I
nte
rna
t
io
na
l J
o
urna
l o
f
P
o
w
er
E
lect
ro
nics
a
nd
Driv
e
Sy
s
t
e
m
(
I
J
P
E
DS
)
Vo
l.
9
,
No
.
3
,
Sep
tem
b
er
2
0
1
8
,
p
p
.
14
1
2
~
14
2
2
I
SS
N:
2
0
8
8
-
8
6
9
4
,
DOI
: 1
0
.
1
1
5
9
1
/
i
j
p
ed
s
.
v9
.i
3
.
pp
14
1
2
-
14
2
2
1412
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//ia
e
s
co
r
e.
co
m/jo
u
r
n
a
ls
/in
d
ex
.
p
h
p
/
I
JP
E
DS
An Adapt
iv
e Neu
ra
l Net
w
o
rk
Cont
ro
ller Ba
sed o
n P
SO
and
G
ra
dien
t
Desc
ent
Metho
d
for PMS
M
Speed
Driv
e
Z
a
i
neb F
rij
et
,
Ali
Z
ri
bi
,
M
o
ha
m
e
d Cht
o
uro
u
Co
n
tr
o
l
&
E
n
e
rg
y
M
a
n
a
g
e
m
e
n
t
L
a
b
(CEM
L
A
B),
Na
ti
o
n
a
l
S
c
h
o
o
l
o
f
En
g
in
e
e
rin
g
o
f
S
f
a
x
,
Un
iv
e
rsit
y
o
f
S
f
a
x
,
T
u
n
isia
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
No
v
24
,
2
0
1
7
R
ev
i
s
ed
Ma
y
16
,
2
0
1
8
A
cc
ep
ted
A
u
g
6
,
2
0
1
8
In
th
is
p
a
p
e
r,
b
a
se
d
o
n
t
h
e
c
o
m
b
in
a
ti
o
n
o
f
p
a
rti
c
le
sw
a
r
m
o
p
ti
m
iza
ti
o
n
(
P
S
O)
a
lg
o
rit
h
m
a
n
d
n
e
u
ra
l
n
e
tw
o
rk
(N
N),
a
n
e
w
a
d
a
p
ti
v
e
sp
e
e
d
c
o
n
tro
l
m
e
th
o
d
f
o
r
a
p
e
rm
a
n
e
n
t
m
a
g
n
e
t
s
y
n
c
h
ro
n
o
u
s
m
o
to
r
(
P
M
S
M
)
is
p
ro
p
o
se
d
.
F
i
rstly
,
P
S
O
a
lg
o
rit
h
m
is
a
d
o
p
ted
t
o
g
e
t
th
e
b
e
st
se
t
o
f
w
e
i
g
h
ts
o
f
n
e
u
ra
l
n
e
tw
o
rk
c
o
n
tro
ll
e
r
(NN
C)
f
o
r
a
c
c
e
lera
ti
n
g
th
e
c
o
n
v
e
rg
e
n
t
sp
e
e
d
a
n
d
p
re
v
e
n
ti
n
g
t
h
e
p
ro
b
lem
s
o
f
trap
p
in
g
in
lo
c
a
l
m
i
n
im
u
m
.
T
h
e
n
,
to
a
c
h
iev
e
h
ig
h
-
p
e
rf
o
r
m
a
n
c
e
sp
e
e
d
trac
k
in
g
d
e
sp
it
e
o
f
th
e
e
x
i
ste
n
c
e
o
f
v
a
r
y
in
g
p
a
ra
m
e
ters
in
th
e
c
o
n
tr
o
l
s
y
ste
m
,
g
ra
d
ien
t
d
e
sc
e
n
t
m
e
th
o
d
is
u
se
d
to
a
d
ju
st
t
h
e
NN
C
p
a
ra
m
e
ters
.
T
h
e
sta
b
il
it
y
o
f
th
e
p
ro
p
o
se
d
c
o
n
t
ro
ll
e
r
is
a
n
a
ly
z
e
d
a
n
d
g
u
a
ra
n
tee
d
f
ro
m
Ly
a
p
u
n
o
v
th
e
o
re
m
.
T
h
e
ro
b
u
stn
e
ss
a
n
d
g
o
o
d
d
y
n
a
m
ic
p
e
r
f
o
rm
a
n
c
e
o
f
th
e
p
ro
p
o
se
d
a
d
a
p
ti
v
e
n
e
u
ra
l
n
e
tw
o
rk
sp
e
e
d
c
o
n
tr
o
l
sc
h
e
m
e
a
re
v
e
ri
f
i
e
d
th
r
o
u
g
h
c
o
m
p
u
ter sim
u
latio
n
s.
