Internati
o
nal
Journal of Ele
c
trical
and Computer
Engineering
(IJE
CE)
Vol
.
4
,
No
. 5, Oct
o
ber
2
0
1
4
,
pp
. 69
1~
69
6
I
S
SN
: 208
8-8
7
0
8
6
91
Jo
urn
a
l
h
o
me
pa
ge
: h
ttp
://iaesjo
u
r
na
l.com/
o
n
lin
e/ind
e
x.ph
p
/
IJECE
Modeling and Simulation of Fuel
cell (Dicks-Larminie Model)
based 3-Phase Voltage Source Inverter
G
a
u
r
av
Sa
chdev
a
Dept. of
E
l
e
c
tri
c
al
Engine
ering
,
Des
h
Bhagat
Un
ivers
i
t
y
Article Info
A
B
STRAC
T
Article histo
r
y:
Received Dec 26, 2013
Rev
i
sed
Jun
12,
201
4
Accepte
d Aug 2, 2014
In the presen
t
work, performance of
thr
e
e ph
ase voltage sou
r
ce
inver
t
er,
while fe
eding d
i
fferent
power fa
ctor lo
ads
,
has
b
een inv
e
s
tiga
t
ed
.
F
u
el c
e
lls
models
namely
Dicks-Larminie model
are
used in input sid
e
as a DC source
while d
y
n
a
m
i
c l
o
ad have been u
s
ed at the output
s
i
de. D
y
nam
i
c l
o
ad us
ed is
induction motor
(IM). Performance of
IM has been investig
ated under
various loading
conditions
. ANN based cont
rol strateg
y
has been proposed to
find the conduction angle of
a
Three Ph
ase VSI and ver
i
fied f
o
r IM load.
Simulations hav
e
been p
e
rformed
using PSIM 7.0.5 and
MATLA
B 7.0.4
.
Keyword:
Dick
s-Larm
in
ie Mod
e
l
Fuel cell
I
ndu
ctio
n m
a
c
h
in
e
VSI
Copyright ©
201
4 Institut
e
o
f
Ad
vanced
Engin
eer
ing and S
c
i
e
nce.
All rights re
se
rve
d
.
Co
rresp
ond
i
ng
Autho
r
:
G
a
u
r
av
Sa
chdev
a
Assistant Profe
ssor, De
pt. of Electrical
Engg,
D
e
sh Bh
ag
at
Un
iv
er
sity, Mand
i Gob
i
nd
g
a
rh
, Punj
ab,
I
n
d
i
a.
+9
1-9
8
7
-
22
2-44
44
Em
a
il: erg
a
u
r
av
sach
d
e
v
a
@gmail.co
m
1.
INTRODUCTION
Fuel cells are
m
o
stly analyzed
electroche
m
ically, this analysis
m
a
thematical
m
odels which a
r
e
im
port
a
nt
not
onl
y
f
o
r
i
m
provi
n
g
t
h
e
desi
g
n
o
f
m
e
m
b
rane
s fl
o
w
fi
el
d
s
, a
nd c
a
t
a
l
y
st
s bu
t
al
so f
o
r
det
e
r
m
i
n
i
n
g
th
e op
tim
al o
p
e
ratin
g cond
itio
n lik
e
fu
el
fl
o
w
rates,
h
u
m
id
ity an
d tem
p
eratu
r
e.
Dick
s-Larm
in
ie
m
o
d
e
l is
deri
ved
f
r
om
t
h
e
fre
que
ncy
r
e
sp
onse
o
f
t
h
e
fuel
cel
l
.
2.
LITERATU
R
E
REVIE
W
{1} F
o
c
u
ses
o
n
a
ve
hi
cul
a
r
s
y
st
em
powe
r
ed
by
a
f
u
el cell
and equippe
d
with
two
secondary e
n
ergy
st
ora
g
e de
vi
ces:
bat
t
e
ry
and ul
t
r
a-ca
paci
t
o
r
(UC
)
. A wa
vel
e
t
and fuzzy
l
o
gi
c (FL)
base
d
energy
m
a
nagem
e
nt
st
rat
e
gy
i
s
pr
o
pos
ed
f
o
r
t
h
e
devel
ope
d
hy
b
r
i
d
ve
hi
cul
a
r
s
y
st
em
. {2} S
h
ows
t
h
e
co
nt
r
o
l
of
t
h
e
f
u
el
c
e
l
l
for
stan
d-
alon
e and
g
r
i
d
co
nn
ect
io
n
.
A pow
er
co
nd
itio
n
i
ng
un
it is d
e
sign
ed fo
r t
h
e so
lid
o
x
i
d
e
fu
el cell, wh
ich
can be use
d
f
o
r ot
he
r fuel
cel
l
s
wi
t
h
con
v
ert
e
r and t
h
e i
n
v
e
rt
er of
di
ffe
re
nt
rat
i
ngs
, b
u
t
t
h
e cont
rol
st
r
a
t
e
gy
u
s
ing
fu
zzy lo
g
i
c (FL)
will re
m
a
in
th
e sa
me fo
r d
e
sign
ing th
e co
n
t
ro
llers. {3
} p
r
esen
ts a fu
el cell h
y
b
r
id
b
u
s
whic
h is equipped
with a
fuel
cell syste
m
and two ene
r
gy
stora
g
e de
vices, i.e., a battery
and a
n
ultra ca
pacitor
(UC
)
. A
n
ene
r
gy
m
a
nagem
e
nt
st
rat
e
gy
based o
n
f
u
zzy
l
ogi
c, whi
c
h i
s
em
pl
oy
ed t
o
c
o
nt
r
o
l
t
h
e p
o
we
r fl
ow
of
t
h
e ve
hi
c
u
l
a
r
p
o
we
r t
r
ai
n, i
s
descri
bed
.
