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.
11, N
o.
1, Mar
ch 20
20,
p
p.
235~
2
4
1
IS
S
N
: 2088-
86
94,
D
O
I
:
10.11
59
1
/ij
ped
s
.
v11
.
i
1.pp
2
35-
24
1
235
Jou
rn
a
l
h
o
me
pa
ge
:
ht
tp:
//
i
j
ped
s
.
i
ae
sc
ore.
c
o
m
R
e
duction of tr
a
nsients i
n
swi
tches usi
n
g embedded machi
ne
learn
i
ng
P.
Suresh
1
,
S.
G
eo
rg
e
Fe
rna
n
de
z
2
,
S. Vid
y
a
s
agar
3
,
V
.
K
a
l
ya
na
su
nd
a
r
a
m
4
,
K
.
V
i
j
a
yak
uma
r
5
,
V
a
i
d
he
eswar
a
n
Ar
chan
a
6
, S
oh
am C
hatter
j
ee
7
1
,
2,
3,
4,
5
Faculty of
E
l
ec
t
r
ical and E
l
ect
ronics Eng
i
n
eeri
ng,
SR
M
Institu
te of S
c
ience
and
Techn
o
l
o
g
y
, In
di
a
6
D
epartm
ent of
El
ectri
cal an
d
E
l
ect
ro
ni
cs E
ng
in
e
e
ring
,
N
a
tio
na
l
u
n
i
v
e
rsity
o
f
Si
ngapore,
S
in
gapo
re
7
D
epartm
ent of
El
ectro
n
i
cs an
d
Co
m
m
u
n
i
cati
on Eng
i
n
eering, N
a
n
y
a
n
g
Tech
nical
Univer
s
i
t
y
, S
ing
a
p
o
re
Art
i
cl
e In
fo
ABSTRACT
A
r
tic
le hist
o
r
y
:
Re
ce
i
v
e
d
Mar
1
,
201
9
R
e
v
i
s
e
d
Jul
1
,
2
019
Ac
ce
p
t
ed
Oc
t
2
3
,
2
019
No
n-lin
e
a
r
lo
ads
can
c
au
se
t
rans
ien
t
s
in
e
l
ectro
nic
s
w
it
ches
.
T
hey
al
so
r
esu
lt
in
a
f
luct
ua
t
i
n
g
out
pu
t
when
t
he
d
evi
ce
i
s
s
witch
e
d
ON
or
O
F
F
.
Thes
e
tran
si
ents
can
h
arm
no
t
o
n
ly
t
he
s
w
i
t
c
hes
but
a
l
s
o
the
d
e
vi
ces
that
t
hey
are
con
n
ect
ed
t
o,
b
y
p
a
ss
in
g
excess
cu
rrents
o
r
vol
tages
t
o
t
h
e
d
ev
ices
.
By
app
l
y
i
ng
m
ach
in
e
l
earnin
g
,
we
can
i
m
p
rov
e
t
he
g
ate
dri
v
e
vo
lt
ag
es
o
f
th
e
sw
it
ches
a
n
d
t
h
e
reby
r
e
d
u
c
e
s
witch
trans
i
en
ts.
A
f
eedback
s
ys
te
m
i
s
built
th
at
m
easu
r
es
t
he
o
utp
u
t
tran
s
i
e
n
ts
a
nd
t
h
e
n
feeds
it
t
o
a
n
e
ur
al
n
et
work
alg
o
rithm
t
h
at
t
h
e
n
gi
ves
a
p
r
ope
r
g
a
te
d
ri
ve
t
o
th
e
devi
ce.
T
h
is
w
i
l
l
reduce
tran
si
ents
a
nd
a
ls
o
im
pro
v
e
p
e
rf
o
r
m
a
nces
o
f
s
w
it
ch
b
ased
d
ev
ice
s
like
in
vert
ers
an
d
convert
ers.
K
eyw
ord
s
:
Emb
e
dd
e
d
s
yste
ms
Mac
h
i
n
e
lear
n
i
ng
ne
ura
l
netw
orks
So
l
i
d
st
at
e
swi
t
ch
es
Tran
si
ent
s
Th
is
is a
n
o
p
en acces
s a
r
ti
cle u
n
d
e
r t
h
e
CC
B
Y
-S
A
li
cens
e
.
Corres
pon
d
i
n
g
Au
th
or:
P
.
S
uresh,
D
e
pa
rtme
nt
o
f
El
e
c
t
rica
l
a
n
d
El
ect
ro
ni
c
s
Eng
in
e
e
ring
,
SRM Inst
i
t
u
t
e
of
S
cie
n
ce
an
d T
ech
no
l
o
g
y
,
Ka
t
t
a
n
kul
at
hu
r, C
h
e
nn
a
i
, In
d
i
a
.
Em
ail:
sur
e
sh.
a
u95
@
g
m
a
i
l
.c
om
1.
I
N
TR
OD
U
C
TI
O
N
Tr
ansie
n
ts
a
r
e
r
andom
a
n
d
rapid
f
l
uc
tua
t
i
o
ns
i
n
vo
l
t
a
g
es
c
a
u
se
d
b
y
c
h
a
n
gi
n
g
l
o
a
ds
a
n
d
f
a
s
t
sw
it
c
h
i
n
g
de
vi
c
e
s.
T
r
a
nsien
t
s
a
r
e
ve
ry
h
ar
mful
i
n
a
c
i
rc
ui
t
bec
a
u
se
t
h
e
y
ca
n
cau
se
h
eat
i
ng
a
nd
i
m
pr
oper
firi
n
g
o
f
so
li
d
st
a
t
e
sw
itc
he
s
.
T
ransients
ca
n
also
c
a
u
se
p
oor
pow
er
f
a
c
t
or,
inc
r
ea
se
c
u
rre
nt
i
n
the
ne
utra
l
con
d
u
ct
or,
i
n
c
r
e
a
se
l
osses
d
u
e
t
o
h
ys
tere
si
s
a
n
d
e
d
dy
c
urre
nt
i
n
mot
o
rs
a
nd
i
t
ca
n
a
l
s
o
s
om
et
ime
s
i
n
t
er
fer
e
w
ith
t
ele
p
ho
ne
n
e
t
w
o
rks.
T
ra
n
s
ien
t
s
a
r
e
very
s
teep
v
o
l
ta
g
e
s
te
p
s
t
h
at
o
cc
u
r
i
n
e
l
ec
t
r
i
cal
c
i
r
cu
i
t
s
d
u
e
t
o
th
e
su
dd
en
r
el
ea
se
o
f
a
p
r
e
v
i
o
u
s
l
y
s
t
o
red
en
ergy
,
e
i
t
h
er
i
nd
u
c
t
i
v
e
or
capa
c
i
t
i
ve
,
w
h
ic
h
resu
lt
s
i
n
a
h
i
gh
v
o
l
t
a
g
e
trans
i
en
t,
o
r
su
rge
be
ing
cre
a
t
e
d.
T
h
i
s
s
u
d
d
e
n
r
elea
se
o
f
e
n
erg
y
b
a
c
k
i
n
t
o
t
h
e
c
i
rc
uit
du
e
t
o
s
o
m
e
swit
ch
i
ng
ac
t
i
o
n
cre
ates
a
t
ra
nsie
nt
v
olt
a
ge
s
pi
ke in t
h
e
form
o
f a s
t
e
e
p
i
mp
ul
se
o
f
e
n
ergy
w
hi
ch
can
i
n
t
h
eo
ry
b
e
o
f
a
ny
in
fini
te va
l
u
e
.
We
m
ust a
l
s
o
r
ealiz
e
t
h
at vo
l
t
a
ge tran
s
ie
n
t
s do
not a
lw
ays
sta
r
t a
t
zer
o v
o
lts or a
t
t
he
b
e
g
i
nni
n
g
of
a
c
yc
le,
b
u
t
c
a
n
be
s
upe
r
i
mpose
d
o
n
t
o
a
n
o
t
her
v
o
l
t
a
ge
l
e
v
e
l
.
