Internati
o
nal
Journal of P
o
wer Elect
roni
cs an
d
Drive
S
y
ste
m
(I
JPE
D
S)
V
o
l.
6, N
o
. 3
,
Sep
t
em
b
e
r
2015
, pp
. 43
3
~
43
8
I
S
SN
: 208
8-8
6
9
4
4
33
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
/
IJPEDS
Influence of Sensorless Contro
l on the Noise of Switched
Reluctan
ce M
o
tor Dri
v
e
Alexander
Petrushin,
Maxim Tchavych
alov
Rostov State Tr
ansport Univ
e
r
sity
, Russia
n
Fe
de
ra
tion
Article Info
A
B
STRAC
T
Article histo
r
y:
Received
Mar 22, 2015
Rev
i
sed
May 18
, 20
15
Accepte
d
J
u
n 2, 2015
The influen
ce o
f
the switched reluc
t
an
ce motor drive sensorless control on
the lev
e
l of i
t
s noise is conside
r
ed.
Th
eoret
i
c
a
ll
y jus
t
i
f
ied
the i
n
creas
e o
f
noise level while sensorless
S
R
M
control
.
The
r
e
s
u
lts
of s
w
it
che
d
relu
ctan
ce
motor noise measurement with
sensor
less control and con
t
rol using the
ph
y
s
ic
al
position
sensor ar
e giv
e
n
.
Keyword:
Mag
n
e
tic asymmetry
Sens
orl
e
ss
co
n
t
rol
Switche
d reluc
t
ance
m
achine
Noise
Copyright ©
201
5 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
:
Al
exa
nde
r Pet
r
ushi
n,
Ro
stov
State Tran
spo
r
t
Un
iv
ersity,
R
u
ssi
an
Fe
dera
t
i
on,
Em
ail: alex331685@yandex.ru
1.
INTRODUCTION
As a rule, m
odern drive system
s are built on the
ba
sis of induction elec
tric
m
achines. Howe
ver in
recent years
ot
her type
s of el
ectric m
o
tors becom
e
in
crea
singly popula
r
. For e
x
am
pl
e, a large
proport
i
on
of
recent
publications is
de
vote
d
to t
h
e study
of s
w
itc
he
d reluctance electric m
achines (SRM) which c
a
n be
cal
l
e
d o
n
e
of
t
h
e m
o
st
pr
om
isi
n
g
t
y
pes
of
electrom
echanical
energy c
o
nverters.
For effective use of
SRM it is necessa
ry to c
ont
ro
l
p
h
a
se ex
citatio
n
cu
rren
t in
acco
r
d
a
nce with
ro
tor
p
o
s
ition
.
Ty
p
i
cally th
is is
d
o
n
e
u
s
ing p
h
y
sical ro
to
r
p
o
sitio
n
sen
s
o
r
fi
x
e
d
on
th
e m
ach
in
e sh
aft
(o
pt
pel
ect
r
o
ni
c senso
r
s, Hal
l
-
effect
sens
or
s,
et
c). Fo
r sy
st
em
s i
n
t
e
nded t
o
operat
e
i
n
ha
r
s
h en
vi
r
onm
ent
s
t
h
e
p
o
s
ition
sen
s
or m
a
y b
eco
m
e
a li
m
i
t
i
n
g
facto
r
. Mean
wh
ile it is po
ssi
b
l
e t
o
d
e
term
in
e roto
r
po
sitio
n ind
i
rectly
u
s
ing
d
e
p
e
nd
en
ce
o
f
ph
ase
fl
u
x
of the ph
ase curren
t
and
ro
to
r
po
sitio
n
ψ
=
f
(
i
,
θ
) [1]. C
ontrol system
s based
on
th
e ind
i
rect
ro
t
o
r po
sitio
n estimatio
n
are call
e
d
sen
s
o
r
less syste
m
s.
C
u
r
r
ent
l
y
rese
arche
r
s
pu
bl
i
s
hed
m
a
ny
wor
k
s
on
sy
nt
hesi
s an
d st
u
d
y
o
f
sens
orl
e
ss c
o
nt
r
o
l
o
f
SR
M
[2].
Typically, the m
e
thods
of indirect rot
o
r position est
i
m
a
tion a
r
e
cl
assified
on m
e
thods
of active phase
m
easurem
ent
and m
e
t
hods
ba
sed o
n
passi
ve
phase m
easur
em
ent
.
