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
.
5
,
Octo
ber
201
9,
pp. 352
2
~3
530
IS
S
N: 20
88
-
8708
,
DOI: 10
.11
591/
i
j
ece
.
v9
i
5
.
pp3522
-
35
30
3522
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Predi
ctive t
or
qu
e contr
ol o
f
electr
ic vehicl
e
Moham
med E
l Amin
Ab
del
ko
ui,
Abdel
dje
ba
r
Ha
z
z
ab
La
bora
toi
re
de
R
ec
her
che Com
ma
nde, Ana
l
y
se
e
t
Opt
imis
at
ion
de
s S
y
stèmes E
lect
ro
-
éne
rgé
ti
qu
es,
U
nive
rsité T
AH
RI
Moham
ed
de
Bec
har
,
Alg
eria
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
Dec
25
, 201
8
Re
vised
Ma
r
2
5
, 2
01
9
Accepte
d
Apr
4
, 2
01
9
The
foll
owing
a
rti
cle
rep
r
ese
nts
the
dev
el
opm
ent
of
a
tra
c
ti
on
s
ystem
of
an
el
e
ct
ri
ca
l
veh
ic
l
e
(EV)
that
consis
t
of
two
Thre
e
-
p
hase
squire
l
-
ca
g
e
induction
m
otors
(IM)
tha
t
per
m
it
the
dr
iv
e
of
th
e
two
f
ro
nt
drivi
ng
whee
l
s.
The
two
mo
tors
are
controlle
d
usin
g
th
e
Predicti
ve
Tor
que
Contro
l
(P
TC)
m
et
hod;
A
te
chn
ique
ba
sed
on
th
e
n
e
xt
step
pre
d
ic
t
i
on
and
evalua
t
ion
of
the
el
e
ct
rom
agne
t
ic
torque
and
stat
or
flux
In
a
cost
func
ti
on
in
orde
r
to
det
ermina
te
the
inve
rt
er
sw
it
ch
i
ng
vector
th
at
m
ini
m
iz
e
th
e
er
ror
bet
we
e
n
ref
ere
n
ce
s
and
pre
dicted
v
al
ues
.
PTC
is
what
we
tri
ed
to
und
erl
in
e
in
th
is
pape
r,
so
we
e
xpla
in
be
low
the
princ
ip
le
of
t
he
m
et
hod;
and
the
s
y
stem
m
at
hemati
c
al
d
e
script
ion
is
provide
d.
An
el
e
ct
ro
nic
diff
ere
n
ti
a
l
i
s
appl
ie
d
o
n
the
s
y
s
te
m
to
co
ntrol
ind
epe
n
d
en
tly
th
e
spee
d
of
the
two
whe
el
s
at
diff
ere
n
t
oper
ating
cond
i
ti
ons
in
ord
er
to
ch
ara
c
te
r
iz
e
the
dr
ivi
ng
wh
ee
l
s
y
st
em
beha
vior
,
the
rob
ustness i
n
ste
ad
y
state
and
in
tr
an
sient
st
at
e
.
Ke
yw
or
d
s
:
Dr
i
ving
w
heels
Ele
ct
ro
nic
d
if
f
eren
ti
al
Ele
ct
ric
v
e
hicle
(E
V)
Ind
uction
m
ac
hin
e
(I
M
)
Pr
e
dicti
ve
t
orq
ue
c
on
t
ro
l
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
:
Moh
am
m
ed
El Am
in A
bdel
koui,
Lab
or
at
oi
re
de R
echerc
he
C
om
m
and
e,
An
al
yse
et
O
pt
i
m
isa
t
ion
des
S
yst
è
m
es Elec
tro
-
é
nergéti
ques,
U
ni
ver
sit
é T
A
HRI
M
oh
am
ed
,
BP 417, Be
c
ha
r
(
0800
0),
Alge
ria
.
Em
a
il
: dr
.abde
lkoui@
gm
ail.co
m
1.
INTROD
U
CTION
Ele
ct
ric
veh
ic
le
s
(EV)
ha
ve
gaine
d
incr
eas
ing
popula
rity
ov
e
r
the
la
st
decad
e
i
n
the
autom
otive
sect
or
,
it
re
pr
e
sents
a
ne
w
al
te
rn
at
ive
t
ow
a
rd
s
w
hich
to
t
urn
to
en
sure
pr
ese
r
vation
of
the
e
nvir
on
m
ent
by
reducin
g
em
issi
on
s
ca
us
ed
by
the
us
e
of
i
nter
nal
com
busti
on
e
ngine
ve
hicle
s
(
ICVs
)
,
a
nd
so
l
ving
energy
pro
blem
s
du
e
to
t
he
de
pleti
on
of
f
os
sil
fu
e
ls
[1
,
2].
The
dev
el
op
m
ent
of
(E
V’
s
)
ha
s
know
n
a
sig
nificant
adv
a
nce
d
an
d
get
enou
gh
perform
ance,
boost
ed
by
the
big
i
m
pr
ov
em
ent
and
dev
el
op
m
ent
of
el
ect
ric
m
oto
rs,
batte
ries,
an
d
high
co
ntr
ol
te
chnolo
gies.
Howev
e
r,
the
m
os
t
rem
ark
able
adv
a
ntage
of
t
he
EV
is
that
we
can
con
t
ro
l
the
m
otor
to
rque
m
uch
m
or
e
quic
kly
an
d
preci
s
el
y
com
par
i
ng
with
c
onve
ntion
al
(I
C
V’
s
)
[
3
,
4].