E
it
her
w
a
y,
t
r
a
ns
ie
nts
ar
e
b
a
d
a
s
t
h
e
y
c
a
n
dam
a
ge
e
lec
t
r
o
n
i
c
e
qui
pm
ent
a
n
d
t
h
ere
f
o
r
e
ne
ed
t
o
be
s
u
p
pr
esse
d
o
r
contr
o
l
l
e
d
.
Tr
an
sient
s
uppression
dev
i
ce
s
can
t
a
k
e
on
m
a
ny
fo
rm
s
from
a
r
c
c
on
t
a
c
t
s,
t
o
fil
t
e
r
s,
t
o
s
o
li
d
s
t
at
e
sem
i
c
o
n
duc
t
o
r
de
vic
e
s
.
D
i
s
c
r
e
te
sem
i
c
o
nd
uc
tor
trans
i
e
n
t
s
u
pp
ressio
n
d
ev
ice
s
s
uc
h
a
s
t
h
e
M
e
t
al-o
xide
V
aris
tor,
o
r
MOV,
a
re
by
f
a
r
the
m
o
st
c
o
mmo
n
as
t
hey
are
av
ai
l
a
ble
i
n
a
v
a
r
i
e
t
y
o
f
en
e
r
g
y
ab
so
rbi
n
g
a
nd
vo
l
t
age
ra
ti
ngs
m
aki
n
g
i
t
pos
si
ble
to
exe
r
cise
t
i
g
h
t
c
ontro
l
ove
r
u
n
w
a
nte
d
a
nd p
o
t
ent
i
a
l
l
y
d
es
truc
ti
v
e tra
n
s
i
e
n
t
s
o
r ove
r vo
lta
ge
s
pike
s.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
nt
J
P
o
w
E
l
e
c
&
D
r
i
S
yst
V
o
l.
11,
N
o.
1
,
Mar
202
0
:
235
–
24
1
23
6
The
most
c
om
m
on
tr
ans
i
e
n
t
is
t
he
“
osc
i
l
l
a
t
or
y
t
r
ans
i
e
n
t
”
.
I
t
i
s
s
o
m
et
i
m
e
s
d
esc
r
i
b
ed
a
s
“
r
i
n
gin
g
tr
a
n
sie
n
t
”
.
Th
i
s
t
y
p
e
of
t
r
a
ns
ien
t
s
is
c
ha
r
a
c
t
e
r
ize
d
by
sw
in
gs
ab
o
v
e
a
n
d
bel
o
w
t
h
e
nor
ma
l
line
vo
l
t
a
g
e.
T
he
ot
her
ty
pe
(
im
pu
l
s
e
)
t
r
a
ns
i
e
n
t
g
e
n
e
r
ate
d
i
n
ind
u
c
t
i
o
n
mo
t
o
r
[1]
,
is
m
or
e
e
a
si
ly
e
x
p
l
a
i
ne
d
as
a
“
one-
s
h
o
t
”
t
y
pe
of
e
ve
n
t
,
an
d
it
i
s
c
h
a
r
ac
t
e
r
i
ze
d
by
h
a
v
in
g
m
o
r
e
t
ha
n
7
7
%
of
i
t
b
e
i
n
g
one
p
ul
s
e
a
bo
ve
t
h
e
line
v
o
lta
ge
[
2]
.
A
li
g
h
tn
in
g
s
t
ri
k
e
c
an
b
e
c
o
m
p
ose
d
o
f
mu
lt
ip
le
t
rans
ien
t
s
o
f
t
his
t
ype
.
I
t
i
s
beca
use
of
a
l
l
t
he
se
p
r
o
b
l
em
s
t
h
at
re
se
arc
h
ers
are
t
r
y
i
ng
t
o
f
i
nd
w
a
y
s
to
r
e
d
u
c
e
t
r
an
si
ent
s
i
n
c
i
r
cu
its.
One
of
t
he
w
a
y
s
i
s
b
y
hav
i
ng
filte
rs
i
n
t
h
e
c
i
r
c
u
its
t
ha
t
c
a
n
r
educ
e
a
nd
s
m
oothe
n
the
ef
f
ects
of
t
hese
t
r
a
n
sie
n
t
s
.
Howeve
r,
f
ilter
s
a
re
t
y
p
ica
lly
l
ar
g
e
a
nd
the
y
a
ls
o
c
ons
ume
a
l
o
t
o
f
p
ow
e
r
a
nd
ha
ve
l
osse
s
due
t
o
he
at.
A
b
ett
e
r
me
t
h
od
t
o
re
du
c
e
t
ran
s
i
e
nt
s
i
s
by
gi
v
i
n
g
p
r
o
pe
r
ga
t
e
p
u
l
ses
to
t
he
s
o
l
i
d
s
t
a
te
s
w
itc
he
s
l
i
ke
I
G
B
T
s
a
n
d
M
O
SFETs.
Howe
ver,
it
is
d
iffi
c
u
lt
t
o
ju
d
g
e
w
h
a
t
t
he
p
r
ope
r
sw
i
t
c
h
i
ng
vo
l
t
a
g
es
h
a
v
e
to
b
e
a
n
d i
t
i
s a
l
s
o
di
ff
ic
ul
t
to
d
e
s
ig
n
a
co
ntr
o
lle
r
t
h
at
c
a
n
r
e
act
q
u
i
c
k
l
y
en
ough
t
o
c
h
a
n
g
e
t
he
s
wi
t
c
h
i
ng
vol
t
a
g
e
b
ased
o
n
t
h
e
i
n
p
u
t
v
olt
a
ge
s
[
3
]
.
These
a
r
e
used
i
n
v
a
r
i
o
u
s
a
p
p
lica
tio
ns su
c
h
a
s
e
lec
t
r
i
c v
e
hic
l
e c
h
ar
gi
n
g
a
nd i
n
duc
t
i
o
n
c
o
ok
in
g a
p
pl
i
c
at
i
o
n
s
[
4-
7]
.
O
ur
appr
oa
ch
is to
u
se
a
Ne
u
r
al
N
et
wo
rk
b
ased
c
o
n
t
r
oll
e
r
t
h
at
c
a
n
qui
c
k
ly
r
e
a
c
t
to
t
h
e
v
o
lta
ge
t
r
a
ns
ie
nt
s
a
nd
c
a
n
r
e
d
u
c
e
the
tr
a
n
sie
n
t
s
b
y
g
i
vin
g
p
r
o
per
swit
c
h
i
n
g
ga
te
s
i
g
nal
s
.
2.
LITERATURE REVIEW
Tra
n
sie
n
t
a
c
t
i
v
it
y
is
b
el
ie
ve
d
to
a
cc
oun
t
f
o
r
8
0
%
of
a
l
l
elec
t
r
i
c
a
l
l
y
-re
l
a
t
e
d
d
o
wn
ti
m
e
.
Lig
h
tn
in
g
a
ccou
n
t
s
a
t
l
e
a
st
5
%
of
I
ns
ur
anc
e
c
laims
a
nd
c
o
s
t
s
a
n
a
ver
a
ge
o
f
$
13,
00
0
pe
r
occ
u
r
r
e
nce
[
8
]
.
E
ff
e
c
ti
ve
tr
a
n
sie
n
t
v
o
l
ta
ge
s
u
ppr
e
s
s
i
o
n
e
quipm
en
t
can
doub
le
o
r
t
r
i
p
l
e
t
h
e
l
i
f
e
of
e
l
ect
ri
c
a
l
an
d
el
ect
ron
i
c
e
qui
p
m
e
n
t
.
Tr
ansie
n
ts
a
r
e
v
e
r
y
s
t
e
e
p
v
olt
a
ge
s
t
e
ps
t
hat
occur
i
n
e
lec
t
r
i
c
a
l
c
i
r
c
ui
ts
due
t
o
t
h
e
s
u
dd
e
n
r
elea
se
o
f
a
pr
ev
io
u
s
ly
s
t
o
r
e
d
e
n
er
gy,
e
i
t
her
i
n
d
u
c
t
i
v
e
or
c
apa
c
iti
ve
,
w
h
i
c
h
r
esu
lts
i
n
a
hi
g
h
v
o
lta
ge
t
r
a
ns
ien
t
,
o
r
s
u
r
ge
b
e
i
n
g
c
r
e
a
t
e
d
.