Usi
n
g t
h
e m
e
t
hods o
f
t
h
e fi
rst
g
r
ou
p
l
i
m
i
t
s
t
h
e act
i
v
e pha
se curre
nt
co
n
t
rol
and
hence
t
h
e SR
M
t
o
rq
ue co
nt
rol
.
A
m
ong t
h
e
m
e
tho
d
s o
f
seco
n
d
gr
o
u
p
t
h
ere are m
e
t
h
ods
, base
d o
n
sort
v
o
l
t
a
ge
pulses (sen
sing
pu
lses) app
licatio
n
[3
]. In
th
is case it
is p
o
ssib
l
e to
u
s
e effectiv
e alg
o
rith
m
s
o
f
activ
e ph
ase con
t
ro
l.
Howev
e
r it is nece
ssary to consi
d
er t
h
e im
pact of s
e
nsing
current on ene
r
gy effective
n
es
s inde
xe
s and torque curve when the se
nsin
g curre
n
t is commensurate with the
activ
e ph
ase cu
rren
t. On
th
e
o
t
h
e
r h
a
nd
, if t
h
e sen
s
ing
cu
rren
t
is sm
al
l it i
s
n
ecessary to
p
r
o
t
ect u
s
efu
ll
sig
n
a
l
fr
om
noi
se.
An
ot
he
r ap
pr
o
ach al
l
o
ws t
o
al
l
o
cat
e
m
e
t
hods base
d o
n
t
h
e use of m
a
rk
ers an
d m
e
t
hods base
d o
n
l
o
o
k
-
u
p t
a
bl
es
whi
c
h re
qui
re
not
onl
y
a l
a
rg
e am
ount
of m
e
m
o
ry
of t
h
e m
i
cropr
ocess
o
r co
nt
r
o
l
sy
st
em
s, but
al
so t
i
m
e
-cons
um
i
ng t
o
det
e
r
m
i
n
e t
h
e l
ook
-
up t
a
bl
e dat
a
.
I
n
ad
di
t
i
on, l
o
o
k
-
u
p t
a
bl
es do
not
t
a
ke i
n
t
o
a
ccou
n
t
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l.
6, No
. 3, Sep
t
em
b
e
r
2
015
:
43
3 – 438
43
4
the diffe
re
nces
in the characte
r
istics
of the phases which m
a
y arise as a
result of the inac
curacies
during SRM
m
a
nufact
uri
n
g
proce
ss [4]
.
M
e
t
h
o
d
s base
d o
n
t
h
e use of m
a
rkers ar
e general
l
y
l
i
m
i
t
e
d t
o
t
h
e r
o
t
a
t
i
o
n
fre
que
ncy
or
e
x
ci
t
a
t
i
on c
u
r
r
e
n
t
co
nt
r
o
l
.
SR
M
sens
orl
e
s
s
co
nt
r
o
l
en
ha
nces a
ppl
i
cat
i
o
n
of t
h
e s
w
itched rel
u
ctance
dri
v
e system
s
in the m
o
st
ad
v
e
rse
o
p
e
ratin
g
cond
itio
ns: h
i
gh
tem
p
e
r
atu
r
e, ch
em
ically ag
g
r
essi
ve en
v
i
ron
m
en
t, in
ten
s
e
rad
i
atio
n
expos
u
re
to the m
o
tor.
In thi
s
case, of
pa
ra
m
ount
im
portance
t
h
e fact of
the
possibility of
SRM
operation
with
ou
t
p
h
y
sical ro
t
o
r
po
sition
sen
s
o
r
. Accord
i
n
g to
t
h
e level o
f
reliab
ility and
d
u
rab
ility
sen
s
o
r
less
swit
ch
ed
relu
ctan
ce drive ex
ceed
s
its
nearest co
m
p
etito
rs. Th
e situ
atio
n
is
d
i
fferent in
th
e
field
of gen
e
ral ind
u
strial
electric d
r
iv
e, wh
en
ex
cep
t
reliab
ility
an
d
du
rab
ility
o
t
h
e
r ind
i
cato
r
s are i
m
p
o
r
tant (en
e
rg
y effi
cien
cy,
electrom
a
gnetic torque ri
pple
and the
noise l
e
vel).
2.