Ind
uction
m
ac
hin
es
(
IM)
are
widely
us
ed
in
(E
V)
ap
plica
ti
on
s
due
to
th
ei
r
low
cost
an
d
high
pe
rform
ances
and r
obus
t
ness
and als
o
f
or th
ei
r
hi
gh starti
ng
to
r
qu
e
.
The
Direct
T
orq
ue
Con
t
ro
l
(D
TC)
re
plac
es
the
fiel
d
ori
ented
c
on
tr
ol
(F
OC)
in
hig
h
dynam
ic
app
li
cat
io
ns
,
t
his
m
et
ho
d
re
qu
i
res
neit
her
a
m
od
ulat
or
nor
a
n
inte
rn
al
current
P
I
c
ontrolle
r.
T
hese
featur
e
s
m
ake
the
syst
em
i
m
ple
m
entation
easi
e
r
an
d
le
ad
to
a
fast
dynam
ic
resp
onse.
H
ow
e
ver,
du
e
to
it
s
str
uc
ture
,
the
m
ai
n
prob
l
e
m
s
of
t
his
str
at
egy
are
t
he
high
le
vel
of
tor
que
a
nd
flu
x
rip
ples
a
nd
th
e
var
ia
ble
s
witc
hing
fr
e
qu
e
ncy.
Wh
ic
h
can
le
a
d
to
decr
ease
the
c
on
t
ro
l
perform
ances;
an
d
in
cr
ease
noise
s
an
d
co
n
tr
ol
di
ff
ic
ulty
at
low
s
peed
[5
-
7].
I
n
the
la
st
few
decad
es
,
var
io
us
m
et
ho
ds
were
pro
pos
e
d
to
overc
om
e
these
dr
a
wb
ac
ks,
su
c
h
as
m
ulti
l
evel
co
nverter
s,
arti
fici
al
int
el
li
gen
ce
te
ch
niques
a
nd
fi
xe
d
s
witc
hing
f
reque
ncy
m
od
ulati
on
te
chn
iq
ues
li
ke
the
sp
ace
ve
c
tor
m
od
ulati
on
(S
VM
),
th
os
e
t
echn
i
qu
e
s
al
w
ay
s
In
crea
se
th
e
con
tr
ol
al
gor
it
h
m
com
plexity
lea
ding to
an exte
ns
ive
softwa
re/hard
war
e c
ompu
ta
ti
onal
r
e
qu
irem
ent [
8].
The
m
od
el
pr
edict
ive
co
ntr
ol
(MPC)
m
e
thod
ha
s
inc
r
eased
it
s
stu
dy
and
resea
rc
h
since
it
s
introd
uction
i
nt
o
the
fiel
d
of
el
ect
rical
engi
neer
in
g
in
t
he
1980s
by
re
searche
rs
li
nk
ed
to
aca
dem
i
a
and
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
Predict
iv
e tor
que c
on
tr
ol
of elec
tri
c vehicl
e (
Mo
hamm
e
d El
Amin
Abdelk
oui
)
3523
industry,
MPC
is
m
ai
nly
us
ed
to
in
crease
t
he
ef
fici
ency
a
nd
pe
rfor
m
ance
of
syst
em
’s
respon
ses
an
d
t
o
s
olv
e
pro
blem
s
of
a
uto
m
at
ion
an
d
co
ntr
ol
of
a
wide
ra
ng
e
of
de
vices
a
nd
industrial
proc
esses
that
pre
sent
a
com
plex
natu
r
e
[
9],
MPC
c
oncept
us
es
a
s
yst
e
m
m
od
el
to
pr
e
dict
the
f
uture
syst
em
s
ta
te
s
in
disc
rete
tim
e
ste
ps
.
Re
ce
ntly
,
pr
e
dicti
ve
tor
qu
e
co
ntr
ol
(P
TC
)
strat
e
gies
ha
ve
rec
ei
ved
wide
at
te
ntion
i
n
r
es
earc
h
com
m
un
it
ie
s
du
e
to
t
heir
int
uiti
ve
featu
res,
easy
i
m
ple
m
e
ntati
on
,
an
d
ea
sy
inclusio
n
of
no
nlinearit
ie
s
an
d
const
raints
of
m
od
el
pr
edict
ive
co
ntro
l
(MP
C)
[10,
11]
.
It
is
a
te
chn
iqu
e
base
d
on
DTC
con
t
ro
l
pri
ncip
le
s
on
the
one
ha
nd;
and
on
the
pre
dicti
on
of
t
he
s
ta
tor
c
urren
t
a
nd
fl
ow
usi
ng
t
he
MPC
m
et
ho
ds
on
the
oth
er
ha
nd,
to
al
low
the
ca
lc
ulati
on
of
the
torque
in
t
he
nex
t
insta
nt,
a
nd
al
s
o
to
pr
e
ve
nt
the
f
uture
beh
a
vior
a
nd
r
eact
ion
of the syst
em
i
n order
to
i
ncr
e
ase it
s p
e
rfor
m
ance.
Ele
ct
ric
vh
ic
le
s
con
t
r
ol
te
chni
qu
es
are
vast
fiel
d
an
d
there
is
too
m
any
m
et
ho
ds
that
m
ay
be
ei
ther
slow
pe
rfor
m
ance
or
e
xpe
ncive
e
quipm
ent
ho
wev
e
r
in
the
w
ork
that
f
ollow
s
a
si
m
ulati
on
unde
r
m
at
la
b/Si
m
uli
nk
of
PTC
int
egr
at
io
n
f
or
th
e
con
t
ro
l
of
t
wo
i
nductio
n
m
achines
ens
uri
ng
the
tract
i
on
of
an
el
ect
ric
veh
ic
l
e
is
pr
e
sente
d.