K
ey
w
o
r
d
:
A
d
ap
tiv
e
n
e
u
r
al
n
et
w
o
r
k
co
n
tr
o
ller
Gr
ad
ien
t d
escen
t
P
ar
ticle
s
w
ar
m
o
p
ti
m
izatio
n
P
er
m
a
n
en
t
m
a
g
n
e
t s
y
n
c
h
r
o
n
o
u
s
m
o
to
r
Sp
ee
d
co
n
tr
o
ller
Co
p
y
rig
h
t
©
201
8
In
s
t
it
u
te o
f
A
d
v
a
n
c
e
d
E
n
g
i
n
e
e
rin
g
a
n
d
S
c
ien
c
e
.
All
rig
h
ts re
se
rv
e
d
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
A
li
Z
r
ib
i
,
C
o
n
tr
o
l &
E
n
er
g
y
Ma
n
a
g
e
m
e
n
t
L
ab
(
C
E
M
L
A
B
)
,
Natio
n
al
Sch
o
o
l o
f
E
n
g
i
n
ee
r
in
g
o
f
S
f
a
x
,
Un
i
v
er
s
it
y
o
f
S
f
a
x
,
T
u
n
is
ia
.
E
m
ail:
ali_
zr
ib
i@
y
a
h
o
o
.
f
r
1.
I
NT
RO
D
UCT
I
O
N
I
n
r
ec
en
t
y
ea
r
s
,
m
ec
h
an
i
s
m
s
d
r
iv
en
b
y
a
p
er
m
a
n
e
n
t
m
a
g
n
e
t
s
y
n
ch
r
o
n
o
u
s
m
o
to
r
(
P
MSM
)
h
av
e
b
ee
n
w
id
el
y
u
s
ed
th
a
n
k
s
to
it
s
ad
v
an
tag
eo
u
s
m
er
it
s
o
f
p
er
f
o
r
m
a
n
ce
s
,
co
s
t
a
n
d
r
eliab
ilit
y
.
Ho
w
e
v
er
,
b
y
r
ea
s
o
n
o
f
th
e
co
n
ti
n
u
o
u
s
v
ar
iatio
n
o
f
p
ar
a
m
eter
s
,
th
e
n
o
n
li
n
ea
r
it
y
o
f
s
y
s
te
m
a
n
d
th
e
in
ac
ce
s
s
ib
il
it
y
o
f
s
o
m
e
s
tates
a
n
d
o
u
tp
u
t
f
o
r
m
ea
s
u
r
e
m
e
n
ts
,
i
m
p
r
o
v
in
g
t
h
e
P
MSM
p
er
f
o
r
m
a
n
ce
s
co
n
tr
o
l
m
a
y
b
ec
o
m
e
ch
allen
g
in
g
[
1
-
5
]
.
T
o
s
u
r
p
ass
t
h
e
ab
o
v
e
p
r
o
b
le
m
s
,
m
an
y
ar
ti
f
icial
i
n
tel
lig
e
n
ce
t
ec
h
n
iq
u
es
h
a
v
e
b
ee
n
p
r
o
p
o
s
ed
.
Neu
r
al
n
et
w
o
r
k
(
NN)
,
d
u
e
to
its
s
i
m
p
le
co
n
ce
p
tio
n
an
d
ea
s
y
i
m
p
le
m
en
t
atio
n
,
h
a
s
b
ee
n
w
id
el
y
u
s
ed
in
i
n
d
u
s
tr
ial
d
r
iv
e
ap
p
licatio
n
s
o
f
P
MSM
[
6
,
7
]
.
R
ec
en
t
l
y
,
to
i
m
p
r
o
v
e
t
h
e
co
n
tr
o
l
p
r
ec
is
io
n
,
C
h
ao
et
al.
p
r
o
p
o
s
ed
a
n
e
w
s
tr
ate
g
y
o
f
s
p
ee
d
co
n
tr
o
l
f
o
r
P
MSM
w
h
er
e
a
co
m
b
i
n
atio
n
o
f
ad
ap
tiv
e
b
ac
k
p
r
o
p
ag
atio
n
n
e
u
r
al
n
et
w
o
r
k
(
B
P
NN)
an
d
P
I
D
w
a
s
ad
o
p
ted
[
8
]
.
T
h
e
c
o
n
v
er
g
e
n
ce
o
f
t
h
e
p
r
o
p
o
s
ed
co
n
tr
o
l
s
ch
e
m
e
w
as
p
r
o
v
ed
u
s
i
n
g
th
e
L
y
ap
u
n
o
v
Stab
ilit
y
An
al
y
s
i
s
.