R
E
SEARC
H M
ETHOD
To
con
s
id
er the in
flu
e
n
ce o
f
sen
s
o
r
less co
n
t
ro
l on
th
e SR
M n
o
i
se lev
e
l it is n
ecessary t
o
id
en
tify th
e
m
a
i
n
reaso
n
s
o
f
i
t
s
occ
u
rre
nc
e. SR
M
noi
se
s
o
urces a
r
e clas
sified as
follows
[5,
6]:
Noi
s
e
f
r
om
t
h
e st
at
or
vi
b
r
at
i
o
n;
No
ise fro
m
radial fo
rces acting
o
n
th
e
ro
to
r;
No
ise fro
m
th
e lo
ad and
add
itio
n
a
l sou
r
ces;
Aer
o
dy
nam
i
c noi
se;
Electronic
noise sources
(s
ources
of
noise
associated
wi
th powe
r el
ect
ronic com
p
one
n
ts operating i
n
switch
i
ng
m
o
de and
with
the
ch
ar
acteristics o
f
con
t
ro
l algorith
m
)
.
In add
itio
n to
t
h
e classification
as a so
urce
of
n
o
i
se t
h
e torqu
e
ripp
le are co
n
s
i
d
ered
[7
].
C
onsi
d
eri
ng t
h
e i
n
fl
ue
nce
of
sens
orl
e
ss c
o
nt
rol
on
n
o
i
s
e an
d vi
brat
i
o
n l
e
v
e
l
i
t
i
s
necessar
y
t
o
speci
fy
that the greate
s
t i
m
pact on t
h
ese in
dicators will provide inaccurate indi
r
ect rotor
position estim
at
ion.
Using
position se
nsor control
system
receives inform
ation about the rotor
position
with a
relat
i
vely low latency due
to the spee
d of CPU. In this case
the error
of rotor position estim
at
i
on depe
nds on the accuracy of
position
sens
or calibration.
When se
nsorless c
ontrol is use
d
the
accuracy of l
o
ok-up ta
ble,
the large am
ount
of
com
put
at
i
on
r
e
qui
red
t
o
det
e
rm
i
n
e t
h
e val
u
e
of
θ
a
n
d t
h
e accuracy
of
measur
em
ent equi
pm
ent impact
on
ro
t
o
r
p
o
sitio
n
esti
m
a
t
i
o
n
accu
r
acy.
All th
is is th
e cau
se
of a tim
e-v
a
ryin
g
error
o
f
estimatin
g
ro
t
o
r
po
sitio
n,
and
as a re
sul
t
- cha
n
gi
n
g
t
h
e act
ual
a
ngl
es of
o
p
e
n
i
n
g
and
cl
osi
ng t
h
e sem
i
cond
u
c
t
o
r s
w
i
t
c
hes
of t
h
e
co
nv
erter.
In
t
h
is case, th
e m
a
g
n
e
tic system is in
a
st
ate of artificial asy
mmetry,
as d
u
ri
n
g
dwel
l
a
ngl
e
pha
se
flux
lin
k
a
g
e
t
a
k
e
on
d
i
fferen
t
v
a
lu
es. Th
us, th
e m
a
in
d
i
fferen
ce
o
f
SRM sen
s
orless co
n
t
ro
l
will b
e
th
e
artificial p
h
a
se ex
citatio
n unbalan
ce wh
ich
serv
es as
a source of ad
d
ition
a
l n
o
i
se and
v
i
bratio
n [8
].
It should als
o
t
a
ke int
o
acc
ount the e
ffect
of th
e sen
s
ing
curren
t
pu
lses app
lied
to estim
a
t
e th
e ro
tor
p
o
s
ition
.
Th
ese p
u
l
ses are typ
i
cally o
f
a h
i
g
h
e
r frequ
e
n
c
y th
an
th
e pu
lses o
f
th
e
p
o
wer su
pp
ly an
d
affect on
une
ve
nne
ss el
e
c
t
r
om
agnet
i
c
t
o
r
q
ue, si
nce t
h
e sensi
n
g
cu
rr
ent
fl
ows
t
h
ro
ug
h t
h
e
pha
se,
rel
a
t
i
v
e t
o
w
h
i
c
h t
h
e
rot
o
r i
s
i
n
t
h
e
area o
f
t
h
e
ge
nerat
o
r
ope
rat
i
on m
ode.