This
te
ch
ni
que
al
lows
est
im
at
ion
a
nd
t
he
nex
t
ste
p
pr
e
di
ct
ion
of
the
dri
ving
par
am
et
ers
when
cha
ng
i
ng
re
sist
ive
torque
or
dr
i
ving
co
nd
i
ct
ion
s
are
ap
plied,
it
per
m
it
s
t
o
antic
ipate
ch
ang
e
s
and
react
quic
kly
in
order
to
pro
vid
e
a
bette
r
c
on
tr
ol
for
t
he
dri
ver,
faster
acce
le
rati
on
a
nd
gr
eat
er
sta
bil
it
y
of
the E
V.
This
pap
e
r
is
s
tructu
red
as
fol
lows
:
in
sect
ion
I
I,
th
e
m
at
he
m
at
ic
al
m
od
el
of
a
n
i
nduction
m
achine
(I
M)
a
nd
the
a
pp
li
ed
volt
age
so
urce
in
ver
te
r
are
p
re
sente
d.
In
sect
io
n
I
II,
the
ve
hicle
m
od
el
is
prese
nted
as
well
as
the
char
ges
a
nd
re
sist
ive
torques
eq
uations.
I
n
sect
ion
IV,
the
P
TC
m
et
ho
d
is
exp
la
ine
d.
Sec
ti
on
V
pr
ese
nts t
he
re
su
lt
s and t
he
c
orres
pondin
g
a
naly
sis. Finall
y, the c
oncl
us
i
ons a
re
giv
e
n
i
n
sect
ion
VI.
2.
INDU
CTIO
N MA
CHINE
A
ND INVE
RTER’S M
ODE
LS
The
m
achine
consi
der
e
d
i
n
this
pap
e
r,
is
a
three
-
phase
s
qu
i
rr
el
-
cage
in
duct
io
n
m
achine.
The
m
at
he
m
at
i
cal
m
od
el
of
a
n
in
duct
ion m
oto
r
ca
n be e
xpr
essed by t
he
fo
ll
ow
in
g
e
qu
at
i
on
s
[1
1]:
=
.
+
(1)
0
=
.
+
−
.
.
(2)
=
.
+
.
(3)
=
.
+
.
(4)
=
3
2
.
.
{
∗
.
}
(5)
=
−
(6)
w
he
re
is
the
st
at
or
vo
lt
ag
e
ve
ct
or
;
and
are
t
he
sta
to
r
an
d
r
otor
cu
rr
e
nts.
and
rep
re
sent
the
sta
to
r
flu
x
an
d
ro
t
or
f
lux
,
res
pecti
vel
y.
an
d
are
the
sta
tor
a
nd
r
otor
resist
an
ces.
,
and
are
sta
tor
,
r
oto
r
and
m
utu
al
in
du
ct
a
nce,
respec
ti
vely
,
and
is
the
el
ect
rica
l
sp
eed.
is
the
nu
m
ber
of
pole
pairs
,
an
d
denotes t
he
el
e
ct
ro
m
agn
et
ic
torq
ue.
The
to
po
l
og
y
of
the
tw
o
-
le
ve
l
vo
lt
age
s
ource
inv
e
rter
a
ppli
ed
in
this
work
for
the
PTC
and
it
s
feasible
vo
lt
a
ge
v
ect
ors a
re
presented
in Fi
gu
re
1.
Figure
1. Left:
Tw
o
-
le
vel
volt
age s
ource i
nverter,
Ri
gh
t:
V
oltage
vecto
rs
2
3
1
4
0
7
5
6
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.
5
,
Oct
ober
20
19 :
3522
-
3530
3524
The
s
witc
hing
sta
te
S
can
b
e
expresse
d by t
he follo
wing
ve
ct
or
[
7
]:
=
2
3
(
+
+
²
)
(7)
w
he
re
=
2
/
3
,
w
hen
=
1
m
eans
on,
m
eans
off,
a
nd
=
,
,
.
The
vo
lt
age
ve
ct
or
is
relat
ed
t
o
the s
witc
hing s
ta
te
by
=
,
(8)
w
he
re
is t
he d
c li
nk
volt
age.
3.
THE
V
EHI
C
LE
M
O
DEL
The
ve
hicle
con
si
der
e
d
i
n
t
he
a
naly
sis
an
d
ta
r
get
f
or
th
e
i
m
ple
m
entat
i
on
of
the
pr
opos
e
d
co
ntr
ol
syst
e
m
is
a
tra
ct
ion
syst
em
F
igure
2.
Starti
ng
f
ro
m
an
usua
l
veh
ic
le
str
uctur
e
,
so
m
e
adapt
at
ion
s
are
i
n
c
our
s
e
with the
ob
j
ect
ive of i
ntrod
uc
ing
t
wo in
dep
e
nd
e
nt
fron
t
w
he
el
s p
r
opulsi
on syst
e
m
u
sin
g
e
le
ct
ric d
ri
ves
.
Figure
2. Ve
hi
cl
e g
eom
et
ry and
dri
vi
ng whe
el
s contro
l
syst
e
m
4.
DIFFE
RENT
IAL SPEE
D REFE
RE
NCE
S C
O
MP
UT
ATIO
N
It
is
possible
t
o
determ
ine
the
sp
ee
d
re
fere
nces
ver
s
us
t
he
requirem
ents
of
the
dr
i
ver.
Wh
en
t
he
veh
ic
le
a
rr
i
ves
at
the
be
ginn
ing
of
a
cu
r
ve
,
the
dri
ve
r
a
ppli
es
a
c
urve
ang
le
on
it
s
w
heel.