T
h
i
s
st
rat
e
gy
i
s
ca
pabl
e
o
f
det
e
r
m
i
n
i
ng t
h
e
des
i
red
o
u
t
p
ut
p
o
w
er
o
f
th
e fu
el cell sy
ste
m
, b
a
ttery an
d u
l
t
r
a
capaci
t
o
r acc
o
r
di
ng
t
o
t
h
e
pr
op
ul
si
on
p
o
w
er
an
d
recu
perat
e
d
br
aki
n
g
po
we
r. A c
o
nt
rol
st
rat
e
gy
i
s
prese
n
t
e
d
by
{4} w
h
i
c
h i
s
s
u
i
t
a
bl
e for m
i
ni
at
ure hy
dr
o
g
en
/
a
i
r
pr
ot
o
n
-e
xc
han
g
e
m
e
m
b
rane (P
EM
) f
u
el
cel
l
s
. The c
ont
rol
app
r
oach i
s
ba
sed o
n
pr
oces
s
m
odel
i
n
g us
i
ng f
u
zzy
l
o
gi
c. The
param
e
t
e
rs of t
h
e f
u
zzy
r
u
l
e
base a
r
e det
e
r
m
i
n
ed by
pl
o
ttin
g
ch
aracteristic d
i
ag
ram
s
of the fuel cell stack at
con
s
t
a
nt
t
e
m
p
erat
ure
s
. FC
sy
s
t
em
and UC
b
a
nk s
u
ppl
y
p
o
w
er
dem
a
nd u
s
i
ng a c
u
r
r
ent
-
fed
ful
l
b
r
i
d
ge
dc-
d
c
con
v
e
r
t
e
r a
n
d
a bi
di
r
ect
i
onal
dc-
d
c c
o
nve
rt
er i
s
p
r
ese
n
t
e
d
by
{5}
.
It
f
o
c
u
ses
o
n
a
no
ve
l
fuzzy
l
o
gi
c c
ont
rol
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE Vo
l. 4
,
N
o
. 5
,
O
c
tob
e
r
20
14
:
691
–
6
96
69
2
alg
o
rith
m
in
te
g
r
ated
in
t
o
th
e po
wer con
d
itio
n
i
n
g
un
it (P
C
U
f
o
r
t
h
e
hy
bri
d
sy
st
em
. The c
ont
rol
st
ra
t
e
gy
i
s
capabl
e
o
f
det
e
rm
i
n
i
ng t
h
e
de
si
red
FC
po
we
r a
n
d
kee
p
s t
h
e
dc
v
o
l
t
a
ge a
r
o
u
n
d
i
t
s
n
o
m
i
nal
val
u
e
by
s
u
pp
l
y
i
n
g
pr
o
pul
si
o
n
p
o
w
er a
nd rec
u
perat
i
n
g b
r
aki
ng e
n
er
gy
. I
n
{6}, a fuzzy
l
ogi
c al
go
ri
t
h
m
has been use
d
t
o
det
e
rm
i
n
e t
h
e
fuel
cel
l
o
u
t
p
u
t
po
we
r
depe
n
d
i
n
g
on
t
h
e
ex
t
e
rnal
po
we
r r
e
qui
rem
e
nt
and t
h
e
bat
t
e
ry
s
t
at
e o
f
char
ge
(S
oC
)
.
If
-t
he
n
ope
rat
i
o
n
r
u
l
e
s a
r
e i
m
pl
em
ent
e
d b
y
fuzzy
l
ogi
c
fo
r t
h
e
e
n
er
gy
m
a
nagem
e
nt
of
t
h
e
hy
b
r
i
d
sy
st
em
. {7}
Prese
n
t
s
an e
nha
nce
d
c
ont
rol
of
t
h
e
po
we
r fl
ow
s
o
n
a
FC
H
V
i
n
or
der
t
o
re
duc
e t
h
e
hy
d
r
o
g
e
n
co
ns
um
pt
i
on, by
ge
nerat
i
n
g an
d st
ori
ng t
h
e el
ect
r
i
cal
energy
o
n
l
y
at
t
h
e
m
o
st
sui
t
a
bl
e m
o
m
e
n
t
s on
a gi
ve
n
dri
v
i
n
g cy
cl
e. T
h
e
p
r
o
p
o
sed
st
rat
e
gy
w
h
i
c
h ca
n
be i
m
pl
em
ent
e
d
on
-l
i
n
e i
s
ba
sed
on
a f
u
zzy
l
ogi
c
deci
si
o
n
sy
st
em
. A FC
/
U
C
hy
bri
d
ve
hi
cul
a
r
po
we
r sy
st
em
usi
n
g a
wavel
e
t
based l
o
ad s
h
ari
n
g an
d f
u
zz
y
l
ogi
c
base
d co
nt
r
o
l
al
gori
t
h
m
i
s
pr
o
pose
d
by
{8}.
Whi
l
e
w
a
vel
e
t
t
r
ansf
o
r
m
s
are sui
t
a
bl
e for a
n
al
y
z
i
n
g an
d
eval
uat
i
n
g t
h
e
dy
nam
i
c l
o
ad
dem
a
nd
pr
o
f
i
l
e
of
a
hy
b
r
i
d
el
ect
ri
c ve
hi
c
l
e (HE
V
), t
h
e
use
o
f
fuzzy
l
ogi
c
cont
rol
l
e
r i
s
a
p
p
r
op
ri
at
e fo
r
t
h
e hy
bri
d
sy
st
em
cont
r
o
l
.