T
h
i
s
s
udd
e
n
r
el
ea
se
o
f
en
erg
y
b
a
ck
i
nt
o
th
e
ci
rc
ui
t
d
u
e
to
s
ome
sw
i
t
c
h
i
n
g
act
i
on
c
r
ea
tes
a
tr
a
n
sie
n
t
vo
l
t
a
g
e
s
p
i
k
e
i
n
t
he
f
or
m
of
a
s
tee
p
i
m
p
u
l
se
o
f
e
n
er
g
y
w
h
ic
h
c
a
n
i
n
t
he
or
y
be
o
f
a
n
y
i
n
f
i
nite
v
a
l
ue
.
Tr
a
n
si
ent
suppre
ssi
on
d
e
v
i
c
e
s
c
an
t
a
k
e
o
n
m
an
y
fo
rms
fro
m
a
r
c
c
ont
a
c
ts,
t
o
f
ilter
s
,
to
s
o
l
i
d
s
t
a
t
e
s
em
i-
c
o
n
duc
t
o
r
dev
i
ce
s.
D
i
s
c
r
ete
se
mi
c
o
n
d
u
c
t
or
t
r
a
ns
i
e
nt
s
up
p
r
essi
o
n
de
vice
s
suc
h
a
s
the
Me
tal-o
x
i
de
V
a
r
istor
,
or
M
O
V
,
a
r
e
b
y
fa
r
the
mo
st
c
om
mon
as
t
he
y
ear
a
va
i
l
a
ble
i
n
a
v
a
r
ie
ty
o
f
e
n
er
g
y
a
bsor
bin
g
a
nd
v
o
l
t
a
ge
r
a
tings
m
ak
i
ng
it
pos
si
b
l
e
t
o
e
xer
c
ise
tig
ht
c
on
tr
o
l
o
ver
u
n
w
a
n
ted
a
n
d
p
o
te
n
tia
l
l
y
d
e
s
tr
uct
i
ve
t
r
a
n
s
i
e
n
t
s
or
ove
r
v
o
l
t
a
g
e
sp
i
k
e
s
.
F
igur
e
1
show
s
t
h
e
me
thods
o
f
r
e
duc
t
i
o
n
o
f
t
ra
nsi
e
nt
s.
F
i
gur
e
1.
P
r
evi
ous
m
et
ho
ds
o
f
r
e
duc
t
i
o
n
o
f
tr
a
n
sie
n
t
s
D
i
ver
t
i
ng
a
tr
a
n
sie
n
t
is us
u
al
l
y
a
cc
om
pl
i
s
he
d
u
s
i
ng
a vo
l
t
a
g
e-
c
l
a
mp
in
g ty
pe
d
e
v
ic
e or
b
y
u
s
in
g
w
h
a
t
a
r
e
com
m
onl
y
c
a
lle
d
a
crowb
a
r
type
d
e
v
ice.
T
hese
p
ar
al
lel
con
n
e
c
t
e
d
d
e
v
i
c
es
e
x
h
i
b
it
a
non
l
i
near
i
m
p
e
d
anc
e
ch
a
r
ac
t
e
ri
sti
c
a
s
t
h
e
c
u
rren
t
f
l
o
win
g
t
h
r
o
ugh
t
h
em
i
s
not
l
i
n
ea
r
to
t
h
e
v
olt
a
g
e
a
cross
th
eir
termin
als
as
g
iv
en
by
O
h
m
s
Law
.
A
v
o
l
tage-
c
la
m
p
i
n
g
de
vice
s
uc
h
as
a
n
M
O
V
,
h
as
a
v
ar
ia
b
l
e
impe
da
n
c
e
de
pen
d
i
n
g
on
t
h
e
c
u
r
r
e
nt
f
low
i
n
g
t
hr
ou
g
h
t
he
d
ev
ice
or
o
n
t
h
e
v
o
lta
ge
acr
oss
i
t
s
t
er
mina
l.
U
nde
r
n
o
r
m
al
s
tea
dy-
s
t
ate
o
p
e
r
a
t
i
n
g
c
o
n
d
it
i
o
n
s
,
the
dev
i
ce
o
ffers
a hi
g
h
im
p
ed
a
n
c
e
a
nd ha
s t
h
ere
f
or
e
no
effe
ct o
n
the c
o
n
n
ec
te
d
circ
uit.
H
o
w
e
ve
r
,
w
hen
a
vo
l
t
age
tr
ans
i
en
t
oc
c
u
r
s
,
the
im
pe
da
nc
e
of
t
he
de
v
i
c
e
c
ha
nge
s
i
n
cr
e
a
si
ng
the
c
u
r
r
e
nt
d
r
a
w
n
t
hr
ou
g
h
t
he
d
e
v
ice
as
t
he
v
o
lta
ge
a
cr
os
s
it
r
ises
.
The
r
e
sul
t
i
s
an
a
p
p
ar
ent
c
l
am
p
i
n
g
o
f
th
e
t
r
an
si
ent
v
o
lt
a
g
e
.
T
h
e
vol
t
-
amp
e
re
c
h
a
ra
c
t
eri
s
t
i
c
o
f
a
cl
a
m
p
i
ng
d
ev
ice
s
i
s
ge
ner
a
lly
t
im
e-
depen
d
e
n
t
as
t
he
lar
g
e
i
n
c
r
ea
se
i
n
c
u
r
r
e
nt
r
e
s
ul
t
s
i
n
t
h
e
dev
i
ce
d
i
s
si
pa
ti
n
g
a
l
o
t
of
ener
g
y
.
C
r
ow
bar
de
v
i
c
e
s
ar
e
ano
t
her
ty
pe
o
f
t
r
ans
i
e
n
t
su
ppr
ess
i
on
de
vic
e
w
h
ic
h
di
ver
t
s
o
v
er
v
o
lta
ge
s
p
i
kes
a
w
ay from
a
c
i
r
c
ui
t
as a r
esu
l
t of a sw
itc
h
i
ng
ty
p
e
tur
n
-on
a
c
t
i
on.
C
r
ow
ba
r
dev
i
ce
s ar
e
si
m
i
l
a
r
i
n
o
per
a
tio
n
t
o
a
Ze
ner
di
o
d
e
i
n
t
ha
t
u
nder
n
o
r
m
al
s
t
e
a
d
y-
sta
t
e
c
o
n
d
i
t
i
ons
t
he
y
h
a
v
e
n
o
e
f
f
e
c
t
o
n
t
h
e
c
i
r
c
u
i
t
.
W
h
e
n
a
tr
a
n
sie
n
t
is
d
e
t
ec
t
e
d,
t
he
y
r
a
p
i
d
l
y
sw
i
t
ch
O
N
off
e
r
i
n
g
a
v
er
y
l
ow
i
m
p
ed
anc
e
p
a
t
h
wh
ic
h
d
i
verts
t
h
e
trans
i
en
t
a
w
ay
f
r
o
m
the
par
a
l
l
e
l
-
c
on
ne
c
t
ed
l
oa
d.
T
r
a
ns
ien
t
s
o
n
a
n
A
C
p
ow
e
r
lin
e
c
an
r
ang
e
from
a
few
v
o
lts
t
o
o
v
e
r
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
Red
u
ct
i
on
o
f
tr
ans
i
e
n
t
s i
n
s
w
i
t
che
s
usin
g em
bed
d
e
d
m
a
ch
in
e lear
nin
g
(P.
Sur
esh)
23
7
se
v
e
ra
l
k
i
l
o
-vo
l
t
s
a
bov
e
th
e
no
rma
l
m
ai
ns
vol
t
a
g
e
.