T
he
el
ect
ronic
diff
e
re
ntial
acts
i
m
m
ediat
el
y
on
t
he
tw
o
m
o
tors
re
duci
ng
the
dri
ving
wheel
sp
eed
sit
ua
te
d
inside
the
cu
r
ve
increasin
g
t
hereby
the
s
peed
of
t
he
dri
vi
ng
wh
eel
ou
tsi
de
the
cu
r
ve.
T
he
dr
i
ving
w
heels
angu
la
r
sp
ee
ds
are
a
s
fo
ll
ows
[
12, 13]
.
=
ℎ
+
∆
(9
)
=
ℎ
−
∆
(10)
=
+
1
co
rr
es
pondin
g
to
a
c
ho
ic
e
of
the
directi
on
of
the
w
heel,
(
−1
)
t
he
rig
ht
t
urn,
an
d
(
+1
)
the
le
ft
t
urn.
The dri
ving
w
heel s
peed va
ri
at
ion
is im
po
se
d by the t
raj
ect
or
y
desire
d by
the drive
r.
It is
giv
e
n by:
∆
=
2
(
+
)
ℎ
(11)
∗
∗
Electr
i
c diff
erentia
l
Co
n
trol
Strategy
Co
n
trol
Strategy
2
.
,
,
,
,
IM
IM
Inv
erter
Inv
erter
Battery
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
Predict
iv
e tor
que c
on
tr
ol
of elec
tri
c vehicl
e (
Mo
hamm
e
d El
Amin
Abdelk
oui
)
3525
The
c
orrelat
io
n
betwee
n
α
w
hich
is
the
c
ur
ve
a
ng
le
gi
ven
by
the
dri
ve
r
w
heel
a
nd
δ
of
t
he
real
cu
r
ve
ang
le
of the
wheel
s is
giv
e
n by:
=
(12)
The
s
pee
d refe
ren
ces
of t
he
t
wo m
oto
rs
a
re:
ω
mR
∗
=
N
red
ω
rR
(13)
ω
mL
∗
=
N
red
ω
rL
(14)
5.
RESIST
AN
T
TORQ
UE O
F
A
N
EL
ECT
R
IC VEHI
CLE
To
determ
ine
the
to
r
qu
e
re
qu
i
red
al
lowi
ng
el
ect
ric
ve
hicle
s
to
ove
rco
m
e
the
resist
ance
f
orces,
s
om
e
factors
hav
e
to be to
ke
n
in
c
onsiderati
on, t
hes
e fact
ors a
re
presented
as
fo
ll
ow [12,
14
,
15
]
:
5.1.
The roll
ing
re
sista
nc
e
=
×
(15)
=
Roll
ing R
esi
sta
nce
=
Gross
V
e
hicle
W
ei
ght
=
Co
-
e
ff
ic
ie
nt
of Rolli
ng Resi
s
ta
nce
5.2.
The gr
ad
e
res
istan
ce
=
×
(16)
=
Grade
r
esi
sta
nc
e
=
Grade
or incli
nation an
gle
5.3.
The accel
er
at
i
on
force
=
×
(17)
=
/
Accele
rati
on
f
or
ce
m
ass o
f
the
v
e
hicle
acce
le
rati
on due to
gra
vity
(
9
.
81
2
⁄
)
re
qu
ire
d
ac
cel
eraio
n
5.4.
The t
otal t
r
ac
tive
e
ffort
The
T
otal T
rac
ti
ve
Ef
fort ca
n be calc
ulate
d
a
s:
=
+
+
(18)
=
Total
att
racti
ve
effo
rt
5.5.
Torque
requir
ed on the d
ri
ve
wheel
The
t
orqu
e
that
is re
qu
i
red o
n t
he
dri
ve
whee
l wil
l i
s:
=
×
×
(19)
=
Torq
ue
=
Fr
ic
ti
on f
act
or that acc
ount for f
rict
ion
al
lo
sses b
et
wee
n b
earin
gs
, a
xles e
tc
.
=
Ra
diu
s
of
dr
i
ve
wheel
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.
5
,
Oct
ober
20
19 :
3522
-
3530
3526
6.
PREDI
CTI
V
E T
ORQUE
CONTR
OL O
F THE E
LE
C
TRIC E
V
The
desi
gn
of
the
pr
e
dicte
d
t
orq
ue
c
on
t
ro
l
(
PTC)
te
c
hn
i
que
is
re
presente
d
i
n
Fi
gure
3,
i
t’s
based
on
the
pr
e
dicti
on
of
the
sta
to
r
c
urren
t
a
nd
fl
ux
fo
r
al
l
feasibl
e
vo
lt
age
vector
s
in
ord
e
r
to
regulat
e
the
tor
que
in
the
tim
e
per
io
d
betwee
n
an
d
+
1
,
an
d
al
so
a
desig
n
of
a
cos
t
functi
on.
on
l
y
the
first
nex
t
volt
age
ve
ct
or
ste
p
i
s c
onside
red in
ne
xt sect
ion
[
1
6
].
Figure
3.
Bl
oc
k diag
ram
o
f
th
e predict
ive t
orqu
e
contr
ol
The
first
ste
p
sta
tor
c
urre
nts
are
cal
culat
ed
us
i
ng
e
qu
at
io
ns
(1),
(
3)
a
nd
(4)
from
the
IM
m
od
e
l
descr
i
bed in
pr
evio
us
secti
on,
it
can be
desc
r
ibed
a
s foll
ows
:
=
−
1
(
(
.
−
.
(
1
−
.
)
.
)
−
)
(20)
w
he
re
=
⁄
,
=
+
2
.
an
d
=
.
Using t
he
f
orw
ard Eule
r discr
et
iz
at
ion
, we c
an pre
dict t
he
nex
t
ste
p value
≈
(
+
1
)
−
(
)
(21)
is t
he
sam
plin
g
ti
m
e o
f
the
s
yst
e
m
.