The
per
f
o
r
m
a
nce
of a
di
rec
t
m
e
t
h
anol
f
u
el
cel
l
(DM
F
C
)
was
m
odel
e
d by
{
9
} usi
n
g
ada
p
t
i
ve-
n
et
w
o
r
k
-ba
s
ed
fu
zzy in
feren
ce system
s (ANFIS). Th
e artificial
neu
r
al
net
w
or
k (A
N
N
) an
d
pol
y
n
o
m
i
al
-based m
odel
s
we
re selected to be com
p
ared with
th
e ANFIS in
respect of qual
ity and accura
cy. Base
d on the ANFIS m
odel obtai
ned, the
cha
r
acteristics of the
DMFC were
stu
d
i
ed
. Th
e
resu
lts show th
at
te
m
p
eratu
r
e an
d m
e
th
an
o
l
c
once
n
tration
greatly affect the
perform
ance
of t
h
e
DMFC. Withi
n
a restricted current
range
,
the m
e
thanol
concentration
does
not
gre
a
tly affect the
stack
v
o
ltag
e
. In
order
to
ob
tain
h
i
g
h
e
r fu
el u
tilizatio
n
effi
cien
cy, th
e m
e
th
an
o
l
con
c
en
tration
s
an
d tem
p
eratu
r
es
shoul
d
be a
d
justed acc
ording to the l
o
ad on t
h
e
system
. Continuous-tim
e
r
ecurre
n
t fuzzy system
s
are
em
pl
oy
ed by
{10} t
o
m
odel
t
h
e el
ect
ri
cal
behavi
or
of a s
o
l
i
d
o
x
i
d
e f
u
el
cel
l
,
whi
c
h b
o
t
h
can
be des
c
ri
be
d
q
u
a
litativ
ely an
d is
k
nown qu
an
titativ
ely fro
m
m
easu
r
em
e
n
ts.
Du
e to th
e tran
sp
aren
cy
o
f
th
e m
o
d
e
l an
easy
esti
m
a
t
i
o
n
of its p
a
ram
e
ters i
s
p
o
ssib
l
e.
Add
itio
n
a
lly, th
e
m
o
d
e
l p
a
ram
e
t
e
rs are
op
ti
m
i
zed
nu
m
e
ricall
y
to
enha
nce t
h
e
m
odel
perf
o
r
m
a
nce. T
h
erm
a
l
m
a
nagem
e
nt
fo
r a s
o
l
i
d
oxi
de f
u
el
cel
l
(SO
F
C
)
i
s
a
c
t
u
al
l
y
te
m
p
eratu
r
e con
t
ro
l, du
e t
o
the i
m
p
o
r
tan
c
e of cell te
m
p
erat
ure
fo
r t
h
e
pe
r
f
o
r
m
a
nce of a
n
S
O
FC
.
A m
odi
fi
e
d
Tak
a
g
i
-Su
g
e
no (T-S)
fu
zzy
m
o
d
e
l th
at is su
itab
l
e fo
r
no
nlin
ear system
s
is b
u
ilt to
m
o
d
e
l th
e SOFC
stack
.
Th
e m
o
d
e
l
p
a
ra
m
e
ters are i
n
itialized
b
y
th
e
fu
zzy c-m
ean
s clu
s
teri
n
g
m
e
t
h
od
, an
d learned
u
s
ing
an off-lin
e
back
-
p
r
opa
gat
i
on al
go
ri
t
h
m
{11}
I
n
o
r
der t
o
o
b
t
a
i
n
t
h
e t
r
ai
ni
ng
dat
a
t
o
i
d
ent
i
f
y
t
h
e m
odi
fi
ed T
-
S m
odel
,
a
SOFC ph
ysical
m
o
d
e
l v
i
a MATLAB is
estab
lish
e
d. The te
m
p
eratu
r
e
m
o
d
e
l is th
e c
e
n
t
re of th
e
ph
ysical
m
odel
and
i
s
devel
ope
d
by
ent
h
al
py
-bal
a
n
ce eq
uat
i
o
ns.
I
t
i
s
sh
ow
n
t
h
a
t
t
h
e m
odi
fi
ed
T-S
f
u
zzy
m
odel
i
s
sufficiently accurate t
o
fo
llow the
tem
p
erature
response
of t
h
e stac
k, a
n
d can be
convenie
ntly utilized t
o
desi
g
n
t
e
m
p
erat
ure c
ont
rol
s
t
rat
e
gi
es. {
12}
Di
scus
ses
ho
w ge
net
i
c
al
g
o
ri
t
h
m
were a
ppl
i
e
d
t
o
opt
i
m
i
ze a
pr
ot
o
n
e
x
c
h
an
ge m
e
m
b
rane
f
u
el
cel
l
st
ack
d
e
si
gn
by
searc
h
i
n
g
fo
r t
h
e
be
st
co
nfi
g
u
r
at
i
o
n i
n
t
e
rm
s of
n
u
m
b
er
of cells and ce
ll surface area. {13} desc
ribe
s the use
of multi obj
ecti
v
e genetic algorithm
s
in the design of
a
fuzzy l
ogic c
o
ntrol system
fo
r a s
o
lid
oxide
fuel cell.
3.