I
n
o
r
d
er
t
o
i
m
p
r
o
v
e
th
e
r
o
b
u
s
tn
e
s
s
p
er
f
o
r
m
an
ce
o
f
a
P
MSM
co
n
tr
o
l,
a
n
en
h
an
ce
d
r
o
b
u
s
t
f
r
ac
tio
n
al
o
r
d
er
p
r
o
p
o
r
tio
n
al
-
p
lu
s
-
i
n
teg
r
al
(
E
R
FOP
I
)
co
n
tr
o
ller
is
p
r
esen
ted
[
9
]
.
T
h
e
p
r
o
p
o
s
ed
co
n
tr
o
l
la
w
is
ac
ted
o
n
a
f
r
ac
tio
n
a
l
o
r
d
er
i
m
p
le
m
en
t
f
u
n
ctio
n
(
FOI
F)
o
f
tr
ac
k
in
g
er
r
o
r
.
Firstl
y
,
to
g
e
t
t
h
e
p
ar
a
m
eter
s
o
f
t
h
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
,
d
if
f
er
en
t
tu
n
i
n
g
r
u
les
w
er
e
ad
o
p
ted
to
g
e
n
er
ate
th
e
E
R
FOP
I
p
ar
am
eter
s
.
Seco
n
d
l
y
,
to
tak
e
c
h
an
g
e
s
o
f
t
h
e
p
r
o
ce
s
s
p
ar
a
m
eter
s
i
n
to
co
n
s
id
er
atio
n
NN
w
as
u
s
ed
to
ad
j
u
s
t
th
e
co
n
tr
o
ller
p
ar
a
m
eter
s
.
Vik
as
et
al.
i
n
[
1
0
]
p
r
o
p
o
s
ed
a
NN
b
ased
PID
(
NNP
I
D)
lik
e
co
n
tr
o
ller
w
h
ic
h
is
t
u
n
ed
w
h
en
t
h
e
co
n
tr
o
ller
i
s
o
p
er
atin
g
i
n
a
n
o
n
l
in
e
m
o
d
e
f
o
r
h
ig
h
p
er
f
o
r
m
a
n
ce
P
MSM
p
o
s
itio
n
co
n
tr
o
l.
I
n
t
h
e
f
ir
s
t
h
a
n
d
,
th
e
g
ai
n
s
o
f
t
h
e
NNP
I
D
ar
e
in
itialized
ac
co
r
d
in
g
to
a
n
e
w
tr
ai
n
i
n
g
al
g
o
r
it
h
m
b
ased
o
n
t
h
e
least
s
q
u
ar
e
s
o
lu
tio
n
.
T
h
en
,
a
s
eq
u
en
tial
tr
ai
n
i
n
g
a
lg
o
r
it
h
m
i
s
u
s
ed
to
ad
j
u
s
t
o
n
l
in
e
th
e
co
n
tr
o
ller
p
ar
a
m
eter
s
.
B
y
e
s
ti
m
atin
g
d
is
t
u
r
b
an
ce
s
,
R
ef
er
e
n
ce
[
1
1
]
s
h
o
w
s
a
s
i
m
p
le
m
e
th
o
d
o
f
i
m
p
r
o
v
i
n
g
co
n
t
r
o
l
p
er
f
o
r
m
an
ce
f
o
r
th
e
s
p
ee
d
o
f
P
MSM
.
T
h
e
y
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2
0
8
8
-
8
694
A
n
A
d
a
p
tive
N
eu
r
a
l Netw
o
r
k
C
o
n
tr
o
ller
B
a
s
ed
o
n
P
S
O
a
n
d
Gra
d
ien
t D
escen
t Meth
o
d
fo
r
…
(
Za
in
eb
F
r
ijet
)
1413
in
tr
o
d
u
ce
d
a
co
n
tr
o
ller
b
u
ild
e
d
b
y
p
r
o
p
o
r
tio
n
al
f
ee
d
b
ac
k
co
n
tr
o
ller
an
d
f
ee
d
f
o
r
w
ar
d
co
m
p
en
s
atio
n
b
ased
o
n
a
r
ad
ial
b
asis
f
u
n
ctio
n
n
e
t
w
o
r
k
d
is
tu
r
b
an
ce
o
b
s
er
v
er
.
T
h
e
p
r
o
p
o
s
ed
c
o
n
tr
o
ller
h
as
s
h
o
w
n
g
o
o
d
p
r
o
p
er
ties
in
d
is
tu
r
b
an
ce
r
ej
ec
tio
n
an
d
i
n
p
ar
a
m
eter
v
ar
iatio
n
s
.