W
it
h (20) an
d (21), the
stat
or
current ca
n be
pr
e
dicte
d
as:
̂
(
+
1
)
=
(
1
−
)
.
(
)
+
1
.
[
.
(
1
−
.
(
)
)
.
(
)
+
(
)
]
(22)
=
⁄
The
ne
xt
-
ste
p
sta
tor
flu
x
̂
(
+
1
)
a
nd
t
he
el
ect
r
oma
gn
et
ic
t
orq
ue
̂
(
+
1
)
are
cal
c
ulate
d,
us
i
ng
(21)
t
o
discreti
ze the
volt
age m
od
el
(1), the
stat
or f
l
ux pre
dicti
on
obta
ined
is as
f
ol
lows
[7
]
̂
(
+
1
)
=
̂
(
)
+
.
(
)
−
.
.
(
)
(23)
Accor
ding
to
(
5),
with
pre
dicti
on
s
of
the
sta
tor
fl
ux
(23)
a
nd
t
he
predict
e
d
cu
rr
e
nt
(
21),
the
el
ect
ro
m
a
gn
et
ic
tor
qu
e
obtai
ne
d:
̂
(
+
1
)
=
2
3
.
.
{
̂
(
+
1
)
∗
.
̂
(
+
1
)
}
(24)
w
it
h
the
pr
e
dicte
d
va
lues
of
tor
que
a
nd
st
at
or
flu
x,
el
ect
ric
tor
que
a
nd
sta
to
r
flu
x
m
agn
it
ud
e
c
ontrol
is
ob
ta
ine
d
by
t
he
m
ini
m
iz
ati
on
of
a
co
st
functi
on
for
wh
ic
h
t
he
in
puts
a
re
t
he
to
rque
re
fer
e
nce
∗
,
the
pr
e
dicte
d
tor
qu
e
̂
(
+
1
)
,
r
efe
r
ence
sta
tor
fl
ux
|
|
∗
an
d
the
pr
e
dicte
d
sta
tor
fl
ux
|
̂
(
+
1
)
|
m
agn
it
ud
e
s,
=
[
(
∗
−
̂
(
+
1
)
]
2
2
+
(
|
|
∗
−
|
̂
(
+
1
)
|
)
2
|
|
2
(25)
IM
Cost f
unct
ion
∗
∗
|
|
∗
̂
(
+
1
)
̂
(
+
1
)
̂
(
)
̂
(
)
,
,
Torque
&
Flux
pre
dictions
Flux
esti
m
at
ions
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
Predict
iv
e tor
que c
on
tr
ol
of elec
tri
c vehicl
e (
Mo
hamm
e
d El
Amin
Abdelk
oui
)
3527
w
he
re,
=
0
…
6
,
co
rr
e
s
pondin
g
to
t
he
7
diff
e
ren
t
volt
age
vect
or
s
of
the
tw
o
le
vel
so
urce
i
nv
e
rter
app
li
e
d
in
this
syst
e
m
;
fo
r
each
sta
t
or
volt
age
ve
ct
or
a
vaila
ble,
this
c
os
t
f
un
ct
io
n
is
evaluate
d,
a
nd
the
sta
to
r
v
ol
ta
ge
⃗
[
]
pro
du
ci
ng
the
m
ini
m
u
m
cost
is
sel
ect
ed
to
be
a
ppli
ed
on
m
oto
r
te
rm
ina
ls.
W
ei
gh
t
gai
ns
an
d
|
|
corres
pond to
the
rated t
orque
v
al
ue
and
flu
x refe
re
nce
duri
ng no
rm
al
sp
eed
op
e
rati
on [1
7
].
7.
SIMULATI
O
N RESULTS
Using
t
he
m
od
el
s
ho
w
n
in
Figure
3,
we
cond
ucted
s
im
ulati
on
s;
i
n
t
he
interest
of
c
har
act
erise
the
dr
i
ving
w
he
el
syst
e
m
behavio
ur.
The
F
i
gure
4
re
prese
nt
s
the
ve
hicle
s
peed
re
spo
nse
us
in
g
cl
assic
DTC
and PTC
.
Figure
4
.
Ve
hi
cl
e sp
ee
d
re
spo
ns
e
us
in
g
cl
ass
ic
D
TC a
nd PT
C
co
ntro
l
T
he
ne
xt
fig
ures
pr
ese
nt
veh
i
cl
e
sp
eed
va
riat
ion
for
PTC
“Pre
dicti
ve
Torqu
e
Co
ntr
ol“.
we
us
e
th
e
PTC
al
gorithm
,
instea
d
of
us
ing
cl
assic
al
con
t
ro
l,
in
t
he
interest
of
im
pr
ov
i
ng
t
he
co
ntro
l
lo
op
rob
us
t
ness.
The
ca
pacit
y
of
the
al
gorithm
to
m
ai
ntain
ideal
tr
ajecto
ries ind
e
pende
ntly
of
t
he
e
xter
nal d
ist
urba
nces
a
nd
th
e
par
am
et
er
var
i
at
ion
s
w
hic
h
m
akes
the
r
obus
tness
t
her
e
f
ore
the
a
dv
a
nta
ge
of
this
c
on
t
ro
l.
T
o
s
how
t
he
ef
fect
of
dist
urba
nce
s
by
resist
ive
tor
que
of
PTC
con
t
ro
l,
Fi
gure
s
below
s
hows
the
syst
e
m
res
pons
es
of
the
two
cases.
7.1.