SIM
U
LATI
O
N
RESULTS
OF DY
NA
MI
C
LO
AD
In
th
is p
a
p
e
r the d
y
n
a
m
i
c
lo
ad
u
s
ed
is in
du
ctio
n
m
o
to
r. The p
a
ram
e
ters o
f
in
du
ction
m
o
to
r u
s
ed
are
p
r
esen
ted in
Tab
l
e 1. Sim
u
latio
n
resu
lts
o
f
3
-
Φ
vol
t
a
ge
sou
r
ce
i
nve
rt
e
r
(
V
S
I
)
fe
d us
i
ng Di
ck
s-La
r
m
i
n
ie
m
odel
of
fuel
cel
l
s
fo
r i
n
duc
t
i
on m
o
t
o
r a
r
e
pre
s
ent
e
d,
Si
m
u
l
a
t
i
on are
p
e
rf
orm
e
d o
n
i
n
duct
i
o
n m
o
t
o
r
t
o
f
i
n
d
lo
ad
lin
e to
lin
e Vo
ltag
e
(V
LL
), starting c
u
rre
nt (I
st
), stea
dy state curre
nt (I
stead
y
)
w
h
i
c
h
i
s
ma
x
i
mu
m
v
a
l
u
e
,
%C
ur
rent
T.
H
.
D (
I
T.H.D
), m
a
xim
u
m
Torq
ue (T
ma
x
), Sp
eed co
rres
p
on
din
g
to m
a
xim
u
m
Torq
ue
(N
Tm
ax
),
Efficiency correspondin
g t
o
m
a
xim
u
m
Torque (
η
T
ma
x
), Operating Speed (N
oper
), i
n
duct
i
on m
o
t
o
r ope
rat
i
n
g
Torque (T
I.M
)
,
Lo
ad
To
rqu
e
(T
load
) Operati
n
g E
fficiency
(
η
oper
)
at v
a
r
i
ous lo
ad
ed cond
itio
n
s
, wh
ich ar
e
N
o
l
o
ad, C
o
nst
a
nt
Tor
q
ue l
o
a
d
(
T
c
), T
load
=
k
1
ω
lo
ad, T
load
=k
2
ω
2
lo
ad
= T
lo
a
d
= k
3
ω
3
loa
d
. Where k
1
, k
2
, k
3
are
lo
ad
con
s
tan
t
s.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Mo
del
i
n
g
&
Si
mul
a
t
i
o
n
of
Fu
el
cel
l
(
D
i
cks-Lar
mi
ni
e M
o
d
e
l)
base
d
3-P
h
ase V
o
l
t
a
ge …
(G
a
u
r
a
v Sa
ch
de
v
a
)
69
3
Fi
gu
re
1.
Di
c
k
s-Larm
i
n
i
e
dy
n
a
m
i
c
m
odel
of
equi
val
e
nt
el
ec
t
r
i
cal
ci
rcui
t
o
f
f
u
el
cel
l
3.
1
Si
mul
a
ti
on
R
e
sul
t
s o
f
Ind
u
c
ti
on
Mo
t
o
r b
e
i
n
g Fed fr
o
m
3-
Φ
VSI
Using Dicks-L
a
rminie Model of
Fuel Cell.
Sim
u
l
a
t
i
on a
r
e
pe
rf
orm
e
d
on
i
n
d
u
ct
i
o
n
m
o
t
o
r
bei
n
g
fed
f
r
o
m
3-
Φ
V
S
I
usi
n
g
Di
c
k
s
-
Larm
i
n
i
e
m
odel
of fuel cell usi
n
g following l
o
ad consta
nts.
T
c
= 31
N-m
K
1
=
0
.
70
794
10
83
K
2
=
8
.
23
407
3*
10
-3
K
3
=
9
.
57
706
4*
10
-5
Table
1 indicates the
following:
I.
The sta
r
ting curre
nt (I
st
)
rem
a
i
n
s alm
o
st sam
e
irresp
ectiv
e
v
a
riou
s lo
ad
ing
co
nd
itio
ns.
II.
The %
Curre
nt
T.H.D (I
T.H.D
)
is m
a
x
i
m
u
m
in
case
o
f
No
lo
ad
co
nd
itio
n.
II
I.
T
h
e
ma
x
i
mu
m T
o
r
q
u
e
(
T
ma
x
) re
m
a
in
s irresp
ectiv
e of
v
a
ri
o
u
s lo
ad
ing
co
nd
i
tio
n
s
.
IV.
The stea
dy state curre
nt (I
stead
y
) i
s
depe
n
d
ent
on
t
y
pe l
o
a
d
.
V.
Th
e slip corresp
ond
ing
to m
a
x
i
m
u
m
to
rq
u
e
is m
a
x
i
m
u
m
u
n
d
e
r lin
early load
ed cond
itio
n.
VI.
The E
fficiency
corres
pond
i
n
g
t
o
ma
x
i
mu
m
T
o
r
q
u
e
(
η
Tm
ax
) is al
m
o
st co
nstan
t
irresp
ective o
f
v
a
riou
s
lo
ad
ing
con
d
iti
o
n
s
is aroun
d 80
%.
VI
I.
The operating Efficiency (
η
oper
) is relatively
m
o
re fo
r c
onst
a
nt T
o
r
que
loa
d
.
Tabl
e
1.
Per
f
o
r
m
a
nce pa
ram
e
ters
fo
r
vari
ous
l
o
ads
o
f
i
n
d
u
ct
i
on m
o
t
o
r
bei
n
g
fed
f
r
om
3-
Φ
VS
I usi
n
g Di
c
k
s-
Larm
in
ie
m
o
d
e
l o
f
fu
el cell.
