol
ca
n
react
in
a
way
and
w
hen
is
m
odulate
d
by
P
WM
change
s
ubsta
ntial
ly
;
therefo
re,
we
s
houl
d
no
t
ass
um
e
that
w
hat
is
fou
nd
in
co
ntin
uous
ti
m
e
can
be
app
li
ed
without
re
stric
ti
on
s
i
n
discrete
tim
e.
The
swit
chin
g
a
nd
sam
pling
f
reque
ncies
aff
ect
a
m
a
jor
pro
portio
n
of
the
integral b
eca
use
it
need
s
t
o
ad
d
the h
ig
hest possible
am
ount
of
points
to
suc
cessf
ully
com
plete
the
integ
r
at
ion
process
.
Ha
ving
ha
rdwar
e
th
a
t
sup
ports
high
sam
pling
a
nd
switc
hing
f
requen
ci
es
is
a
ve
ry
im
po
rtant
f
act
or
wh
e
n
im
ple
m
e
ntin
g
a
syst
em
.
F
or
the
im
ple
m
ented
c
onve
r
te
r,
t
he
P
I
D
c
ontr
oller
did
not
res
pond
in
the
best
way.
It
is
ne
cessary
to
filt
er
or
rem
ov
e
the
de
rivati
ve
an
d
sel
ect
appr
opriat
e
sa
m
pl
ing
a
nd
s
witc
hing
fr
e
qu
e
ncies.
ACKN
OWLE
DGME
NTS
This
w
ork
was
su
pp
or
te
d
by
t
he
U
ni
ver
si
da
d
Nacio
nal
de
C
ol
om
bia,
Sede
Me
dellí
n
unde
r
the
pro
j
ect
s
HERMES
-
34
671
a
nd
H
ERM
ES
-
36911.
T
he
auth
or
s
t
hank
the
Scho
ol
of
Ph
ysi
cs
f
or
t
he
ir
valua
ble
s
upport
to
cond
uct
this re
search
.
T
his
re
search
pa
per
c
orres
ponds
to
“
pro
gr
am
a
reconstr
ucció
n
del te
j
ido
so
ci
al
e
n
zo
nas
de
pos
-
c
onflic
to
e
n
C
olo
m
bia
del
proyect
o
M
od
el
o
ec
os
ist
é
m
ic
o
de
m
ejo
r
a
m
ie
nto
r
ur
al
y
co
ns
tr
ucció
n
de
paz:
instal
aci
ón
de
capaci
da
des
lo
cal
es,”
fina
nce
d
by
the
“
Fon
do
Nacio
nal
de
Finan
ci
am
ien
to
par
a
la
Ci
encia,
la
Tecn
ología y
la I
nnov
aci
ón, Fo
ndo Fra
ncis
co
J
os
é
de
Ca
l
das
c
on c
on
trat
o No 2
13
-
2018
con C
ód
i
go 58
960.
”
REFERE
NCE
S
[1]
Yous
efz
ade
h
V
.
Hait
ao
Hu
.
and
Maksim
ovic
D.
,
“
Nonuniform
A
/D
quant
i
zation
for
improved
d
ynamic
responses
of
digi
t
al
l
y
con
trol
le
d
DC
–
DC
conve
rt
ers
,”
Powe
r
E
le
c
tronic
s,
I
EE
E
Tr
a
nsacti
ons
on
,
Vol.
23
,
no.
4,
pp.
1998
-
2005
,
2008.
[2]
DeGus
sem
e
K.
Vande
S
y
p
e
D.M
.
,
Vand
enBossche
A.P.
DeB
el
i
e
F.M.L
.
L.,
and
Melke
be
ek
J.A.
“
Sm
al
l
-
signal
z
-
dom
ai
n
anal
y
s
is
of
dig
it
a
lly
con
trol
le
d
conv
erter
s,”
Powe
r
El
e
c
tronic
s,
I
EE
E
T
rans
act
ions
on
Vol.
21
,
no
.
2
,
pp.
470
-
478
,
2
006.
[3]
Maksim
ovic
D.
and
Za
n
e
R.
,
“
Sm
al
l
-
signal
dis
cre
t
e
-
ti
m
e
m
ode
li
ng
of
dig
it
a
lly
cont
rol
le
d
DC
-
DC
conve
rte
rs
,
”
Computers i
n
Po
wer
Elec
troni
cs
,
2006.
COM
PEL
’06.
IE
EE Works
hops on,
pp.
23
1
-
235
,
2006
.
[4]
Z
ane
R
.
Maksim
ovic
D.
and
Er
icks
on
R.
“
Im
pac
t
of
dig
it
a
l
cont
r
ol
in
power
el
e
c
troni
cs,
”
Power
Semic
ondu
ct
or
Dev
ices and
ICs
,
2004.
Proce
ed
in
gs.
ISP
SD
’04.
T
he
16th
Int
e
rna
t
i
onal
S
y
m
posium
on,
pp.
13
-
22,
2004.
[5]
Maksim
ovic
D.
Prodic
A.
and
Er
ic
kson
R.
W
.