Sen
s
i
ng c
u
r
r
ent
al
s
o
co
nt
ri
but
es t
o
n
o
i
s
e i
n
t
h
e
up
pe
r
fre
que
ncy
ran
g
e
o
f
heari
n
g d
u
e
t
o
m
a
gnet
o
st
ri
ct
i
on.
Based
on
t
h
e ab
ov
e
p
r
ov
isions, it is p
o
s
sib
l
e to
assu
m
e
th
a
t
th
e no
ise level o
f
sen
s
o
r
less con
t
ro
lled
switch
e
d
relu
ct
an
ce driv
e will b
e
h
i
gh
er th
an wh
en
u
s
ing
the p
h
y
sical ro
tor p
o
sitio
n
sen
s
o
r
. In
th
is case, it is
pos
si
bl
e t
o
che
c
k t
h
e val
i
d
i
t
y
of t
h
i
s
co
ncl
u
s
i
on o
n
l
y
be a p
h
y
s
i
cal
m
odel
as a
m
eans of com
put
er t
ech
nol
ogy
do
n
o
t
al
l
o
w i
n
cl
udi
n
g
al
l
cau
ses of
SR
M
no
i
s
e i
n
t
h
e cal
cu
l
a
t
i
on o
f
t
h
e da
t
a
. It
sh
oul
d be
bo
rne i
n
m
i
nd t
h
at
an a
d
equate as
sessm
ent of the im
pact of se
nsorless c
o
n
t
rol o
n
electric
no
ise in
co
m
p
ariso
n
with
trad
i
tio
n
a
l
syste
m
s u
s
in
g
ro
t
o
r
po
sitio
n sen
s
o
r
can
b
e
g
i
v
e
n und
er th
e
fo
llo
wi
n
g
cond
itio
n
s
:
•
as an object of
study s
h
oul
d s
e
rve
the sa
me
S
R
M
f
e
d
f
r
o
m
t
h
e
s
a
me
p
o
w
e
r
co
n
v
e
r
t
e
r
;
•
measu
r
em
en
t sh
ou
ld b
e
p
e
rform
e
d
with
th
e
sam
e
in
stru
m
e
n
t
s in con
s
tan
t
en
v
i
ron
m
en
tal co
nd
itio
ns;
•
SR
M
sho
u
l
d
n
o
t
be m
oved
b
e
t
w
een m
easur
em
ent
s
, unac
c
e
p
t
a
bl
e sepa
rat
i
on
of SR
M
st
r
u
ct
u
r
al
el
em
ent
s
and
cha
n
ge of shaft
l
o
ad.
To m
easure t
h
e noi
se l
e
vel
SR
M
wi
t
h
t
o
o
t
h zone co
n
f
i
g
urat
i
o
n 6/
4 (
f
i
g
u
r
e 1
)
was ut
i
l
i
zed (SR
M
rat
e
d
p
o
we
r
5k
W,
rat
e
d
spee
d
1
5
0
0
r
p
m
)
. Ex
peri
m
e
nt
al
assem
b
l
y
i
s
gi
ve
n
on
Fi
g
u
r
e
2.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PED
S
I
S
SN
:
208
8-8
6
9
4
In
flu
en
ce
o
f
S
e
n
s
orless Con
t
ro
l on
t
h
e N
o
ise o
f
S
w
itch
e
d Relu
cta
n
c
e Mo
to
r
Drive
(
A
l
exan
d
er
Pet
r
us
hi
n)
43
5
Fi
gu
re
1.
SR
M
6/
4
m
a
gnet
i
c
f
i
el
d
Fi
gu
re
2.
Ex
pe
ri
m
e
nt
al
assem
b
l
y
f
o
r
n
o
i
s
e l
e
vel
m
easurem
ent
1 –
m
o
tor,
2
– encode
r, 3 – re
silient
m
ounts
,
4
– s
h
aft t
o
rque sens
or,
5
– c
ont
rolled loa
d
Po
wer
su
p
p
l
y
of e
n
gi
ne
was
per
f
o
r
m
e
d by
hal
f
-
b
ri
d
g
e c
o
nve
rt
er.
M
i
cro
c
ont
rol
l
e
r
A
V
R
At
m
e
ga16
was u
s
ed
as th
e CPU
o
f
the co
n
t
ro
l system
.