I
n
o
r
d
er
to
m
ee
t
h
ig
h
er
p
er
f
o
r
m
a
n
ce
s
p
ec
if
icatio
n
s
f
o
r
a
P
MSM
,
an
ad
ap
tiv
e
m
o
d
i
f
ie
d
r
ec
u
r
r
en
t
L
e
g
en
d
r
e
n
eu
r
al
n
et
w
o
r
k
co
n
tr
o
l
s
y
s
te
m
i
s
p
r
o
p
o
s
ed
in
[
1
2
]
.
I
t
co
n
s
is
ts
o
f
a
m
o
d
i
f
ied
r
ec
u
r
r
e
n
t
L
e
g
e
n
d
r
e
NN
co
n
tr
o
l
w
it
h
ad
ap
tatio
n
la
w
an
d
a
co
m
p
e
n
s
ated
co
n
tr
o
l
w
it
h
esti
m
atio
n
la
w
,
d
er
iv
ed
b
y
u
s
i
n
g
t
h
e
L
y
ap
u
n
o
v
s
tab
ilit
y
t
h
eo
r
e
m
.
Desp
ite
co
n
tr
o
l
m
et
h
o
d
s
b
ased
o
n
NN
h
av
e
a
g
o
o
d
r
o
b
u
s
tn
e
s
s
ag
a
in
s
t
p
ar
a
m
eter
v
ar
i
atio
n
s
an
d
u
n
k
n
o
w
n
ex
ter
n
al
d
i
s
tu
r
b
a
n
ce
s
,
th
e
y
ar
e
co
n
f
r
o
n
ted
b
y
m
an
y
d
i
f
f
ic
u
ltie
s
.
Fo
r
e
x
a
m
p
le,
NN
p
ar
a
m
eter
s
tr
ain
i
n
g
i
s
ti
m
e
-
co
n
s
u
m
in
g
a
n
d
ea
s
il
y
f
all
s
in
to
lo
ca
l
m
in
i
m
u
m
w
h
ic
h
m
a
k
i
n
g
th
e
m
en
t
ir
el
y
d
ep
en
d
en
t
o
n
in
itial
w
ei
g
h
ts
[
1
3
]
.
P
SO
alg
o
r
ith
m
,
as
a
n
e
w
g
lo
b
e
o
p
ti
m
iza
tio
n
al
g
o
r
ith
m
,
h
as
th
e
ab
il
it
y
to
o
v
er
co
m
e
t
h
ese
p
r
o
b
lem
s
b
y
ac
ce
ler
atin
g
th
e
r
ate
o
f
co
n
v
er
g
e
n
ce
a
n
d
p
r
ev
en
ti
n
g
w
ei
g
h
ts
tr
ap
p
in
g
in
to
lo
ca
l
o
p
ti
m
a
.
R
ec
en
t
l
y
,
v
ar
io
u
s
co
n
tr
o
ller
d
esi
g
n
m
e
th
o
d
s
b
ased
o
n
P
SO
an
d
ar
tif
icial
n
e
u
r
al
n
et
w
o
r
k
(
A
NN)
h
a
v
e
b
ee
n
d
ev
elo
p
ed
[
1
4
-
1
6
]
.
A
s
a
n
e
w
t
r
ain
in
g
m
et
h
o
d
f
o
r
f
ee
d
f
o
r
w
ar
d
n
eu
r
al
n
et
w
o
r
k
s
(
FNN
s
)
,
an
h
y
b
r
id
o
f
P
SO a
n
d
Gr
av
itatio
n
al
Sear
ch
A
l
g
o
r
ith
m
(
GS
A
)
is
p
r
o
p
o
s
ed
in
[
1
7
]
.
I
t
w
a
s
e
m
p
lo
y
ed
to
r
ed
u
ce
th
e
p
r
o
b
lem
s
o
f
tr
ap
p
in
g
i
n
lo
ca
l
m
in
i
m
a
a
n
d
t
o
ac
ce
ler
ate
th
e
co
n
v
er
g
e
n
ce
r
ate
o
f
th
e
lear
n
i
n
g
alg
o
r
it
h
m
s
.
I
n
[
1
8
]
,
Yag
h
i
n
i e
t
al.
h
av
e
co
m
b
i
n
ed
th
e
ab
ilit
y
o
f
m
eta
h
e
u
r
is
tic
s
an
d
g
r
ee
d
y
g
r
ad
ien
t
b
ased
alg
o
r
ith
m
s
to
o
b
tain
a
h
y
b
r
id
i
m
p
r
o
v
ed
o
p
p
o
s
itio
n
b
ased
P
S
O
an
d
a
B
P
alg
o
r
ith
m
w
it
h
t
h
e
m
o
m
en
tu
m
ter
m
.