P
a
ra
m
e
t
e
r Noload
T
c
T
l
oad
=k
1
ω
T
l
oad
= k
2
ω
2
T
l
oad
= k
3
ω
3
V
LL
402.
90
8
259.
79
8
253.
76
6
264.
35
3
267.
00
6
I
st
48.
416
48.
480
1
48.
413
3
48.
411
7
48.
411
6
I
st
e
a
d
y
18.
110
5
16.
101
9
32.
794
5
31.
099
3
30.
708
1
I
T.
H
.
D.
31.
406
6
7.
0982
75
8.
7804
9
10.
284
87
10.
679
33
T
ma
x
60.
866
6
61.
484
61.
083
2
58.
740
8
59.
032
7
N
Tm
a
x
821.
02
826.
48
6
777.
63
7
804.
96
9
810.
62
η
Tm
a
x
-
82.
648
6
77.
763
7
80.
496
9
81.
062
T
I.
M
.
-
32.
443
2
54.
437
1
56.
838
4
55.
953
6
T
l
oad
-
31
53.
953
9
56.
222
8
53.
932
5
N
o
p
er
-
970.
39
4
731.
46
3
789.
07
8
788.
57
7
η
o
p
er
-
97.
039
4
73.
146
3
78.
907
8
78.
857
7
3.
2
Si
mul
a
ti
on R
e
sul
t
s o
f
Ind
u
c
t
i
o
n
Mo
t
o
r be
i
n
g Fed fr
om
3-
Φ
VSI
Usin
g Dicks
-
Lar
m
i
nie Model o
f
Fuel cell when Conduction
Angle Change
s fr
om
140
0
to
15
0
0
with
Constant Tor
que
Load.
Sim
u
l
a
t
i
on ha
v
e
bee
n
p
e
r
f
o
r
m
e
d
on i
n
d
u
ct
i
o
n m
o
t
o
r
bei
n
g
fed
fr
om
3-
Φ
VSI
usi
n
g
Dic
k
s-
Larm
inie
m
odel
of
fuel
cel
l
whe
n
c
o
n
duct
i
n
g a
n
gl
e
chan
ges
fr
om
14
0
0
to 150
0
with
Con
s
tan
t
Torqu
e
l
o
ad to
fi
nd
vari
at
i
o
n of %
C
ur
rent
T
.
H
.
D
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE Vo
l. 4
,
N
o
. 5
,
O
c
tob
e
r
20
14
:
691
–
6
96
69
4
Tabl
e 2. Vari
at
i
on o
f
%
C
u
r
r
e
n
t
T.H
.
D
w
ith
co
ndu
ctio
n
an
gle
f
o
r
3
-
Φ
VSI
fed u
s
i
n
g Dicks-Larm
in
ie m
o
d
e
l
o
f
fu
el cell feed
ing
indu
ction
m
o
to
r with
constan
t
Torq
u
e
load
.
COD
UCTION A
N
GLE
IN DEGREES
% CU
RRENT
T.H
.
D
140
17.
600
32
141
17.
240
81
142
16.
972
02
143
16.
672
95
144
16.
383
34
145
15.
917
79
146
15.
795
01
147
15.
748
32
148
15.
757
7
149
15.
774
44
150
15.
791
63
Tabl
e 2 s
h
ows
vari
at
i
o
n o
f
%
C
ur
rent
T.
H.
D
wi
t
h
C
o
n
duct
i
on
A
ngl
e f
o
r 3
-
Φ
V
S
I
fed
usi
ng m
odel
o
f
fu
el cell feed
i
n
g
indu
ctio
n
m
o
to
r with
con
s
tan
t
Torq
u
e
lo
ad
. Tab
l
e 2
indicates that as the conduction a
ngle
increases the
%Current
T.H.D dec
r
eases for 3-
Φ
VSI fed u
s
ing
Dick
s-Larm
in
ie
m
o
d
e
l o
f
fu
el cell
feed
i
n
g
in
du
ctio
n m
o
to
r wit
h
co
n
s
tan
t
Torqu
e
l
o
ad.
3.
3
ANN B
a
sed
Control
Str
a
te
gy
for
Ind
u
cti
o
n Motor
In t
h
i
s
pape
r, a
cont
r
o
l
st
rat
e
g
y
for i
n
d
u
ct
i
o
n
m
o
t
o
r bei
n
g f
e
d fr
om
3-
Φ
V
S
I usi
ng
Di
cks
-
Larm
i
n
i
e
m
odel
of fuel
c
e
l
l
whi
c
h fee
d
s
C
onst
a
nt
Tor
q
ue l
o
a
d
(T
c
) is
p
r
op
o
s
ed
wh
ich
u
s
es m
u
ltilay
e
r b
a
ck
prop
agatio
n
feed
fo
rwa
r
d neu
r
al
net
w
o
r
k. T
h
i
s
co
nt
r
o
l
st
rat
e
gy
gi
ve
s go
o
d
est
i
m
a
t
e
of co
n
duct
i
on a
n
gl
e. The
neu
r
al
net
w
or
k t
a
kes
% C
u
r
r
ent
T
.
H.