S
u
pp
re
ssi
on
dev
i
c
e
s
w
hich
a
tte
nua
t
e
o
r
b
l
oc
k
t
h
ese
t
r
a
n
si
e
n
t
u
s
e
fil
t
e
r
ci
rc
ui
ts
t
o
ef
f
ect
iv
ely
el
i
m
i
n
a
t
e
th
es
e
ma
ins
bor
n
t
r
ans
i
en
ts
b
y
i
n
se
r
ting
a
1
0
0
H
z
f
i
l
t
er
i
n
series
w
ith
t
h
e
c
on
n
e
c
t
ed
l
oa
d.
T
he
freq
u
e
n
cy
c
om
p
o
n
en
t
of
a
f
a
st
s
wit
c
h
i
ng
vol
t
a
g
e
t
ra
n
s
i
e
nt
c
a
n
b
e
mu
c
h
hi
ghe
r
tha
n
t
he
s
low
mov
i
n
g
f
u
n
d
am
enta
l
fr
eque
nc
y
o
f
t
he
A
C
sou
rc
e.
T
hus,
an
o
b
v
io
u
s
c
ho
ic
e
t
o
a
t
t
e
nua
te
a
n
d
co
nt
rol
th
ese
un
wa
nt
ed
t
ra
n
s
i
e
nt
s
i
s
t
o
us
e
a
l
o
w-p
a
ss
f
i
l
t
er
s
ec
t
i
o
n
b
e
t
we
en
t
he
s
o
u
rc
e
an
d
t
h
e
l
o
ad
.
Low
p
a
ss
fi
lt
e
r
s,
s
uc
h
as
a
n
LC
f
il
t
e
r,
c
an
b
e
use
d
t
o
att
e
nu
a
t
e
an
y
hig
h
f
req
u
e
n
c
y
t
ra
nsie
n
t
s
a
nd
a
llow
the
l
o
w
-
fre
que
nc
y
pow
er
o
r
si
gna
l
t
o
p
a
s
s
t
h
ro
ug
h
un
d
i
s
t
urbe
d.
T
he
s
imp
lest
f
orm
of
t
r
a
ns
ien
t
s
up
press
i
o
n
f
i
l
ter
is
tha
t
o
f
a
re
sis
t
or-c
apac
i
t
or
RC
filter
pl
ac
ed
d
ire
c
tl
y
a
c
ross
t
he
pow
er
l
i
n
e
to
a
t
t
e
nua
te
a
n
y
h
ig
h
fre
que
nc
ies
tra
n
s
i
en
ts.
3.
CIRCUIT
The
Vo
l
t
age
M
e
as
urem
e
n
t
Circui
t
is
u
sed
t
o
m
e
a
s
u
re
t
h
e
o
u
t
pu
t
v
ol
ta
ge
a
c
r
oss
t
h
e
s
w
itc
h.
T
h
i
s
is
the
n
f
e
d
i
n
t
o
t
h
e
m
i
cro
co
ntr
o
l
l
er.
The
M
i
c
r
o
con
t
ro
ll
er
i
s
a
R
a
s
pbe
rry
P
i
t
h
at
h
as
a
n
e
u
r
a
l
ne
tw
ork
ins
i
de
it.
The
ne
ur
al
n
e
t
w
o
rk
t
ake
s
a
s
in
p
u
t
the
o
u
t
p
u
t
of
t
h
e
V
ol
ta
ge
S
e
n
s
i
ng
E
lem
e
nt.
I
t
t
he
n
pre
d
ic
ts
t
he
v
a
l
ue
.
F
i
gure
2
an
d
F
i
g
u
re
3
s
h
o
w
t
h
e
so
ft
s
w
i
tc
hi
ng
by
u
s
i
n
g
high val
ue
g
a
t
e
re
sis
t
or
o
f
the
g
a
te
d
ri
ver
vo
l
t
a
g
e
an
d
the b
l
ock
d
i
a
g
ram
of tra
nsie
n
t
r
educ
tio
n sys
t
em
usin
g
ML.
F
i
gure
2.
S
oft
sw
itc
h
i
n
g
b
y
u
s
i
ng
hi
gh
v
a
l
u
e
ga
t
e
re
s
i
s
t
o
r
of t
he
g
a
t
e
dri
v
e
r
volta
g
e
suc
h
t
h
at t
he
t
r
a
ns
ie
n
t
s
ac
ross the
sw
it
ch
w
ill
be r
edu
c
e
d
w
hic
h
i
n
t
u
r
n
incre
ases
q
u
a
li
ty
p
o
w
e
r
[
9
]
F
i
gure
3.
B
l
o
c
k
d
i
a
gram
of trans
i
en
t r
e
duc
t
i
on sy
stem
us
i
n
g
M
L
The
o
u
t
p
u
t
o
f
the
ne
ural
n
etw
o
rk
w
i
l
l
be
acr
oss
t
h
e
GPI
O
p
i
n
s
o
f
the
Ras
pberr
y
P
i
.
Th
i
s
o
u
t
p
u
t
w
i
l
l
be
i
n
t
h
e
f
o
rm
o
f
a
volt
a
g
e
.
Th
e
ou
t
put
i
s
th
en
f
e
d
i
nto
t
h
e
G
a
t
e
D
riv
e
r
c
i
rcu
it
of
t
he
i
n
v
er
ter.
T
he
re
a
re
t
w
o
m
ode
s
i
n
t
h
e
ci
rcu
it:
Mod
e
1:
In
t
hi
s
mod
e
,
a
Hi
gh
pul
s
e
i
s
g
i
v
e
n
to
t
h
e
f
i
r
s
t
a
nd
f
ou
rth
I
G
B
T
s
i
n
the
inve
rter
c
ir
cui
t
.
The
D
C
c
urr
e
n
t
f
low
s
[1
0]
t
hro
u
gh
th
e
first
IG
BT,
t
h
e
loa
d
a
nd
t
h
e
n
t
h
r
o
u
g
h
t
he
f
o
u
r
t
h
I
G
B
T
.
T
h
e
v
o
l
t
a
g
e
o
u
t
p
u
t
t
h
a
t
w
e
g
e
t
i
s
t
h
e
pos
it
ive
hal
f
o
f
the
S
qua
re
w
a
v
e
A
C
o
u
t
p
u
t
.
The
pu
l
s
e
s
a
c
r
o
ss
t
h
e
o
t
h
e
r
t
w
o
I
G
B
T
s
a
r
e
l
o
w
a
n
d
t
h
e
y
a
r
e
i
n
the sw
i
t
c
h
e
d
O
F
F
c
ond
i
t
i
on w
h
e
r
e
t
h
e
y
do
not
l
e
t
an
y
c
urre
nt pa
ss t
h
ro
ugh.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
nt
J
P
o
w
E
l
e
c
&
D
r
i
S
yst
V
o
l.
11,
N
o.
1
,
Mar
202
0
:
235
–
24
1
23
8
Mo
d
e
2
:
I
n
t
hi
s
m
ode,
a
H
i
g
h
P
u
l
se
i
s
g
i
ve
n
to
I
G
B
Ts
t
w
o
a
n
d
t
hr
ee
.
I
n
t
h
i
s
case
th
e
c
u
rre
n
t
f
l
ows
t
h
ro
ugh
I
GB
T
2
fi
rst
an
d
th
en
a
c
r
o
ss
t
h
e
l
o
a
d
a
nd
t
h
e
n
IGBT
3
.
In
t
hi
s
c
a
s
e
,
t
he
o
u
t
p
u
t
vo
l
t
age
ac
r
o
ss
t
he
L
oa
d
is
r
eve
r
sed
a
nd
w
e
g
e
t
t
he
s
ec
on
d
ha
lf
o
f
the
o
u
t
p
ut
a
c
r
oss
the
lo
ad.
T
h
e
o
t
h
e
r
I
G
B
T
s
a
r
e
i
n
t
h
e
sw
i
t
che
d
O
F
F
c
ondi
ti
o
n
an
d
th
ey
d
o
not
l
e
t
a
ny
c
u
r
r
e
nt
t
h
r
oug
h
.