Th
e ro
t
o
r
p
o
s
ition
at com
p
arativ
e test
s d
e
term
in
ed
u
s
ing
an
abs
o
l
u
t
e
enc
o
der
or
sen
s
o
r
l
e
ss co
nt
r
o
l
al
go
ri
t
h
m
[9]
,
b
a
sed
on t
h
e
u
s
e of
sen
s
i
n
g
pul
ses
an
d a
m
a
rker
.
Am
pl
i
t
ude
o
f
sensi
n
g p
u
l
s
e cur
r
ent
i
n
unal
i
gne
d
r
o
t
o
r
position (180 electrical degrees
) is used as a
marker.
Swi
t
c
hi
n
g
of
pha
ses
occu
rs
aft
e
r m
a
rker,
i
.
e.
whe
n
c
o
nt
r
o
l
sy
st
em
regi
st
ers t
h
e c
u
rre
nt
am
pl
i
t
ude o
f
t
h
e
sensi
n
g p
u
l
s
e,
whi
c
h i
s
l
e
ss than t
h
e cu
rre
n
t
am
pl
it
ude of
t
h
e previ
ous
sensi
n
g p
u
l
s
e.
Pi
ct
ure o
f
t
h
e pha
se
cu
rren
t is shown
in Figure
3
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l.
6, No
. 3, Sep
t
em
b
e
r
2
015
:
43
3 – 438
43
6
Fi
gu
re
3.
Se
ns
orl
e
ss
co
nt
r
o
l
l
e
d SR
M
p
h
ase c
u
r
r
ent
Noi
s
e l
e
vel
m
easurem
ent
s
were car
ri
ed
out
f
o
r t
h
e
ori
e
nt
at
i
on m
e
t
h
o
d
at
fi
ve
poi
nt
s o
f
t
h
e
measuring s
u
rface (Fi
g
ure
4)
for thr
ee
different val
u
es
of rotor s
p
eed.
Fi
gu
re
4.
Locat
i
on
o
f
m
easuri
n
g
p
o
i
n
t
s
o
n
t
h
e m
easuri
ng
su
rface
1-
5
– m
easure
m
ent
poi
nt
s,
S
– m
easurem
ent
su
rface
,
d
–
di
st
ance f
r
o
m
no
i
s
e so
urce
(
1
m
e
t
e
r)
The c
o
m
put
at
i
o
n
o
f
a
v
era
g
e
noi
se
l
e
vel
bas
e
d
on
t
h
e e
q
uat
i
on
Evaluation Warning : The document was created with Spire.PDF for Python.
I
J
PED
S
I
S
SN
:
208
8-8
6
9
4
In
flu
en
ce
o
f
S
e
n
s
orless Con
t
ro
l on
t
h
e N
o
ise o
f
S
w
itch
e
d Relu
cta
n
c
e Mo
to
r
Drive
(
A
l
exan
d
er
Pet
r
us
hi
n)
43
7
n
i
i
K
L
n
L
1
1
(1
)
wh
ere
n
is a nu
m
b
er of m
easu
r
emen
t po
i
n
ts in
wh
ic
h
noise
level m
easurement wa
s,
K is
the prem
ises
con
s
t
a
nt
, c
o
m
put
ed
by
e
quat
i
on
.
v
S
A
A
S
K
1
4
1
lg
10
(2
)
whe
r
e
S is the
area
of selected m
easuri
n
g surface,
S
v
is t
h
e a
r
ea
of lim
iting s
u
rfaces
of the
pre
m
ises,
А
is
the equivale
nt
area
of ac
oustic abs
o
rption.
Equ
i
v
a
len
t
area of aco
u
s
tic ab
sorp
tio
n is
d
e
termin
ed
fro
m
eq
u
a
tion
v
s
S
a
A
(3
)
whe
r
e
a
s
is t
h
e
avera
g
e c
o
e
ffic
i
ent of ac
oustic
ab
so
rp
tion
is a fu
n
c
tion
o
f
pre
m
ises typ
e
.
C
h
an
gi
n
g
t
h
e r
o
t
o
r spee
d was
m
a
de di
rect
l
y
chan
ge t
h
e va
l
u
e of t
h
e s
u
p
p
l
y
vol
t
a
ge vi
a
l
a
borat
or
y
tran
sform
e
r, wh
ile th
e
v
a
lu
e
o
f
t
h
e
phase
c
u
r
r
ent
i
s
l
i
m
ited t
o
1
5
A. T
h
e
noi
se m
easurem
ent
s
were
m
a
de
usi
n
g a
s
o
u
n
d
l
e
vel
m
e
t
e
r C
e
n
t
er 3
2
0
e
x
hi
bi
t
e
d t
e
m
poral
c
h
aracteristics of "Slow",
the measurem
ent of rot
o
r
spee
d - using a
m
echanical tachom
eter. T
h
e
engi
ne
was
running at idle
during t
h
e m
easure
m
ents.