D (
X
) as i
n
put
a
nd
gi
ve
s
C
o
n
d
u
ct
i
on
A
ngl
e
(C
A
)
co
r
r
esp
o
ndi
ng t
o
t
h
at
%
Cu
rren
t T.H.D (X). Th
e n
e
t
w
ork
is set wi
th
‘log
sig’ activ
atio
n
fun
c
tion
at th
e m
i
d
d
l
e layer an
d
pu
relin
activ
atio
n
fu
nctio
n
at th
e o
u
t
pu
t layer. Th
e d
e
sign
o
f
t
h
e net
w
or
k and sel
ect
i
o
n o
f
opt
i
m
u
m
t
r
ai
ni
n
g
param
e
t
e
rs are
per
f
o
r
m
e
d by
t
r
i
a
l
an
d er
r
o
r.
Furt
herm
or
e,
t
a
nsi
n
g
fu
nct
i
o
n i
s
use
d
w
h
i
c
h ca
uses
fewe
r
ep
ochs
as co
m
p
ared
t
o
o
t
h
e
r train
i
ng
fun
c
tio
n
s
. Th
erefore,
wh
en an
inp
u
t
is app
lied
in
t
h
e n
e
twork, it will b
e
g
i
n
t
r
ai
ni
n
g
base
d
on t
h
e
gi
ve
n
dat
a
i
n
o
r
de
r
t
o
pr
o
du
ce t
h
e ap
prox
im
ate
resu
lts. Th
e resu
lts ob
tain
ed
fro
m
Neural Netwo
r
k
are v
e
rified
with
th
e
resu
lts
th
at are ob
tained
fro
m
Ex
cel Cu
rv
e
fittin
g
.
Fi
gu
re
2.
B
l
oc
k
di
ag
ram
of A
N
N
base
d
co
nt
rol
st
rat
e
gy
f
o
r
i
n
d
u
ct
i
o
n m
o
t
o
r
Fi
gu
re 2 sh
o
w
s bl
oc
k di
a
g
ram
of AN
N
base
d cont
rol
st
rat
e
gy
for
i
n
d
u
ct
i
on m
o
t
o
r
.
Fi
gu
re 2
in
d
i
cates th
at th
e con
t
ro
l str
a
teg
y
tak
e
s % Cu
rr
en
t T.H
.
D
(X
)
as inp
u
t
and
g
i
v
e
s Co
nductio
n
An
g
l
e (
C
A
)
as
out
put
.
3.
4
Results
a
nd V
erificati
on of
AN
N
B
a
sed
C
o
nt
rol Str
a
te
g
y
for
Ind
u
cti
o
n
M
o
tor
.
Tabl
e 3
sh
o
w
s
resul
t
s
a
n
d
v
e
ri
fi
cat
i
on
of
cont
rol
st
rat
e
g
y
for
i
n
duct
i
o
n m
o
t
o
r fe
d
fr
om
3-
Φ
VSI
usi
n
g
Di
cks
-
La
rm
i
n
i
e
m
odel
o
f
f
u
el
cel
l
w
h
i
l
e fee
d
i
n
g c
onst
a
nt
T
o
r
que
l
o
a
d
(T
c
).
Tabl
e
3. R
e
s
u
l
t
an
d
veri
fi
cat
i
o
n
of
co
nt
r
o
l
st
r
a
t
e
gy
o
f
i
n
d
u
ct
i
on m
o
t
o
r
X
CA fro
m
neural n
e
tw
ork
CA fro
m
Exce
l C
u
rve fit
t
ing
6.
07
140.
62
5
140.
61
4
6.
03
141.
64
5
141.
64
53
6.
015
142.
12
24
142.
11
89
5.
97
143.
82
83
143.
82
45
5.
95
144.
71
41
144.
71
95
5.
935
145.
43
94
145.
69
88
5.
92
146.
22
7
146.
22
01
5.
89
147.
90
76
147.
91
04
5.
88
148.
50
71
148.
51
6
5.
872
149.
02
37
149.
01
57
X
CA
Artificial Neu
r
al
Netw
or
k
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
ECE
I
S
SN
:
208
8-8
7
0
8
Mo
del
i
n
g
&
Si
mul
a
t
i
o
n
of
Fu
el
cel
l
(
D
i
cks-Lar
mi
ni
e M
o
d
e
l)
base
d
3-P
h
ase V
o
l
t
a
ge …
(G
a
u
r
a
v Sa
ch
de
v
a
)
69
5
Tabl
e
3 i
n
di
cat
es t
h
at
AN
N
b
a
sed c
o
nt
rol
st
r
a
t
e
gy
i
s
gi
vi
n
g
sat
i
s
fact
ory
res
u
l
t
s
f
o
r
i
n
duct
i
on
m
o
t
o
r.
Tabl
e 4. Vari
at
i
on o
f
%C
ur
re
nt
T.
H.
D wi
t
h
con
d
u
ction
a
n
g
l
e
fo
r vari
ous
p
o
we
r facto
r
s (P
Fs) fo
r 3-
Φ
VS
I
fe
d
u
s
ing
Dick
s-Larm
in
ie
m
o
d
e
l of fu
el cell
CA in
degrees
PF=0.9
PF=0.8
PF= 0
.
7
PF =
0
.
6
PF =
0
.
5
PF =
0
.
4
PF =0
.3
PF =0
.2
140
12.
469
89
8.
6965
47
6.
3016
82
5.
5546
18
5.
0603
3
4.
7192
15
4.
4830
04
4.
3270
14
141
12.
052
69
8.
4262
79
6.
2972
05
5.
5631
33
5.
0821
51
4.
7524
83
4.
5284
23
4.
3755
02
142
11.
649
3
8.
1819
5
6.
2983
72
5.
5779
22
5.
1083
83
4.
7907
88
4.
5709
47
4.
4334
25
143
11.