C
a
r
e
mu
st
b
e
t
a
k
e
n
no
t
t
o
sw
it
c
h
O
N
tw
o
I
G
BT
s
in
t
he
s
am
e
l
e
g,
s
o
a
s
t
o
n
o
t
c
a
use
a
shor
t
cir
c
u
i
t.
4.
MACH
INE
L
E
ARNING
We
u
se
N
e
u
ral
Networ
ks
t
o
mode
l
the
tran
sie
n
ts
i
n
the
s
w
itc
hes
[
11]
.
The
ne
ur
a
l
n
e
t
w
o
r
k
i
s
t
r
a
i
n
e
d
us
in
g
o
p
e
n
s
ou
r
c
e
tr
a
i
nin
g
dat
a
tha
t
i
s
f
r
e
e
l
y ava
ila
b
l
e
.
T
he
d
a
t
a
is div
id
ed
in
t
o thr
ee
par
t
s
.
80%
o
f the
da
t
a
i
s
use
d
f
or
t
r
a
i
n
i
ng
t
h
e
neur
a
l
n
etw
o
r
k
[
1
2
]
.
T
he
15%
i
s
us
ed
f
or
tes
t
i
n
g
and
t
h
e
r
e
st
i
s
use
d
f
or
v
a
l
id
ati
o
n.
F
i
g
u
r
e
4
a
nd
Figu
re
5
s
ho
w th
e
b
l
oc
k
diagram
of
t
raining
pro
c
edure
for
ne
u
r
al
n
etwo
rk
a
n
d
t
he
flow
c
hart
of
t
h
e
propose
d
s
y
s
t
e
m.
Figure
4.
B
loc
k
dia
gra
m
of
tra
i
ning proc
edure
for
neural
n
e
t
wor
k
Fig
u
re
5
. Flow c
h
art of
t
he
p
ro
po
se
d s
y
stem
The
da
t
a
i
s
d
i
v
i
de
d
i
n
t
o
b
a
t
c
h
e
s
dur
in
g
tr
ai
n
i
ng
to
r
ed
uc
e
th
e
tr
a
i
n
i
ng
t
i
m
e
a
nd
t
o
d
e
c
r
e
ase
the
c
o
m
puta
tio
na
l
r
e
sour
ce
s
r
e
q
u
i
r
e
d
t
o
t
r
a
in
d
ee
p
le
ar
n
i
n
g
[
13]
t
he
n
e
u
r
a
l
ne
t
w
o
r
k
.
We
u
se
a
F
ee
d
F
o
r
w
a
r
d
N
e
ur
a
l
N
e
t
w
o
r
k
i
n
th
is
w
or
k.
T
he
n
etw
o
r
k
i
s
tr
ai
ned
usi
ng
G
r
a
d
i
e
n
t
D
e
s
cen
t
base
d
Bac
k
P
r
opa
ga
ti
o
n
.
Bac
k
P
r
opa
gat
i
on
i
s
u
se
d
to
u
p
d
a
t
e
t
h
e
weig
hts
to
i
nc
rea
s
e
the
ac
c
u
ra
cy
o
f
t
h
e
ne
u
r
al
n
e
tw
or
k.
T
he
l
oss
us
ed
t
o
tr
ai
n
t
h
e
ne
t
w
or
k
is
M
e
a
n
Ab
s
ol
ut
e Er
ro
r
.
5.
S
I
MULA
TION
To
t
e
s
t
our
s
ys
tem
,
a
n
in
ver
t
e
r
c
ir
c
u
it
w
i
t
h
t
h
e
pr
opose
d
M
a
c
hi
ne
L
ea
r
n
in
g
m
o
del
w
a
s
s
i
mula
t
e
d
i
n
MA
TLA
B
for
tw
o
ty
pe
s
o
f
l
o
a
d.
T
he
n
e
u
r
a
l
netw
or
k
w
a
s
tr
a
i
ne
d
us
i
n
g
M
A
TLA
Bs
N
e
u
r
a
l
N
e
tw
or
k
T
o
ol
k
it
a
nd
w
a
s
use
d
t
o
pr
e
d
i
c
t
t
h
e
gate
p
ulse
s
tha
t
a
r
e
r
e
qui
r
e
d
t
o
r
e
d
u
c
e
t
h
e
t
r
an
si
e
n
t
s
i
n
th
e
s
w
i
t
ch
es.
T
h
e
sim
u
la
ti
on
r
e
s
u
l
t
s
ar
e
sh
ow
n
i
n
t
he
f
i
g
ur
es.
The
ver
tica
l
a
xi
s
r
e
pr
ese
n
ts
t
he
v
ol
ta
ge
w
h
e
r
e
as
t
he
hor
iz
on
ta
l
a
x
is
i
s
t
h
e
t
i
m
e
.
The
fi
r
s
t
grap
h
sh
ows
t
h
e
Gate
P
u
l
se
s
an
d
t
h
e
S
ec
on
d
G
r
ap
h
sh
ow
s
t
h
e
r
e
d
u
c
t
i
o
n
i
n
tr
a
n
sie
n
t
s
a
s
t
h
e
neur
a
l
ne
t
w
o
r
k
t
r
a
i
n
s.
T
he
s
e
c
on
d
g
r
ap
h
sh
ow
s
t
he
G
a
t
e P
u
l
s
e
s
an
d
t
he
f
ir
st
G
r
a
ph
s
how
s t
h
e
r
e
duc
ti
o
n
i
n
tr
a
n
sie
n
t
s
a
s
t
h
e
neur
al
n
e
t
w
o
r
k
t
r
a
i
n
s.
F
r
o
m
the
F
i
g
u
res
i
t
can
b
e
se
en
t
ha
t
the
t
r
ans
i
e
n
t
s
a
r
e
m
o
r
e
i
n t
h
e f
i
r
s
t ha
lf
of
t
h
e
gr
aph.
T
hi
s
i
s
b
e
cause
th
i
s
is t
h
e tim
e when th
e
n
e
u
r
a
l ne
tw
or
k
is tra
i
n
in
g.
Durin
g
th
is
tim
e,
t
he
g
ate
p
u
l
ses
ar
e
als
o
v
er
y
e
rra
ti
c
a
nd
fluc
t
u
at
i
n
g.
F
igur
e
6
t
o
F
i
g
ur
e
8
sh
ow
t
he
s
im
ul
a
tio
n
dia
g
r
a
m
,
s
im
ulatio
n
r
e
sul
t
s
fo
r
RL
l
oad
an
d
R
l
o
ad.
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
Re
d
u
ct
ion o
f
tr
a
n
s
i
e
n
ts
in swi
t
c
h
e
s
usi
n
g
e
m
bed
d
e
d
m
a
ch
in
e l
e
ar
nin
g
(
P
.
Sur
esh)
23
9
Figure 6.
Sim
ulat
ion
diagram
Figure
7.
S
im
ulation results
f
or RL l
o
a
d
F
igure 8.
Sim
ula
tion
r
e
s
ults for
R
L
oad
Aft
e
r
t
h
e
Ne
ur
al
N
etw
o
rk
h
as
c
omp
l
ete
d
t
ra
ini
n
g,
t
he
g
a
t
e
p
u
lse
s
ar
e
m
o
re
s
ta
b
l
e
a
n
d
t
h
e
o
u
t
p
ut
h
a
s
a
l
e
s
se
r
num
ber
of
t
r
a
n
s
ie
nts
an
d
f
l
uc
t
u
at
ion
s
.
6.
H
A
R
D
W
A
RE
IMPLEM
E
N
T
ATIO
N
T
h
er
e
ar
e
th
r
e
e
m
a
jo
r
p
a
r
t
s
to
t
h
e
h
a
r
d
w
a
r
e
im
p
l
e
m
en
t
a
tio
n
wh
i
c
h
is
s
hown
in
F
ig
ure
9
:
a.
Inv
e
r
t
er
C
ir
cu
it:
T
h
e
I
n
v
er
ter
is
a
n
e
l
e
c
tr
i
cal
d
ev
ice
wh
ich
co
nve
r
t
s
d
i
r
ect
c
u
rrent
(
D
C
)
t
o
a
lte
r
n
ate
cu
rren
t
(AC)
.