E
f
f
ec
ti
v
en
ess
o
f
t
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
is
co
m
p
ar
ed
w
it
h
s
e
v
e
r
al
o
th
er
f
a
m
o
u
s
A
N
N
tr
ain
i
n
g
alg
o
r
it
h
m
s
o
n
t
h
e
v
ar
io
u
s
b
en
ch
m
ar
k
p
r
o
b
lem
s
.
T
h
e
p
u
r
p
o
s
e
o
f
th
is
s
t
u
d
y
i
s
t
o
d
ev
elo
p
an
ad
a
p
tiv
e
NN
co
n
tr
o
ller
to
im
p
r
o
v
e
P
MSM
s
p
ee
d
co
n
tr
o
l.
I
n
o
n
e
h
an
d
,
i
n
o
r
d
er
to
av
o
id
tap
p
in
g
i
n
lo
ca
l
m
in
i
m
u
m
an
d
r
ed
u
cin
g
t
h
e
tr
ai
n
in
g
t
i
m
e,
P
SO
alg
o
r
ith
m
i
s
ad
o
p
ted
f
o
r
s
elec
tin
g
t
h
e
o
p
tim
al
s
o
lu
tio
n
o
f
in
i
tial
w
ei
g
h
ts
o
f
t
h
e
n
e
u
r
al
n
et
w
o
r
k
co
n
tr
o
l
ler
(
NNC).
T
h
en
,
to
ad
ap
t
th
e
i
n
itiall
y
u
n
ce
r
tain
a
n
d
v
ar
y
i
n
g
p
ar
a
m
eter
s
in
th
e
co
n
tr
o
l
s
y
s
te
m
,
t
h
e
g
r
ad
ien
t
d
escen
t
m
et
h
o
d
i
s
u
s
ed
to
o
p
ti
m
ize
t
h
e
w
ei
g
h
ts
w
h
er
e
t
h
e
co
n
v
er
g
en
ce
o
f
t
h
e
NN
co
n
tr
o
ller
is
g
u
ar
an
te
ed
f
r
o
m
L
y
ap
u
n
o
v
th
eo
r
e
m
.
T
h
is
p
ap
er
is
o
r
g
a
n
ized
as
f
o
llo
w
s
.
T
h
e
P
MSM
m
o
d
el
is
d
escr
ib
ed
in
s
ec
tio
n
2
.
Sectio
n
3
p
r
esen
t
s
th
e
NN
s
p
ee
d
co
n
tr
o
ller
p
ar
am
eter
s
t
u
n
i
n
g
u
s
in
g
P
SO
a
lg
o
r
ith
m
an
d
g
r
ad
ien
t
d
escen
t
m
eth
o
d
.
I
n
s
ec
t
io
n
4
,
th
e
s
tab
ilit
y
an
a
l
y
s
is
m
et
h
o
d
is
d
ev
elo
p
ed
.
T
h
e
s
i
m
u
lati
o
n
r
esu
lt
s
ar
e
p
r
esen
ted
in
Sectio
n
5
an
d
th
e
co
n
clu
s
io
n
is
p
r
ese
n
ted
in
Sec
tio
n
6
.
2.
B
ASI
C
P
M
S
M
M
O
DE
L
T
h
e
m
ac
h
in
e
m
o
d
el
o
f
a
P
M
SM
ca
n
b
e
d
escr
ib
ed
in
th
e
r
o
to
r
r
o
tatin
g
r
e
f
er
en
ce
f
r
a
m
e
as
f
o
llo
w
s
[
1
9
]
:
d
d
s
d
s
r
q
q
q
q
s
q
q
r
d
d
r
dI
V
R
I
L
L
I
dt
dI
V
R
I
L
L
I
dt
(
1
)
w
h
er
e
V
d
an
d
V
q
ar
e
d
-
q
ax
is
s
tato
r
v
o
lta
g
es,
I
d
an
d
I
q
ar
e
d
-
q
a
x
is
s
tato
r
c
u
r
r
en
t
s
,
L
d
a
n
d
L
q
ar
e
d
-
q
ax
is
s
tato
r
in
d
u
ctan
ce
s
,
r
is
th
e
elec
tr
ical
r
o
to
r
s
p
ee
d
an
d
is
th
e
f
lu
x
li
n
k
ag
e.
T
h
e
elec
tr
ic
to
r
q
u
e
ca
n
b
e
ex
p
r
ess
ed
as:
e
d
q
d
q
q
T
p
L
L
I
I
I
(
2
)
an
d
th
e
eq
u
atio
n
o
f
t
h
e
m
o
to
r
d
y
n
a
m
ics i
s
:
r
e
L
r
d
J
T
T
f
dt
(
3
)
J
is
m
o
m
e
n
t
o
f
i
n
er
tia.