3.
R
E
SU
LTS AN
D ANA
LY
SIS
R
e
sul
t
s
o
f
noi
s
e
l
e
vel
m
easur
em
ent
are
gi
ve
n i
n
Ta
bl
e 1
.
Tab
l
e
1
Resu
lts of SRM
no
ise lev
e
l m
easu
r
e
m
en
t
P
a
r
a
me
t
e
r
S
e
r
i
a
l
n
u
mb
e
r
o
f
me
a
s
u
r
e
m
e
n
t
№
1
№
2
№
3
SRM shaft rotation speed (m
in
-1
)
800
1000
1500
SRM
contr
o
l using phy
sical r
o
tor
pos
ition sensor
Noise level in
m
e
asur
ing point 1(
dB
A)
57,
7
60,
1
64,
7
Noise level in
m
e
asur
ing point 2 (
d
B
A
)
57,
8
59,
4
65,
0
Noise level in
m
e
asur
ing point 3 (
d
B
A
)
57,
3
60,
4
65,
6
Noise level in
m
e
asur
ing point 4 (
d
B
A
)
57,
7
60,
7
64,
7
Noise level in
m
e
asur
ing point 5 (
d
B
A
)
57,
6
60,
9
64,
8
Average noise leve
l (dBA)
52,
49
55,
17
59,
83
SRM
sensor
less co
ntr
o
l
Noise level in
m
e
asur
ing point 1 (
d
B
A
)
59,
2
61,
9
66,
4
Noise level in
m
e
asur
ing point 2 (
d
B
A
)
58,
9
62,
1
66,
8
Noise level in
m
e
asur
ing point 3 (
d
B
A
)
58,
2
62,
7
66,
3
Noise level in
m
e
asur
ing point 4 (
d
B
A
)
59,
7
61,
6
66,
8
Noise level in
m
e
asur
ing point 5 (
d
B
A
)
58,
3
61,
1
67,
4
Average noise leve
l (dBA)
53,
73
56,
75
61,
61
As see
n
f
r
om
Tabl
e
1, t
h
e
n
o
i
s
e l
e
vel
o
f
s
e
ns
orl
e
ss
SR
M
i
s
hi
ghe
r i
n
c
o
m
p
ari
s
on
wi
t
h
SR
M
wi
t
h
encode
r. T
h
e
noise le
vel increases
with t
h
e SRM s
h
af
t
spee
d inc
r
eas
e. Conside
r
i
ng that an i
n
cre
a
se in
rotation freque
ncy increase
s
the inacc
ur
acy
of rot
o
r po
si
t
i
on
est
i
m
at
i
on,
we
conclude t
h
at the inacc
uracy of
o
f
ro
tor
p
o
s
ition
estim
at
io
n
con
t
ribu
tes to th
e in
crease in
no
i
s
e
4.
CO
NCL
USI
O
N
From
these theoretical findi
ngs
and re
sult
s of
noise m
e
asurem
ents can be
seen t
h
a
t
sensorless
cont
rol
i
s
carri
ed o
u
t
at
a hi
gher
noi
se l
e
vel
of SR
M
.
The
noi
se i
n
c
r
ease
d
by
1,
2
4
-
1
,
7
8
dB
A,
we can
assum
e
th
at th
e ch
aracteristics o
f
SR
M are
v
e
ry sen
s
itiv
e to
sm
al
l ch
ang
e
s i
n
ti
min
g
of t
h
e pulse v
o
ltag
e
.
Artificial
mag
n
e
tic asymmetry th
at arises as a
resu
lt of error i
n
estimatin
g
th
e ro
tor
p
o
s
ition
sho
u
l
d
recogn
ized
as m
a
in
reason
of n
o
i
se
lev
e
l
in
creasin
g
.