Ca
se
of s
tr
aight wa
y
In
t
his
te
st,
we
hav
e
tw
o
cases
,
the
first
on
e
is
strai
ght
r
oa
d
without
a
sl
op
e
show
n
i
n
F
i
gures
5,
7,
9
;
than
we
ha
ve
s
trai
gh
t
ro
a
d
wi
th
10%
a
slo
pe
show
n
i
n
F
i
gures
6,
8,
10
;
th
e
syst
e
m
is
subm
i
tt
ed
to
the
sam
e
sp
ee
d
ste
p.
O
nl
y
a
cha
ng
e
of
the
de
velo
ped
m
oto
r
to
rque
is
noti
ced
si
nc
e
the
dr
i
ving
wh
eel
s
sp
ee
ds
sta
y
al
ways
the
sa
m
e
and
t
he
r
oa
d
slo
pe
does
not
af
fect
the
c
ontr
ol
of
the
w
he
el
.
The
slo
pe
eff
ect
resu
lt
s
i
n
high
i
m
pr
ovem
ent i
n
the
elec
trom
agn
et
ic
m
o
tor
t
orq
ue.
Figure
5
.
Ve
hi
cl
e sp
ee
d
in
a s
trai
gh
t
way
without sl
op
e
Figure
6
.
Ve
hi
cl
e Sp
ee
d
in
a
strai
gh
t
way
w
it
h
slo
pe
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.
5
,
Oct
ober
20
19 :
3522
-
3530
3528
Figure
7
.
I
M T
orq
ue
in
a strai
gh
t
way
without s
lo
pe
Figure
8
.
I
M T
orq
ue
in
straig
ht w
ay
with
slo
pe
Figure
9
.
I
M
C
urren
t
s
in
a
str
ai
gh
t
way
with
ou
t
Slo
pe
Figure
10
. IM
Current
i
n
a st
r
ai
gh
t
way
with
Slo
pe
7.2.
Ca
se
of c
urve
d wa
y
The
ve
hicle
is
dri
vi
ng
on
a
curve
d
r
oa
d
on
the
le
ft
si
de
with
50km
/h
sp
eed
.
We
as
sum
e
that
tw
o
m
oto
rs
are
not
disturbe
d.
T
he
dr
i
ving
w
hee
ls
fo
ll
ow
diff
e
r
ent
pat
hs
,
a
nd
they
tur
n
in
t
he
sam
e
directio
n
bu
t
with
diff
e
re
nt
sp
ee
ds
du
e
to
t
he
el
ect
ronic
diff
e
re
ntial
who
act
s
on
the
two
m
oto
r
s
pee
ds
by
decr
easi
ng
th
e
sp
ee
d
of
t
he
dri
vin
g
w
heel
on
the
le
ft
side
sit
uated
i
ns
ide
t
he
c
urve,
an
d
on
the
ot
her
side
by
inc
reasi
ng
th
e
wh
eel
m
oto
r
s
peed
i
n
the
e
xt
ern
al
sid
e
of
th
e
curve.
I
n
this
te
st,
we
ha
ve
two
cases
,
the
first
on
e
is
a
c
urve
d
ro
a
d
without
a
slop
e
sho
wn
i
n
F
ig
ures
11,
13,
15
;
t
han
we
ha
ve
c
urve
d
r
oa
d
with
10%
a
sl
op
e
show
n
i
n
F
igures
12,
14,
16
.
Figure
1
1
. Ve
hi
cl
e’s
w
heels s
peed in a
cur
ve
d
way with
out sl
op
e
Figure
1
2
. Ve
hi
cl
e’s
w
heels s
peed in a
cur
ve
d way
with sl
op
e
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
Predict
iv
e tor
que c
on
tr
ol
of elec
tri
c vehicl
e (
Mo
hamm
e
d El
Amin
Abdelk
oui
)
3529
Figure
1
3
.
Ri
ght
IM
To
rque i
n
a c
urve
d way
without Sl
op
e
Figure
1
4
.
Ri
ght
IM
To
rque i
n
a c
urve
d way
with Sl
op
e
Figure
15
. Rig
ht I
M
’s
sta
to
r
Currents i
n
a
C
urve
d
way
with
out
Sl
op
e
Figure
16
.
Ri
ght
I
M
’s
sta
to
r C
urren
ts i
n
a c
urve
d
way with
Slo
pe
Althou
gh
that
the
con
tr
ol
with
Pr
e
dicti
ve
Torqu
e
co
nt
ro
ll
er
offe
rs
be
tt
er
per
f
or
m
ances
in
both
con
t
ro
l
an
d
tra
ckin
g,
but
thes
e
figures
show
that
the
eff
ect
of
the
distu
rbance
is
ver
y
low
in
the
case
of
the
PTC.
Mo
reove
r
to
these
dyna
m
ic
per
form
ances,
the
i
m
po
sed
co
ns
trai
nts
are
res
pected
by
the
dr
i
ving
s
yst
e
m
su
c
h
as t
he ro
bustness
a
nd p
a
r
a
m
et
er v
ariat
io
ns
.
8.
CONCL
US
I
O
N
In
t
his
pa
per,
the
pre
dicte
d
to
rque
c
on
tr
ol
te
chn
i
qu
e
has
be
en
ap
plied
a
nd
si
m
ulate
d
to
con
t
ro
l
tw
o
inducti
on
m
ac
hin
es
with
el
e
ct
ronic
dif
f
ere
ntial
us
ed
in
t
he
str
uctu
re
of
a
tract
io
n
sy
stem
fo
r
a
n
el
ect
ric
veh
ic
le
.
PTC
i
s
a
si
m
ple
te
c
hn
i
qu
e
base
d
on
tw
o
c
onditi
on
s;
the
first
one
is
DTC
swi
tc
hin
g
ta
ble,
a
nd
t
he
seco
nd
one
is
the
pr
e
dicti
on
of
the
cu
rr
e
nt,
torque
an
d
flux
nex
t
ste
p.