T
e
,
T
L
ar
e
elec
tr
o
m
a
g
n
etic
a
n
d
lo
a
d
to
r
q
u
es.
f
is
f
r
ictio
n
co
ef
f
i
cien
t.
p
i
s
n
u
m
b
er
o
f
p
o
le
p
air
s
.
Ass
u
m
e
th
at
L
d
=
L
q
in
a
n
o
n
-
s
a
lien
t p
o
le
m
ac
h
i
n
e
(
s
u
r
f
ac
e
m
o
u
n
ted
)
P
MSM
,
th
e
m
o
d
el
w
il
l b
e
ev
en
s
i
m
p
ler
as i
n
d
icat
ed
b
y
th
e
f
o
llo
w
i
n
g
eq
u
atio
n
s
[
2
0
]
:
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
694
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
,
Vo
l.
9
,
No
.
3
,
Sep
tem
b
er
2
0
1
8
:
14
12
–
14
2
2
1414
1
1
31
2
1.5
ds
q
r
q
d
qq
q
s
q
r
d
r
q
q
q
q
r
q
L
r
q
eq
dI
R
I
p
I
V
dt
L
L
dI
R
I
p
I
p
V
dt
L
L
L
d
pB
IT
dt
J
L
J
J
T
p
I
(
4
)
A
cc
o
r
d
in
g
to
t
h
e
s
y
s
te
m
o
f
eq
u
atio
n
s
(
4
)
,
th
e
s
p
ee
d
co
n
tr
o
l
is
ac
h
ie
v
ed
b
y
co
n
tr
o
llin
g
t
h
e
cu
r
r
en
t
o
f
t
w
o
ax
e
s
(
I
d
an
d
I
q
)
.
T
h
e
m
ac
h
in
e,
s
p
ee
d
f
ee
d
b
ac
k
,
s
p
ee
d
an
d
cu
r
r
en
t
co
n
tr
o
ller
s
,
an
d
in
v
er
ter
co
n
s
tit
u
te
th
e
P
MSM
d
r
iv
e
s
y
s
te
m
a
s
s
h
o
w
n
in
Fi
g
u
r
e
1
.
Fig
u
r
e
1
.
B
lo
ck
d
iag
r
a
m
o
f
P
MSM
d
r
iv
e
3.
CO
NT
RO
L
ST
R
AT
E
G
Y
3
.
1
.
Ada
ptiv
e
NN
Sp
ee
d Co
ntr
o
ll
er
T
h
e
p
r
o
p
o
s
ed
s
p
ee
d
co
n
tr
o
lle
r
f
o
r
th
e
P
MSM
i
s
b
u
ilt
b
ase
d
o
n
an
ad
ap
tiv
e
N
N.
I
n
o
r
d
er
to
av
o
id
tr
ap
p
in
g
i
n
a
lo
ca
l
m
i
n
i
m
u
m
an
d
ac
ce
ler
ate
tr
ain
i
n
g
co
n
v
er
g
en
ce
,
P
SO
al
g
o
r
it
h
m
is
ad
o
p
ted
f
ir
s
tl
y
to
s
elec
t
in
itial
w
ei
g
h
ts
.
T
h
e
g
o
al
o
f
P
SO
alg
o
r
ith
m
is
to
g
et
th
e
o
p
ti
m
al
s
et
o
f
w
ei
g
t
h
s
(
p
ar
ticles
’
p
o
s
itio
n
)
b
y
ad
j
u
s
tin
g
t
h
e
tr
aj
ec
to
r
ies
o
f
t
h
e
p
ar
ticle
b
ased
o
n
t
h
e
b
est
p
o
s
itio
n
s
.
T
h
en
,
to
ac
h
ie
v
e
h
ig
h
-
p
er
f
o
r
m
a
n
ce
s
p
ee
d
tr
ac
k
in
g
d
esp
ite
o
f
t
h
e
e
x
is
te
n
ce
v
ar
y
i
n
g
p
ar
a
m
eter
s
i
n
th
e
c
o
n
tr
o
l
s
y
s
te
m
,
g
r
ad
ien
t
d
esce
n
t
m
et
h
o
d
is
u
s
ed
to
ad
j
u
s
t th
e
n
et
w
o
r
k
p
ar
a
m
eter
s
.
T
h
e
p
r
o
p
o
s
ed
m
et
h
o
d
f
o
r
d
eter
m
in
a
tio
n
t
h
e
NN
s
p
ee
d
co
n
tr
o
ller
p
ar
am
eter
s
co
m
p
r
i
s
es t
h
e
f
o
llo
w
in
g
s
tep
s
.
S
tep
1
:
I
n
p
u
t th
e
o
b
j
ec
t v
alu
e
o
f
co
n
tr
o
lled
s
y
s
te
m
i
n
to
t
h
e
co
n
tr
o
ller
.