Mod
e
rn
ro
tor p
o
s
itio
n
sen
s
o
r
s
can prov
id
e accurate
measu
r
em
en
ts with
an
error
not exce
eding 0,25
o
-0
,1
25
o
. In this
case, accordi
n
g to prelim
in
ary estim
a
tes
of th
e authors at the
sen
s
o
r
less con
t
ro
l erro
r in d
e
t
e
rm
in
in
g
th
e
po
sitio
n of the ro
tor is
2
,
5
o
-3
o
.
B
a
sed
on
t
h
i
s
,
we ca
n c
o
n
c
l
u
de t
h
at
Evaluation Warning : The document was created with Spire.PDF for Python.
I
S
SN
:
2
088
-86
94
I
J
PED
S
Vo
l.
6, No
. 3, Sep
t
em
b
e
r
2
015
:
43
3 – 438
43
8
for co
m
p
etitiv
e adv
a
n
t
ag
e sen
s
orless switch
e
d rel
u
ctan
ce
driv
e it is
n
e
cessary to tak
e
measu
r
es t
o
i
m
p
r
o
v
e
the accuracy of the indirect
rotor
position esti
m
a
tion. Im
provem
e
nt of
sensorless c
ont
rol is possi
ble by
cho
o
si
ng
t
h
e
ra
t
i
onal
pa
ram
e
ters
of
sensi
n
g
pul
ses
,
inc
r
easi
n
g the acc
urac
y of
lookup tables, etc.
ACKNOWLE
DGE
M
ENTS
The w
o
r
k
i
s
sup
p
o
rt
e
d
by
The
M
i
ni
st
ry
of
education
and science
of Russia.
Agree
m
ent
№
1
4
.604
.2
1.0040
, i
d
en
tif
ier
RFMEFI
6041
4X
00
40
.
REFERE
NC
ES
[1]
Acarnley
P, Hi
ll R, Hooper C
.
Detection of
rotor pos
ition i
n
stepping and
switched relu
ctance m
o
tors by
monitoring of
cu
rrent wav
e
forms.
IEEE Trans. In
dus
trial El
ec
tronics
. 1985
; 3
;
215
–222.
[2]
Tchav
y
chalov
MV. Methods o
f
switched reluctan
ce electric machin
es
sensorless control.
Rostov State Transp
o
rt
University Studies
. 2011; 4; 86-8
8
.
[3]
Mvungi NH. Sensorless Commutation Cont
r
o
l of Switched
Reluctance M
o
tor.
World Academy of Scien
ce,
Engineering and
Technolog
y
. 20
07; 25; 325-330.
[4]
Petrushin A, Tchav
y
chalov M, Miros
hnichenko
E. The S
w
itche
d Reluctan
ce E
l
ectr
i
c M
achin
e with Cons
tructiv
e
Asy
m
etry
.
I
n
ter
national Journal
of Powe
r
Electr
onics and Drive
Systems
. 2015; 6
(
1): 86-91
[5]
Krishnan R. Switch
e
d reluc
t
an
c
e
m
o
tor drives: m
odeli
ng, simulation
,
analy
s
is
, design, and app
l
ications.
Magna
Physics Publishing
. 2001; 416.
[6]
Ghousia SF. The Impact Analy
s
is of
Dwell Angles on Current S
h
ape and
Torqu
e
in Switched R
e
luctance Motor
s
.
International Jo
urnal of
Power
Electronics and
Drive Systems
. 2
012; 2(2): 160-1
69.
[7]
Bösing M. Acoustic modeling
of elec
tric
al
drives. Noise a
nd vibrati
on s
yntesis based on force respose
superposition.
Dis
s
.
RWT
H
Aa
ch
en Univ
er
s
ity
. 2
013; 208.
[8]
Nasirian V, Dav
oudi A, Kabo
li
S, Ed
ring
ton CS. Ex
citation
shif
ting:
a g
e
ne
r
a
l
low-cost solution
for eliminating
ultra-
low-frequency
torqu
e
ripp
le
in
switched
r
e
luctance mach
ines.
IEEE Transactions on ma
gnetic
. 2013
; 4
9
;
5135-5149.
[9]
Tchav
y
chalov
,
MV. Complex algorithm of switc
hed relu
ctance
motor sernsorless control.
Science and education
.
Electronic resea
r
ch and
develop
m
ent publication
.
2012; 12. http:/
/technomag.edu
.ru/doc/496400
.h
tml
Evaluation Warning : The document was created with Spire.PDF for Python.