T
h
e
i
n
v
e
rte
r
i
s
us
ed
f
or
e
m
e
rgen
c
y
b
ackup
powe
r
i
n
a
ho
m
e
.
Th
e
i
n
ve
rt
er
i
s
u
s
e
d
i
n
s
o
me
aircraft
s
yste
m
s
t
o c
onvert
a
p
o
r
t
i
o
n
of
t
he
a
irc
r
a
f
t
DC
[
1
4
]
p
ower t
o
AC
.
The AC
powe
r
i
s
used
m
ain
l
y
for
e
l
ec
t
r
ic
al
d
ev
i
ces
lik
e
l
i
gh
ts
,
r
a
d
a
r
,
r
ad
io
,
m
o
tor
,
a
nd
o
t
he
r
de
vic
e
s
.
A
H-b
r
i
d
ge
b
ri
d
g
e
i
n
ve
rte
r
w
a
s
u
s
e
d
i
n
t
h
e
p
r
oje
c
t
.
H
b
ri
d
g
e
s
a
re
a
va
ila
ble
a
s
i
n
t
e
g
r
a
te
d
ci
rc
uits
,
or
c
a
n
b
e
built
f
r
o
m
disc
ret
e
co
m
p
o
n
en
ts
.
T
h
e
term
H
b
ridg
e
is
d
eriv
e
d
f
rom
t
h
e
ty
pica
l
gra
p
h
ica
l
r
e
p
re
se
nta
t
i
o
n
o
f
s
uc
h
a
ci
rc
uit.
A
n
H
b
r
i
d
g
e
i
s
b
u
i
l
t
w
i
t
h
f
o
u
r
s
w
i
t
c
h
e
s
[
1
5
]
(
s
o
l
i
d
-
s
t
a
t
e
o
r
m
e
c
h
an
ical).
W
h
e
n
th
e
switch
es
[
16
]
S1
a
nd
S4
(
acco
rd
ing
to
t
h
e
F
igu
r
e
1)
a
re
c
los
e
d
(and
S
2
and
S3
a
re
o
p
en)
a
positi
ve
v
olta
ge
w
ill
be
a
ppl
i
e
d
across
th
e
motor
.
B
y
op
en
ing
S1
a
nd
S
4
switches
a
nd
closi
n
g
S2
a
n
d
S3
s
witc
he
s,
t
his
v
o
l
t
a
g
e
is
rev
e
r
s
ed
,
allowing
r
e
v
ers
e
op
e
ra
ti
on
of t
h
e mo
tor.
M
any
a
p
p
l
ica
tion
s are
g
i
ven
in
[17-24]
.
b.
R
a
spber
r
y
P
i:
A
R
aspber
r
y
P
i
is
a
c
re
di
t
car
d
-size
d
c
om
put
e
r
or
i
g
in
al
ly
d
e
s
ign
e
d
for
edu
c
ation
,
inspi
r
e
d
b
y
t
h
e
1981
BBC
Micro.
C
reator
E
ben
Upton’s
goal
was
t
o
create
a
l
ow-cost
d
e
v
i
ce
t
hat
w
o
ul
d
i
m
pr
o
v
e
p
r
o
g
r
am
m
i
n
g
s
kill
s
a
n
d
ha
r
d
wa
re
u
n
d
e
r
s
t
an
di
ng
at
t
h
e
pre
-
un
iv
e
r
sity
l
ev
el.
Bu
t
th
a
nks
t
o
i
ts
s
mal
l
s
i
z
e
and
a
c
c
e
ss
ib
le
p
r
i
c
e
,
i
t
w
a
s
q
u
i
ck
ly
a
dop
ted
by
t
inker
e
r
s
,
mak
e
rs,
and
electron
ics
en
thus
ia
s
t
s
for
pro
j
ects
t
h
a
t
r
e
q
u
i
re
m
ore
than
a
b
a
s
ic
m
icro
co
nt
r
o
ll
e
r
(
s
u
c
h
a
s
A
r
d
u
i
n
o
de
vic
e
s
).
Th
e
Ra
s
p
be
r
r
y
P
i
i
s
o
pe
n
h
a
r
d
wa
re
,
wit
h
t
he
e
x
c
epti
o
n
of
t
h
e
p
rim
a
r
y
ch
ip
o
n
the
R
a
spb
e
rry
P
i,
t
h
e
B
road
c
o
m
m
.
S
o
C
(
S
y
ste
m
o
n
a
Ch
i
p
),
w
hic
h
r
u
n
s
m
a
n
y
o
f
t
he
m
ai
n
c
o
m
p
o
ne
nt
s
o
f
t
he
b
oa
r
d
C
P
U
,
g
r
a
p
hi
cs
,
m
e
mory
,
the
USB
con
t
ro
ller,
etc
.
M
any
of
t
h
e
p
ro
je
cts
m
a
d
e
w
ith
a
R
a
s
pb
erry
P
i
are
op
e
n
a
nd
w
ell-
do
cumen
t
e
d
a
s w
e
ll and
are t
h
i
ngs you
ca
n
bu
ild
a
n
d
m
od
ify
your
se
lf.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
S
N: 2
0
8
8
-
86
94
I
nt
J
P
o
w
E
l
e
c
&
D
r
i
S
yst
V
o
l.
11,
N
o.
1
,
Mar
202
0
:
235
–
24
1
24
0
c.
Po
we
r
Ci
rc
ui
t:
T
he
p
o
w
e
r
t
o
the
c
i
r
c
ui
t
i
s
g
i
v
e
n
b
y
a
t
r
a
n
sf
o
r
m
er
[
25]
.
The
ou
tpu
t
f
rom
th
e
tr
ansfo
r
m
e
r
is
r
e
c
t
i
f
ie
d
us
i
n
g
a
f
u
ll
w
a
v
e
re
ct
i
f
ier.
I
t
is
t
he
n
filt
er
ed
u
sing
c
a
p
acito
rs
t
o
g
i
ve
a
D
C
inpu
t
to
th
e Inv
e
r
t
er.
Figure
9
. Hardw
a
r
e
Imp
l
eme
n
tation
of
T
ra
n
s
ie
nt
s
Re
d
u
c
t
i
o
n
u
s
i
n
g
M
L
7.
H
A
R
D
W
A
RE
R
ES
ULTS
The
r
e
sul
t
s
fr
o
m
t
he
h
a
r
dw
ar
e
ar
e
sh
ow
n
F
i
gur
e
1
0
a
nd
F
i
g
u
r
e
1
1
.
The
g
r
a
ph
s
h
ow
n
i
n
F
ig
ur
e
1
0
i
s
the
ou
t
p
u
t
o
f
t
h
e
i
n
ve
r
t
er
w
i
t
ho
u
t
our
n
e
u
r
a
l
ne
tw
or
k
mo
d
e
l.
I
n
t
h
i
s
gr
a
p
h,
t
he
t
r
a
ns
ie
nt
s
ar
e
m
o
r
e
a
n
d
t
h
e
y
ha
ve
a
h
ig
h
pe
ak
o
f
a
b
ou
t
1
7
V
.
This
c
a
n
l
e
a
d
to
f
a
l
se
s
w
itch
i
ng
i
n
t
h
e
s
w
itc
hes
a
nd
it
c
a
n
a
l
s
o
c
a
u
se
f
a
s
t
e
r
f
a
i
l
ur
e
a
nd
de
gr
adat
io
n
of
s
w
i
t
c
he
s.
The
ver
t
ica
l
a
xi
s
in
t
he
g
r
a
p
h
i
s
the
v
o
l
t
a
g
e
an
d
the
hor
iz
on
t
a
l
a
x
is
i
s
the
t
i
me
.
In
t
he
F
igur
e
1
1
,
howe
v
er,
the
t
r
ans
i
en
t pe
a
k
h
a
s
b
ee
n re
duce
d
a
nd i
t
c
an
b
e
se
en
t
hat
the
m
a
xim
u
m
t
r
ans
i
ent
s
i
s ju
st
1
5 V.
F
i
gur
e
1
0
.