S
tep
2
:
Sear
ch
o
p
ti
m
al
NN
s
p
ee
d
co
n
tr
o
ller
p
ar
am
eter
s
b
y
P
SO a
lg
o
r
it
h
m
.
S
tep
3
:
Use th
e
o
b
tain
ed
co
n
tr
o
ller
to
co
n
tr
o
l th
e
P
MSM
.
S
tep
4
:
Feed
b
ac
k
th
e
o
u
tp
u
t o
f
th
e
P
MSM
.
S
tep
5
:
A
d
j
u
s
t p
ar
am
eter
s
o
f
NN
s
p
ee
d
co
n
tr
o
ller
b
y
t
h
e
g
r
ad
ien
t d
escen
t
m
et
h
o
d
.
S
tep
6
:
I
f
t
h
e
s
p
ee
d
er
r
o
r
is
s
m
all
en
o
u
g
h
,
t
h
e
n
s
to
p
else g
o
t
o
Step
4
.
3
.
2
.
P
SO
f
o
r
f
ee
dfo
r
w
a
rd
NN
t
r
a
ini
ng
T
h
e
tr
ain
i
n
g
s
y
s
te
m
o
f
t
h
e
NN
s
p
ee
d
co
n
tr
o
ller
i
s
d
eter
m
i
n
ed
as
d
escr
ib
ed
in
Fi
g
u
r
e
2
.
I
n
t
h
i
s
s
tr
u
ct
u
r
e,
th
e
NN
s
p
ee
d
co
n
tr
o
ller
p
r
ec
e
d
es
th
e
P
MSM
m
o
d
el
an
d
r
ec
eiv
es
as
in
p
u
t
th
e
s
y
s
te
m
r
e
f
er
en
ce
w
it
h
th
e
p
ast
s
y
s
te
m
o
u
tp
u
ts
a
n
d
p
ast
in
p
u
ts
.
T
h
e
er
r
o
r
s
ig
n
al,
th
e
d
if
f
er
e
n
ce
b
et
w
ee
n
th
e
r
e
f
er
en
ce
s
i
g
n
al
an
d
t
h
e
s
y
s
te
m
o
u
tp
u
t,
w
il
l b
e
u
s
ed
b
y
P
SO a
lg
o
r
ith
m
to
tr
ain
th
e
n
et
w
o
r
k
.
C
u
rre
n
t
c
o
n
t
ro
l
l
e
r
PW
M
g
e
n
e
ra
t
o
r
I
n
v
e
rt
e
r
PM
SM
Sp
e
e
d
s
e
n
s
o
r
+
-
r
e
f
q
r
e
f
i
d
r
e
f
i
q
i
d
i
dr
e
f
v
qr
e
f
v
Pa
rk
t
ra
n
s
f
o
rm
a
t
i
o
n
A
V
B
V
C
V
G
ra
d
i
e
n
t
d
e
s
c
e
n
t
PSO
N
N
s
p
e
e
d
c
o
n
t
ro
l
l
e
r
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2
0
8
8
-
8
694
A
n
A
d
a
p
tive
N
eu
r
a
l Netw
o
r
k
C
o
n
tr
o
ller
B
a
s
ed
o
n
P
S
O
a
n
d
Gra
d
ien
t D
escen
t Meth
o
d
fo
r
…
(
Za
in
eb
F
r
ijet
)
1415
Fig
u
r
e
2
.
C
o
n
tr
o
l lo
o
p
im
p
le
m
en
ted
d
u
r
in
g
t
h
e
P
SO o
f
th
e
N
N
s
p
ee
d
co
n
tr
o
ller
T
h
e
tr
ain
in
g
p
r
o
ce
s
s
o
f
N
N
is
u
s
u
all
y
co
m
p
l
icate
d
a
n
d
h
i
g
h
d
i
m
en
s
io
n
al.
T
h
e
P
SO
as
an
ev
o
lu
tio
n
ar
y
al
g
o
r
ith
m
co
u
ld
b
e
u
s
ed
in
n
e
u
r
al
n
e
t
w
o
r
k
tr
ain
i
n
g
[
2
2
]
.
I
t
is
a
p
o
p
u
latio
n
(
s
w
ar
m
)
b
ased
o
p
tim
izatio
n
to
o
l
.
E
v
er
y
s
i
n
g
le
s
o
lu
tio
n
(
ca
lled
a
p
ar
ticle)
“f
lies
”
o
v
er
t
h
e
s
o
lu
tio
n
s
p
ac
e
in
s
ea
r
ch
f
o
r
t
h
e
o
p
tim
a
l
s
o
l
u
tio
n
.