The
res
ult
sh
ows
that
this
con
t
ro
l
pro
vid
es
a
r
ob
us
t
co
ntr
ol;
this
adv
a
ntage
gi
ves
PTC
m
eth
od
ve
ry
go
od
perform
ances
in
bo
t
h
ste
ad
y
and
transient
sta
te
s
.
In
a
dd
it
io
n
to
that;
the
PTC
te
chn
iq
ues
s
ho
ws
bette
r
to
rque
an
d
sp
ee
d
r
esp
on
se
a
nd
a
m
or
e
flexible
c
ontro
l
schem
e
than
DTC.
This
pa
per
gi
ves
a
n
instru
ct
io
n
f
or
the
sel
ect
io
n
of
M
PC
m
et
h
od
s
f
or
eng
i
neer
i
ng pr
act
ic
es.
REFERE
NCE
S
[1]
X.
Shi
and
M.
Krishnam
urth
y
,
“
Digit
al
Control
of
Induc
ti
on
M
ac
hin
e
--
A
Potent
ia
l
Ba
ckup
Control
Strat
eg
y
fo
r
Fault
To
le
r
ant
C
ontrol
in
Elec
t
ri
c
Vehi
cl
e
Appli
ca
t
ions
,
”
I
EEE
J
ournal
of
Eme
rg
ing
and
Se
lecte
d
Topics
in
Power
El
e
ct
ronics
,
v
ol
.
2,
pp.
651
-
658,
Sep
2014
.
[2]
T.
Furn
y
a,
et
al
.
,
“
Im
ple
m
ent
at
i
on
of
adva
nce
d
adhe
sion
cont
ro
l
for
el
ectri
c
veh
ic
l
e,
”
in
Proc.
I
EE
E
Workshop
Adv
anc
ed
Mo
ti
o
n
Control
,
vol
.
2
,
pp
.
430
-
435
,
1
996.
[3]
Y.
Hori,
e
t
al
.
,
“
Tra
c
ti
on
Contro
l
of
El
e
ct
ri
c
Vehi
cl
e
:
Basic
Expe
r
imenta
l
R
esult
s Us
ing
th
e
Te
st EV
“
UO
T
El
ec
tric
Marc
h
,
”
IE
EE Tr
ansacti
on
s On I
ndustry
Applications
,
vol
.
34
,
pp
.
1131
-
1138,
199
8
.
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.
5
,
Oct
ober
20
19 :
3522
-
3530
3530
[4]
M.
Doum
ia
ti
,
et
al
.
,
“
D
y
namics
Control
of
an
In
-
W
hee
l
Elec
t
r
ic
Vehic
l
e
with
St
ee
r
-
by
-
W
ire
,
”
1
7th
Inte
rnationa
l
IEE
E
Confe
ren
c
e
on
In
telli
g
ent
Tr
anspo
rtati
on
S
yste
ms
(
ITSC
)
,
pp.
348
-
353
,
Oc
t 2014
.
[5]
D.
Casade
i
,
e
t
a
l
.
,
“
FO
C
and
DTC:
Two
Viable
Schemes for
Indu
ct
ion
Motors T
o
rque
Control
,
”
I
EE
E
Tr
ansacti
o
ns
On P
ower
E
lectr
onic
s
,
v
o
l. 17, p
p.
779
-
787
,
Sep
2002
.
[6]
A.
Amm
ar
,
et
a
l
.
,
“
Closed
loop
torque
SV
M
-
D
TC
base
d
on
robust
super
twisti
ng
spee
d
cont
ro
ll
er
for
induction
m
otor
drive
wit
h
eff
icien
c
y
opt
i
m
iz
at
ion
,
”
In
te
r
nati
onal
Journal
of
Hydrogen
E
nergy
,
v
ol
.
42
,
pp
.
17940
-
1795
2,
Jul
2017.
[7]
F.
W
ang,
et
al
.
,
“
Model
Based
Predic
ti
v
e
Dir
ec
t
Contr
ol
Str
at
eg
ie
s
for
Elec
tri
c
al
Drive
s:
A
n
Expe
riment
al
Eva
lu
at
ion
of
PTC
and
PC
C
M
et
hods
,
”
IE
EE
Tr
ansacti
ons
on
I
ndustrial
Informatic
s
,
v
ol
.
11,
p
p
.
671
-
681,
Jun
2015
.
[8]
A.
A
mmar
,
et
a
l
.
,
“
Predic
t
ive
Di
rec
t
Torque
Con
trol
with
Redu
ced
Rippl
es
and
Fuzz
y
Logic
Spee
d
Control
le
r
fo
r
Induc
ti
on
Motor
Drive
,
”
The
5
t
h
Inte
rnational
Confe
renc
e
on
El
e
ct
rica
l
Engi
n
ee
ring
–
Boume
rdes
(
ICEE
-
B)
,
Alger
ia
,
Oc
t
201
7.
[9]
J.
L.
D
.
Ma
drid,
et
al
.
,
“
Predic
t
i
ve
Contro
l
of
a
Furuta
Pendulu
m
,
”
IE
EE
3rd
C
olombian
Conf
e
renc
e
on
Au
tomatic
Control
(
CCAC)
,
Oct
2017
.
[10]
M.
Preindl
and
S.
Bologna
ni
,
“
Model
Predictiv
e
Dire
c
t
Torqu
e
Control
W
it
h
Fi
nit
e
Con
trol
Set
for
PM
SM
Driv
e
S
y
stems
,
Part
2:
Fiel
d
W
ea
k
eni
n
g
Opera
ti
on
,
”
I
E
EE
T
rans
act
ions
On
Industrial
Informatic
s
,
v
ol
.