S
i
mul
a
t
i
o
n
o
utp
u
t
w
it
ho
u
t
n
eur
a
l
netw
or
k
F
i
g
u
r
e
11.
Si
m
u
l
a
ti
o
n
o
u
t
p
u
t
o
f
i
n
v
e
r
t
e
r
w
i
t
h
neur
a
l
n
etw
o
r
k
8.
CONCLUSION
I
n
t
h
i
s
pa
per
a
cir
c
ui
t
f
o
r
r
e
duci
n
g
tr
ans
i
e
n
t
s
i
n
sw
i
t
che
s
w
a
s
p
r
o
p
o
sed
us
i
n
g
ne
ur
al
n
etw
o
r
k
s.
T
his
w
a
s
d
one
b
y
u
s
i
n
g
a
ha
r
d
w
a
r
e
i
mp
lem
e
n
t
ed
i
nver
t
er
c
i
r
cu
i
t
a
lo
n
g
w
it
h
r
a
spb
e
r
r
y
p
i
w
hi
ch
i
s use
d
f
or
r
un
nin
g
the
ne
ur
a
l
n
e
t
w
o
r
k
on
i
t
.
Thi
s
n
eur
a
l
ne
tw
o
r
k
fur
t
her
w
a
s
suc
c
e
ss
f
u
l
i
n
r
educ
ing
the
tr
ans
i
en
ts
a
n
d
o
b
t
a
i
ni
ng
an
a
cc
u
r
ac
y
of
8
0%
. T
h
i
s
ci
r
c
u
it
was
si
m
u
l
ate
d
i
n
M
A
T
L
A
B
to
b
ac
k
t
he
resu
lts.
REFERE
NC
E
S
[1]
P
.
S
uresh
an
d
A.
S
ures
hkum
ar,
“M
o
d
elli
ng
and
s
i
m
u
lation
of
S
p
a
c
e
V
e
ctor
M
od
ul
ated
M
atri
x
Conv
erter
F
e
d
In
du
c
t
io
n
mo
to
r,”
In
te
rn
at
io
na
l J
o
urna
l o
f
Co
nt
ro
l
Th
e
o
ry
an
d
Ap
p
l
ic
a
t
ion
, Vol
. 9
,
no
.
6
, 20
1
6
.
[2]
M
easu
r
emen
t
of
T
ran
s
i
e
nt
V
ol
tages
Indu
ced
b
y
Dis
c
on
n
ect
S
w
i
tch
O
p
erat
io
n.
[
Online]
.
A
v
a
il
ab
l
e
:
ht
tps://
ieeexp
l
o
r
e.ieee.
o
r
g/
do
c
u
m
e
nt/4
112
91
3/
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
Red
u
ct
i
on
o
f
tr
ans
i
e
n
t
s i
n
s
w
i
t
che
s
usin
g em
bed
d
e
d
m
a
ch
in
e lear
nin
g
(P.
Sur
esh)
24
1
[3]
T
h
e
m
eas
urem
ent
o
f
t
rans
isto
r
tra
n
sien
t
sw
i
t
c
h
in
g
p
a
ramet
e
rs.
[
On
lin
e]
.
A
v
ail
a
bl
e:
h
t
t
p
s
:
//i
e
eex
pl
ore.i
eee.org
/
d
o
cum
e
n
t
/
524
44
04
/
[4]
G.
F
.
S
a
v
a
r
i
,
V
.
K
r
is
hn
a
s
a
m
y,
J
.
S
a
th
ik
,
et al.
,
"Int
ernet
o
f
T
h
i
ngs
b
ased
r
eal-t
im
e
electri
c
v
e
h
i
cl
e
l
o
ad
f
or
ecastin
g
an
d
ch
argi
ng
s
t
a
tio
n
rec
o
m
m
e
ndatio
n,
"
ISA Transaction
s
,
2
019
.
[5]
Ge
o
r
ge
F
e
r
n
a
n
d
e
z
Sa
va
ri,
Vij
a
ya
k
u
ma
r
Krishn
a
s
a
m
y,
a
nd
J
ose
p
M
.
Gu
errero
,
"
O
pt
im
a
l
S
ched
uli
n
g
an
d
E
c
on
omic
A
nal
y
si
s
of
H
y
b
rid
El
ectric
V
e
hi
cles
i
n
a
M
i
cro
g
rid,
"
In
te
rna
t
io
na
l J
o
u
r
n
a
l of
E
m
e
r
ging
E
l
e
c
t
ri
c
Po
w
e
r S
y
s
t
e
m
s
, v
ol
.
1
9
,
n
o
.
6
,
2018.
[6]
P
r
ad
eep
V
is
hn
uram
,
S
r
i
dhar
Ramas
a
m
y
,
et
a
l
.
,
“A
s
imple
di
g
i
t
a
l
control
f
o
r
mitiga
ting
vo
l
t
ag
est
r
ess
on
s
i
n
gle
s
w
itch
reso
nant
i
nvert
er
f
ori
n
d
u
cti
o
n
coo
k
i
ng
app
l
i
c
at
io
n
s
,
”
In
t
e
rnat
io
na
l
Jo
urn
a
l of
E
l
ectr
o
n
i
cs
Lett
e
r
s
,
20
19.
[7]
Ge
o
r
ge
F
e
r
na
nd
e
z
Sa
v
a
ri
a
nd
Vija
ya
k
u
m
a
r
Krishn
a
s
a
m
y,
"
Un
ma
nn
e
d
an
d
aut
ono
m
o
u
s
g
ro
un
d
veh
i
cle,
"
In
ter
n
a
t
io
nal Jour
na
l o
f
Elect
ri
cal
an
d
Co
mputer En
g
i
neeri
n
g
,
Vo
l
.
9
, No.
5,
p
p
.
4
4
6
6
-
44
7
2
, Octo
b
er, 20
1
9
.
[8]
M
eas
uri
ng
O
u
tput
R
ipple
and
S
w
itchi
ng
T
rans
ien
t
s
i
n
S
witch
i
ng
R
egulato
r
sh
tt
p.
[
On
line]
.
A
va
il
abl
e
:
/
/
w
ww.anal
og.
c
o
m
/
m
e
d
i
a
/
en/t
ec
hn
ical
-d
ocu
m
entat
i
o
n
/
a
pp
licat
io
n
-
n
otes
/A
N-11
44
.pd
f
[9]
Ri
yat
r
i
Roy,
A
b
i
sanka
Bhatt
a
c
h
arya,
et a
l
.
,
“M
it
iga
t
io
n
of
P
ower
Q
u
a
li
ty
D
i
s
t
u
rb
a
n
ces
i
n
P
o
we
r
Sy
st
em
u
si
ng
DV
R
,
”
Inter
n
a
t
ion
a
l
Jour
nal of Recent
T
echn
o
logy
and
En
g
i
neer
ing
,
V
o
l
.
8
,
No
. 1
4,
J
un
e
20
19.
[10]
V
.
N
air
and
G.
E
.
H
i
n
t
o
n
,
“Recti
f
ied
li
near
u
n
i
t
s
i
m
p
ro
ve
r
estr
i
c
ted
bo
ltzmann
m
achi
n
es,”
Proc
. 27
th
In
ter
n
a
t
io
nal Conf
eren
c
e
on
M
a
chin
e L
e
ar
n
i
ng
,
2
0
10
.
[11]
M
ach
in
e
Learn
i
n
g
M
astery
.
[O
nlin
e].
A
v
ailabl
e:
h
t
t
ps://
m
ach
in
el
earni
ng
mastery
.
com/
sav
e
-l
o
a
d
-
keras-d
e
ep-
l
earni
ng
-mo
d
els/
[12]
B.
K
.
Bose,
“N
eural
N
e
tw
ork
A
p
p
lications
i
n
Power
E
l
e
c
t
r
oni
cs
a
n
d
M
ot
or
D
rives:
I
n
t
ro
du
ctio
n
and
P
e
rs
pect
ive,
”
IEE
E
Tr
ans
actio
ns on
Ind
u
stria
l
Electr
o
n
i
cs
,
v
o
l
.