261
25
7.
9666
68
6.
2996
36
5.
5947
47
5.
1383
54
4.
8423
9
4.
6247
55
4.
4963
15
144
10.
892
6
7.
7852
48
6.
3060
23
5.
6156
26
5.
1721
62
4.
8723
9
4.
6795
38
4.
5597
52
145
10.
545
5
7.
6411
38
6.
3148
57
5.
6404
86
5.
2097
4.
9265
7
4.
7347
68
4.
6238
3
146
10.
224
33
7.
5382
92
6.
3285
74
5.
6688
72
5.
2517
82
4.
9796
54
4.
8058
66
4.
7013
32
147
9.
9322
66
7.
4812
16
6.
3452
51
5.
7010
33
5.
2971
48
5.
036
4.
8732
53
4.
7778
11
148
9.
6732
27
7.
4739
28
6.
3637
78
5.
7419
58
5.
3468
96
5.
0965
24
4.
9438
1
4.
8542
74
149
9.
4510
17
7.
4779
56
6.
3894
38
5.
7762
37
5.
4021
95
5.
1615
74
5.
0129
27
4.
9321
71
150
9.
2690
96
7.
4868
92
6.
4164
92
5.
8183
02
5.
4530
68
5.
2244
31
5.
0881
61
5.
0131
14
Table 4
sh
o
w
s
t
h
e vari
at
i
on
of % cu
rre
nt
T.H
.
D wi
t
h
co
nd
uct
i
o
n an
gl
e
for
vari
ous
p
o
we
r fact
o
r
s
(PFs
) fo
r 3-
Φ
VSI
fe
d
usi
n
g
Di
cks
-
Larm
i
n
i
e
m
odel
of
f
u
el
cell. He
re the
current ta
ke
n i
s
phase c
u
rre
nt.
Table
4 indicates the
following:
1.
For
PF
=
0.
9, a
s
t
h
e c
o
nd
uct
i
o
n a
ngl
e
va
ri
o
u
s
fr
om
14
0
0
-1
50
0
the Curre
nt T
.
H.D
dec
r
eases
.
2.
For
PF =
0.
8, a
s
t
h
e co
nd
uct
i
o
n an
gl
e va
ri
o
u
s
fr
om
140
0
-15
0
0
th
e Cu
rren
t T.H.D
d
ecreases u
n
til 14
0
0
and
then i
n
creas
es.
3.
For
PF =
0.
7, a
s
t
h
e co
nd
uct
i
o
n an
gl
e va
ri
o
u
s
fr
om
140
0
-15
0
0
th
e Cu
rren
t T.H.D
d
ecreases u
n
til 14
1
0
and
then i
n
creas
es.
4.
Fo
r PF = 0.6, 0
.
5
,
0
.
4
,
0.3, 0
.
2
,
as th
e co
ndu
ctio
n angle v
a
r
i
ou
s
f
r
om
1
4
0
0
-1
50
0
t
h
e Cu
rre
nt T
.
H.
D
increases
.
5.
As the Power
Factor (PF) de
creases t
h
e Current T
.
H.D dec
r
eases.
6.
T
h
e
% C
u
r
r
e
n
t
T
.
H
.
D
i
s
mi
n
i
mu
m a
t
1
4
0
0
fo
r PF = 0.
2.
7.
T
h
e
% C
u
r
r
e
n
t
T
.
H
.
D
i
s
ma
x
i
mu
m a
t
1
4
0
0
fo
r PF = 0.
9.
4.
CO
NCL
USI
O
N
In
t
h
e prese
n
t
wo
rk
, per
f
o
r
m
a
nce of
i
nve
rt
er
w
h
i
l
e
fee
d
i
n
g
di
f
f
ere
n
t
p
o
we
r fact
or
l
o
ads has bee
n
investigate
d
. Fuel cells are
used in
t
h
e inpu
t sid
e
as a so
urce
wh
ile in th
e ou
tpu
t
si
d
e
d
y
n
a
m
i
c lo
ad
is
considere
d
. T
h
e followi
ng conclu
si
ons
are
d
r
aw
n
fo
r
In
d
u
c
t
i
on m
o
t
o
r l
o
a
d
i
.
e.
dy
nam
i
c
l
o
ad.
1.
The Starting
c
u
rrent (I
st
) remain
s alm
o
st irresp
ectiv
e
o
f
variou
s lo
ad
ing
co
nd
itio
ns.
2.
The %
Curre
nt
T.H.D (I
T.H.D
)
is m
a
x
i
m
u
m
in
case
o
f
No
lo
ad
co
nd
itio
n.
3.
T
h
e
ma
x
i
mu
m T
o
r
q
u
e
(
T
ma
x
) re
m
a
in
s irresp
ectiv
e of
v
a
ri
o
u
s lo
ad
ing
co
nd
i
tio
n
s
.
4.
The stea
dy state curre
nt (I
stead
y
) i
s
depe
n
d
ent
on
t
y
pe l
o
a
d
.
5.
Th
e slip corresp
ond
ing
to m
a
x
i
m
u
m
to
rq
u
e
is m
a
x
i
m
u
m
u
n
d
e
r lin
early load
ed cond
itio
n.
6.
The E
fficienc
y
correspondi
ng to m
a
xim
u
m
Torque
(
η
Tma
x
) is al
m
o
st co
nstan
t
irrespectiv
e of v
a
rious
lo
ad
ing
con
d
iti
o
n
s
.
7.
The operating Efficiency (
η
oper
) is relatively
m
o
re fo
r c
onst
a
nt T
o
r
que
loa
d
.