,
“
Digit
a
l
con
trol
l
er
chi
p
set
for
isolated
dc
power
su
ppli
es,
”
Applie
d
Powe
r
E
lectroni
cs
Confe
ren
ce
a
nd
Ex
posi
ti
on
,
2
003.
AP
EC
’03
.
Ei
gh
te
en
t
h
Ann
ual
I
EE
E
Vol
.
2,
pp
.
866
-
872
,
2003.
[6]
Pete
rch
ev
A.V.
a
nd
Sander
s
S.R.
,
“
Quanti
zation
r
e
soluti
on
and
li
m
i
t
c
y
c
li
ng
in
dig
it
al
l
y
cont
rol
le
d
p
wm
conve
rte
rs,
”
Powe
r E
le
c
troni
cs,
IE
EE Tr
ansacti
ons on
,
Vol
.
1
8
,
no.
1
,
Part
:
2
,
pp
.
301
-
308
,
2003.
[7]
M.
Habb
ab,
A.
Haz
za
b
,
and
P.
Sica
rd.
,
“
Rea
l
T
ime
Im
ple
m
enta
ti
on
of
Fuzz
y
A
dapt
iv
e
PI
-
slid
in
g
Mode
Contro
ller
for
Indu
ct
io
n
Mac
hin
e
Contro
l
,”
Int
ernati
onal
Journal
of
El
e
ctr
ic
al
and
Comp
ute
r
Engi
n
ee
rin
g
(
IJE
CE)
,
Vol.
8
,
No.
5,
pp.
2884
-
2894
,
Octobe
r
2
018.
[8]
E.
Rakhman,
Fer
i
y
on
ika.
,
“
Distribut
ed
Con
trol
Sy
stem
Appl
ie
d
i
n
T
empera
tur
C
ontrol
b
y
Coo
rdi
nat
ing
Multi
-
loo
p
Con
trol
le
r
,
”
Te
le
communic
a
ti
o
n
Computing
El
e
ct
ronics
an
d
Control
(
TEL
KOMNIKA
)
,
Vol.
16,
No.
4,
pp.
1568
-
1576
,
Augus
t
2018.
[9]
A.
Pra
y
it
no
,
V.
Indra
wati,
I
.
I.
Trusula
w,
“
Fuzz
y
Ga
in
Schedu
ling
PID
Contro
l
for
Pos
it
ion
of
the
AR.Dron
e,”
Inte
rnational
Jo
urnal
of
Elec
trical
and
Computer
Engi
n
ee
ring
(
IJ
ECE
)
.
Vol
.
8
,
N
o.
4
,
pp
.
1939
-
19
46,
Augus
t
2018.
[10]
F.
E.
Ho
y
os,
J
.
E.
C
ande
lo
,
an
d
J.
A.
T
abor
da
,
“
Sele
c
ti
on
an
d
Vali
da
ti
on
of
Mathe
m
at
i
cal
Models
of
Pow
er
Convert
ers
usin
g
Rapi
d
Modeling
and
Control
Protot
y
p
ing
Methods,
”
In
te
rna
ti
onal
Journal
of
El
e
ct
ri
cal
an
d
Computer
Engi
n
ee
ring (
IJE
C
E)
,
Vol.
8
,
No.
3,
pp
.
1551
-
1568
,
Jun
e
2018.
[11]
Prodic
A.
Hao
P,
Maksim
ovic
D
.
and
Al
arc
on
E.
,
“
Modeli
ng
of
q
uant
i
za
t
ion
ef
fects
in
dig
it
a
lly
co
ntrol
le
d
DC
-
DC
conve
rt
ers,
”
Po
wer
Elec
troni
cs,
IEE
E
Tr
ansacti
o
ns on
,
Vol
.
18
,
No.
1
,
Par
t. 2, p
p.
301
-
308
,
20
03.
[12]
Li
u
C.
P
.
Fung
C.
W
.
and
Pong
M.H.
“
A
dia
gr
a
m
m
at
ic
appr
oa
c
h
to
sea
rch
for
m
ini
m
um
sa
m
pli
ng
fre
qu
ency
a
nd
quant
i
za
t
ion
res
olut
ion
for
dig
ital
con
trol
of
po
wer
conve
rt
ers
,”
Powe
r
Elec
tro
nic
s
Speciali
sts
Confe
renc
e
,
200
7.
IEE
E
,
pp
.
826
-
832
,
2007
.
[13]
Gregor
y
Chong.
Kiam
Heong
Ang.
and
Yun
Li.
,
“
PID
cont
ro
l
s
y
stem
a
na
l
y
sis
,
design,
and
tech
nolog
y
,
”
I
EEE
Tr
ansacti
ons on control
syst
ems t
ec
hnolog
y
,
Vol.
13,
No.
4,
Jul
y
2
005.
[14]
E.
Fos
sas
and
G.
Oliva
r
,
“
Stud
y
o
f
ch
aos
in
th
e
buck
conv
ert
e
r,
”
I
EEE
transa
ct
ions
on
Circuits
and
Syst
ems
-
I:
Fundament
al
Theory
and
Applic
ati
ons
,
43
,
13
–
2
5,
1996
.
Evaluation Warning : The document was created with Spire.PDF for Python.