5
4
,
n
o
.
1
,
p
p
.
1
4
-3
3,
F
e
b
2
007.
[13]
M
i
chael
A
. Nielsen
,
Neur
a
l
Netwo
r
ks
an
d
D
eep L
e
a
r
ni
ng
. Determ
i
n
a
tio
n
Pres
s
,
20
1
5
.
[14]
P
.
S
uresh
and
Ki
rub
a
karan
D
h
an
dapan
i
,
“Enh
anced
Z
eta
Co
nv
erter
f
o
r
DC
B
us
V
oltage
R
egu
l
at
io
n,”
In
ter
n
a
t
io
nal Jour
na
l o
f
P
o
wer
El
ectro
n
i
cs
a
n
d
Dr
ive S
y
s
t
ems
(
I
JPED
S
)
,
vol. 8
, N
o.
4
, p. 1
50
3
, Decem
ber 2
0
1
7
.
[15]
M
i
n
i
R
,
M
a
n
j
i
r
i
J
o
s
h
i
,
B
.
H
a
r
i
r
a
m
S
a
t
h
e
e
s
h
,
a
n
d
D
i
n
e
s
h
M
.
N
,
“
A
ctive
L
C
C
lamp
d
v
/
dt
F
i
l
t
e
r
f
o
r
Vol
t
age
Ref
l
ecti
on
du
e
to
L
o
n
g
Cabl
e
in
I
nd
uctio
n
Mo
to
r
Dri
v
es
,
”
In
d
o
n
e
sia
n
Jour
n
a
l o
f
El
ectr
i
cal
an
d Co
mp
u
t
er
Sc
ie
nc
e
, Vol. 6
,
No
.
4
,
A
ug
u
s
t
2
0
1
6
.
[16]
S
u
resh
k
u
m
ar
A
,
S
u
re
s
h
P
,
Vis
hnuram
P
,
et a
l
.,
“P
o
w
er
F
act
or
M
ai
nt
ain
f
o
r
LED
D
r
iv
erUs
i
n
g
Is
o
l
at
ed
C
onv
erte
r
w
i
t
h
S
of
t
S
w
it
chi
ng,
”
Jou
r
o
f
A
d
v R
e
sea
r
ch
in D
y
n
a
mi
cal
&
Contr
o
l
System
s
, Vo
l
.
10
,
Special Issu
e, 20
1
8
.
[17]
G
.
F.
S
av
ari,
V
.
Kris
hn
asa
m
y,
et al
.,
"Op
t
i
m
a
l
C
h
a
rgin
g
S
c
hedu
li
ng
o
f
Elect
ric
V
e
hi
cles
i
n
Micro
G
ri
ds
Using
Pri
o
r
i
ty
A
l
g
or
ith
ms
and
Par
ticl
e
S
warm
O
pt
imiza
t
ion,"
M
o
b
ile Ne
t
w
or
k
s
a
nd Ap
pli
c
atio
ns
,
vol.
1
7
,
2
019
.
[18]
P
.
S
u
r
esh
,
G
.F
S
avari,
S
u
r
es
h
Kum
a
r
A
.
,
“Des
ign
a
s
i
ngle
st
age
A
C
t
o
D
C
c
o
n
v
e
r
t
e
r
f
o
r
L
E
D
d
r
i
v
e
r
w
i
t
h
p
o
w
e
r
f
act
or
i
m
p
ro
vement
,
”
In
tern
a
t
io
n
a
l Jo
urn
a
l o
f
Recent
T
echno
lo
gy a
n
d
En
gi
neering
,
v
o
l
.
8
(
2
S
p
e
c
i
a
l
I
s
s
u
e
1
1
)
,
p
p
.
33
12-3
3
1
8
,
20
19.
[19]
D.
S
attianadan,
“
Techno
E
cono
mic
E
v
al
uation
of
a
H
ybr
i
d
energy
system,
”
Inter
n
a
t
i
onal Jo
ur
na
l of
Recen
t
Te
c
h
n
o
lo
gy
an
d
En
gine
e
r
in
g
,
vol.
8(2
S
p
ecial
I
ssue
11),
pp.
2
575-25
79
,
2
0
1
9
.
[20]
K
.
V
ij
ayak
um
ar,
“H
yb
rid
En
ergy
S
ou
rce
F
e
d
Th
ree
L
e
v
e
l
NP
C
w
i
t
h
Quasi
Z
So
urce
Netw
ork
,
”
Internationa
l
Jo
ur
na
l of
P
u
r
e
an
d A
ppli
e
d
M
a
thema
t
i
c
s
, Jan
2
0
1
8
.
[21]
K
.
S
el
vak
u
m
a
r,
K
.
Vi
jayak
u
m
a
r,
D
.
Kart
hi
keya
n
,
D
.
S
e
lvab
harat
h
i
,
an
d
V.
K
u
b
end
r
an,
“
H
y
s
teres
i
s
co
ntrol
3-
l
e
vel
S
I-NP
C
i
nvert
e
r
w
it
h
wind
energ
y
s
ys
tem
,
”
Int
e
rn
at
ion
a
l
Jou
r
n
a
l o
f
P
o
wer
Elect
ro
nics
and D
r
i
v
e
S
y
st
em
(I
J
P
E
D
S
)
,
vo
l
.
8
, n
o.
4,
Dec 2
0
1
7
.
[22]
K
a
bi
ru
M
a
i
d
a
lailu
a
nd
G.
F.
S
av
ari,
“
M
o
d
e
li
ng
a
nd
S
im
u
l
ati
o
n
o
f
Dy
nami
c
Vo
l
t
ag
e
Rest
orer
(
DV
R)
U
si
ng
ZSI
f
o
r
M
i
t
i
gation
of
V
ol
tage
S
ags/S
w
ells,”
In
te
rna
t
i
o
na
l jo
urna
l of App
lie
d
Eng
i
n
e
e
r
in
g
Re
se
arc
h
,
V
o
l
.
1
0
,
N
o
.
4
4
,
20
15
.
[23]
G
o
u
r
ab
S
ah
a,
“
O
p
tim
a
l
p
l
acemen
t
of
D
istri
b
u
t
e
d
G
en
e
r
at
io
n
i
n
a
D
i
st
rib
u
tio
n
s
y
s
t
em
u
si
ng
H
y
b
ri
d
Big
Brunch
&
Bi
g Crunch
A
l
g
orithm,”
In
tern
a
t
io
nal Jo
urn
a
l of Con
t
ro
l Th
e
o
r
y
an
d
A
p
pl
i
c
at
i
o
ns
, vo
l
.
9
(
16
),
2
01
6
.
[24]
Kunal
Ma
li
k
,
S
at
yajit
Dora,
an
d
K.
V
i
j
ayakumar,
“A
N
ew
S
ymmet
r
i
c
&
Asymmetr
i
c
M
u
lti
level
Inverte
r
T
o
p
o
l
ogy
w
it
h
Reduced
M
axi
m
um
B
l
o
ckin
g
Vol
t
ag
e
S
w
itch
e
s,”
Jou
r
na
l of
Advan
ced Resea
r
c
h
in
D
y
nam
ical
an
d Co
nt
ro
l S
y
st
e
m
s
,
Vo
l. 7
,
No.
1
1
,
2
0
1
8
.
[25]
A
n
u
p
ri
ya
K
R
an
d
S
a
s
i
lat
h
a
T.,
“Tras
n
f
o
rm
er
E
xc
h
a
ng
in
g
w
i
t
h
V
a
c
u
u
m
E
l
e
c
t
r
i
ca
l
S
w
it
ch
,
”
In
do
nes
i
an
Jo
ur
na
l
o
f
E
l
ect
ri
cal En
gineeri
n
g
an
d Comp
uter
Sci
e
nce
,
V
o
l
.
8
, N
o
. 3
,
p
p 679
-68
0
, D
ecember 2
017
.
Evaluation Warning : The document was created with Spire.PDF for Python.