8.
ANN
gives
very fast and acc
urate res
u
lts
and
it
can
b
e
u
s
ed
for o
n
lin
e
calcu
latio
n
s
.
REFERE
NC
ES
[1]
O Erding, B Vu
ral, M Uzunog
lu
. “A wave
let log
i
c based
energ
y
management st
r
a
teg
y
for
a fu
el
cell/batter
y
/ultr
a-
capa
c
itor
h
y
brid
vehicu
lar pow
er
s
y
s
t
em
”.
Journa
l of Power
Sources
. 2009; 194(1): 369-380.
[2]
A Sakhare,
A Davari,
A Feliach
i. “Fuzzy
logic
control of fuel
cell
for stand-alon
e and grid conn
ection
”
.
Journal of
Power Sources
. 2004;
135(1-2): 165-176.
[3]
D Gao, Z Jin,
Q Lu. “Energ
y
mana
gement strateg
y
based on fuzzy
l
ogi
c for a fuel cel
l h
y
br
i
d
bus
”.
Journal of
Power Sources
.
2008; 185(1): 31
1-317.
[4]
J
O
S
c
hum
acher,
P
Gem
m
a
r, M
De
nne, M
Zedd
a, M Stueb
e
r. “Control of
mini
atu
r
e proton
ex
chan
ge membrane fu
el
cells based
on fu
zzy
logic”.
Jour
nal of Power
So
urces
. 129(2)
: 1
43-151.
[5]
MC Kisacikoglu, M Uzunoglu, MS Alam
.
“Load sharing using fuzzy
logic co
ntrol in a fuel cell/u
ltr
a capaci
t
o
r
h
y
brid
veh
i
cle”.
International Jo
urnal
of Hydrog
en En
ergy
. 2009
; 34(3): 1497-15
07.
[6]
KS Jeong, WY Lee, CS Kim.
“Energ
y
manag
e
ment strategies
of a fuel
cel
l\b
atter
y
h
y
brid
s
y
stem using fuzzy
logics”.
Journal
of Pow
e
r Sources
. 2005; 145
(2):
319-326.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-87
08
I
J
ECE Vo
l. 4
,
N
o
. 5
,
O
c
tob
e
r
20
14
:
691
–
6
96
69
6
[7]
S Caux, W Ha
nkache, M Fadel,
D His
s
e
l. “
O
n-line fuzz
y
e
n
erg
y
m
a
nag
e
m
e
nt for h
y
br
id fuel ce
ll s
y
s
t
em
s
”
.
International Jo
urnal
of Hydrog
en En
ergy
. In Pr
ess, Corrected
Proof, Availab
l
e o
n
line 13 Janu
ar
y 2010.
[8]
O Erdinc, B Vur
a
l, M Uzunoglu
,
Y Ates. “Modeling and an
aly
s
is
of an FC/UC h
y
brid vehicular p
o
wer s
y
stem using
a wavelet-
fuzzy
logic b
a
sed
load
sharing
and con
t
rol
algorithm”.
International Jo
urnal of Hydrog
en En
ergy
. 2009
;
34(12): 5223-52
33.
[9]
R Wang, L Qi,
X Xie, Q Ding,
C Li, CCM
Ma. “Modeling of
a 5-cell dire
ct
m
e
thanol fuel c
e
ll us
ing adapt
i
v
e
-
network-based
f
u
zzy
infe
re
nc
e s
y
ste
m
s”
.
Journa
l of Power
Sources
. 2008; 185(2): 1201-1208.
[10]
A Flemming, J Adam
y
.
“Modeling
solid ox
ide fuel
cel
ls using continuous-time r
ecurren
t
fuzzy
s
y
s
t
ems”.
Engineering App
lications of
Ar
ti
f
i
cial
Int
e
ll
igen
ce
. 2008; 21(8): 12
89-1300.
[11]
J Yang, X Li, HG Mou, L
Jian. “Control-oriented thermal ma
nagement of solid o
x
ide fuel cells based on a modified
Takag
i
-Sugeno f
u
zzy
model”.
Jo
urnal of
Power S
ources
. 2009; 18
8(2): 475-482
.
[12]
I Mohamed, N Jenkins. “Proton exchang
e
memb
rane (PEM) fuel cell stack c
onfiguration using genetic algor
ithm”.
Journal of Power Sources
. 2004
; 131(1-2): 1
42-1
46.
[13]
F Jurado, M Valverda. “E
nhancing the electrical
perform
ance of
a s
o
lid oxid
e
fue
l
ce
ll using multi objective genetic
algorithm”.
Ren
e
wable
Energ
y
.
2005; 30(6): 881
-902.
BI
O
G
R
A
P
HY
OF
A
U
T
HO
R
Prof. GK Sachdeva has don
e B.Tech (
E
lectr
i
cal
E
ngg), M.Tech
(
P
ower Engg.).
He is curren
tly
pursuing Ph.D
in Elec
tri
cal E
ngg. He has more than 10
y
e
ars of Research
& Teaching
experi
enc
e
a
t
T
h
apar Univ
ers
i
t
y
,
P
unjabi
Univ
ers
i
t
y
,
Love
l
y
P
r
ofes
s
i
onal Univ
ers
i
t
y
.
He has
authored 3 inter
n
ation
a
l and 2 nation
a
l books.
He has published various
papers in reputed
nation
a
l &
i
n
ter
n
ation
a
l confer
e
n
ce & journals.
Evaluation Warning : The document was created with Spire.PDF for Python.