9,
pp
.
648
-
65
7
,
M
ay
2013
.
[11]
Md.
Habibul
la
h
,
et
al
.
,
“
A
Sim
p
li
fie
d
Fini
te
-
St
ate
Predic
t
ive
Dir
ec
t
Torqu
e
Con
t
rol
for
Induc
ti
o
n
Motor
Drive
,
”
IEE
E
Tr
ansacti
o
ns On Industrial
El
e
ct
ronics
,
v
ol
.
63,
pp
3964
-
39
75,
Jun 2016
.
[12]
A.
Nasri,
e
t
al
.
,
“
Two
W
hee
l
Speed
Robust
Slidi
ng
ModeCont
r
ol
for
El
e
ct
ri
c
Ve
hic
l
e
Drive
,
”
Se
rbian
Journal
o
f
El
e
ct
rica
l
Eng
in
ee
ring
,
v
ol
.
5
,
p
p
.
199
-
216
,
Nov
2008.
[13]
J.
La
rm
inie
and
J.
Lowr
y
,
“
El
e
ctric
Veh
ic
l
e Tech
nolog
y
Exp
la
in
e
d,
”
John
W
il
e
y
&
Sons
,
Eng
la
nd
,
2003
.
[14]
M.
khessam
,
et
a
l
.
, “
Fuzz
y
Adap
t
ive
Control
for
Dire
ct
Torque
in
El
e
ct
ri
c
Vehi
cl
e
,
”
Int
ernati
onal Journal
of
Pow
e
r
El
e
ct
ronics
and
Dr
iv
e
Syst
em
(
IJ
PE
DS)
,
v
ol
.
4,
p
p.
557
-
566
,
De
c 2014
.
[15]
S.
Chauha
n,
“
Motor
Torque
C
alculations
For
Elec
tr
ic
Vehi
cle
,
”
Inte
rnational
Jo
urnal
Of
Sci
entif
ic
&
Technol
ogy
Re
search
,
v
ol
.
4
,
pp
.
126
-
127,
Aug 2015.
[16]
B.
Zhu
,
et
al
.
,
“
Predic
ti
ve
Tor
que
Control
wi
t
h
Ze
ro
-
Sequ
ence
Curre
nt
Suppress
ion
for
Ope
n
-
End
W
indi
ng
Induc
ti
on
Mac
hi
ne
,
”
2015
I
EEE
Industry
Applica
ti
ons Soc
ie
t
y An
nual
Me
et
ing
,
D
ec
2015
.
[17]
H.
Miranda,
e
t
al
.
,
“
Predictive
Torque
Contro
l
of
Induc
t
ion
Mac
hine
s
Bas
ed
on
State
-
Space
Models
,
”
I
EEE
Tr
ansacti
ons On Industrial
El
e
ctr
onic
s
,
v
ol
.
56
,
p
p
1916
-
1924,
Jun 2009
.
BIOGR
AP
H
I
ES
OF
A
UTH
OR
S
Mohamme
d
El
Amin
Abdel
kou
i
was
born
in
1
990
at
Bec
h
ar
-
Alger
i
a,
re
ceive
d
t
he
stat
e
Mast
er
degr
ee
in
Auto
m
at
ic
&
Tele
co
m
m
unic
at
ion
en
gine
er
ing
in
20
13
from
the
Uni
ver
sit
y
of
Ta
hri
Moham
m
ed
of
Bec
har
,
Alger
ia
.
He’s
cur
r
ent
l
y
pre
p
ari
ng
hi
s
Ph.d.
degr
ee
in
el
e
ct
roni
c
engi
ne
eri
ng.
Ab
deldjebar
H
a
z
z
a
b
rec
ei
v
ed
his
Stat
e
Eng
ine
er
,
M.S.
,
an
d
Ph.D
degr
ee
s
in
Elec
tr
ical
Engi
ne
eri
ng
fro
m
the
Elec
tr
ica
l
Engi
n
ee
ring
I
nstit
ute
of
Th
e
Univer
sit
y
of
Scie
nc
es
and
Te
chno
log
y
of
Oran
(US
TO),
Alger
ia
in
1995
,
1999,
and
2006,
respe
c
ti
ve
l
y
.
H
e
is
cur
ren
t
l
y
a
Profess
or
of
El
ec
trica
l
Eng
ineeri
ng
at
the
Univer
sit
y
of
Be
cha
r
(
Alger
ia
)
,
where
he
has
bee
n
the
Dire
ct
or
of
th
e
Resea
rch
L
abo
rat
or
y
of
Com
m
and,
Anal
y
s
es,
and
Optimizat
ion
of
El
e
ct
ro
-
Ene
rge
ti
c
S
y
st
e
m
s
since
2009
a
t
2014.
His
rese
arc
h
int
er
ests
in
cl
ude
power
qu
a
li
t
y
,
m
odel
ing
,
m
oder
n
cont
rol
l
er
and
observe
r
design
for
non
linear
s
y
stems
,
co
ntrol
of
power
e
le
c
troni
cs
and
m
ult
idri
ve
s
y
s
tem
s,
cont
rol
of
p
ower
el
e
ct
roni
cs
,
m
ult
idri
v
e
s
y
st
ems
and
el
ectri
c
al
veh
ic
l
e,
and
ada
pt
ive
con
trol
and
nonli
ne
ar
sy
stems
dia
gnosti
c,
ad
apt
iv
e
cont
r
ol,
neur
al
ne
tworks
and
fuz
z
y
logi
c
s
y
st
ems
.
Evaluation Warning : The document was created with Spire.PDF for Python.