I
n
ea
ch
g
e
n
er
atio
n
,
ea
ch
p
ar
ticle
i
s
u
p
d
ated
b
ased
o
n
2
s
p
ec
ial
p
ar
ticle
s
:
P
b
is
th
e
p
er
s
o
n
al
b
est
s
o
lu
t
io
n
o
f
ea
ch
p
ar
ticle,
an
d
P
g
is
th
e
g
lo
b
al
b
est
s
o
lu
tio
n
b
y
a
n
y
p
ar
ticle
in
t
h
e
s
w
ar
m
(
p
o
p
u
latio
n
)
.
T
h
e
p
er
f
o
r
m
a
n
ce
o
f
ea
c
h
p
ar
ticle
is
m
ea
s
u
r
ed
u
s
i
n
g
a
f
it
n
e
s
s
f
u
n
c
tio
n
th
at
v
ar
ie
s
d
ep
en
d
in
g
o
n
th
e
o
p
tim
izatio
n
p
r
o
b
le
m
[
2
3
,
2
4
]
.
T
h
ese
p
ar
ticles
v
elo
cit
y
an
d
p
o
s
itio
n
ar
e
u
p
d
ated
th
r
o
u
g
h
t
h
e
f
o
ll
o
w
in
g
eq
u
atio
n
s
:
1
1
2
2
(
1
)
(
)
(
)
(
)
(
)
(
)
i
d
i
d
i
d
i
d
g
d
i
d
v
k
w
v
k
C
r
p
k
x
k
C
r
p
k
x
k
(
5
)
(
1
)
(
)
(
1
)
1
id
id
id
x
k
x
k
v
k
d
D
(
6
)
W
h
er
e,
D
is
th
e
d
i
m
en
s
io
n
f
o
r
a
s
ea
r
ch
in
g
s
p
ac
e,
X
i
=(
x
i1
,x
i2
,…,x
iD
)
r
ep
r
esen
ts
t
h
e
p
o
s
itio
n
o
f
ea
c
h
p
ar
ticle,
V
i
=(
v
i1
,v
i2
,…,v
iD
)
is
th
e
v
elo
cit
y
o
f
th
e
t
h
p
ar
ticle,
P
ib
=(
p
i1
,p
i2
,…,p
iD
)
i
s
t
h
e
b
est p
o
s
itio
n
e
n
co
u
n
ter
ed
b
y
t
h
p
ar
ticle
,
P
g
s
h
o
w
s
th
e
b
est
p
o
s
i
tio
n
f
o
u
n
d
b
y
an
y
m
e
m
b
er
i
n
t
h
e
e
n
tire
s
w
ar
m
p
o
p
u
latio
n
P
g
=(
p
g1
,p
g2
,…,p
gD
)
,
k
is
i
ter
at
io
n
co
u
n
ter
;
C
1
,
C
2
ar
e
ac
ce
ler
atio
n
co
ef
f
icie
n
t
s
a
n
d
1
,
2
ar
e
t
w
o
s
i
m
i
lar
r
an
d
o
m
n
u
m
b
er
s
i
n
[
0
,
1
]
.
w
is
ca
lled
t
h
e
in
er
tia
f
ac
to
r
.
I
t
r
ed
u
ce
s
d
u
r
in
g
a
r
u
n
f
r
o
m
1
to
n
ea
r
0
in
ea
ch
g
en
er
atio
n
w
h
ic
h
f
ac
ilit
ate
s
a
b
alan
ce
i
n
th
e
e
x
p
lo
r
atio
n
an
d
ex
p
lo
itatio
n
o
f
th
e
s
ea
r
ch
s
p
ac
e,
it is
d
eter
m
i
n
e
d
as f
o
llo
w
s
:
m
a
x
m
in
m
a
x
m
a
x
ww
w
w
i
t
e
r
i
t
e
r
(
7
)
w
h
er
e
iter
max
i
s
th
e
m
a
x
i
m
u
m
n
u
m
b
er
o
f
iter
atio
n
s
,
an
d
iter
i
s
th
e
c
u
r
r
en
t n
u
m
b
er
o
f
iter
ati
o
n
.
P
SOB
P
A
lg
o
r
ith
m
I
n
th
is
s
tu
d
y
,
P
SO
i
s
u
s
ed
to
tr
ain
a
NN
to
o
b
tain
a
n
o
p
ti
m
u
m
n
et
w
o
r
k
m
o
d
el
a
n
d
to
i
m
p
r
o
v
e
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
NN.
D
u
r
in
g
t
h
e
tr
ain
i
n
g
p
h
a
s
e,
th
e
m
e
an
s
q
u
ar
ed
er
r
o
r
(
MSE
)
is
u
s
ed
to
ca
lcu
late
th
e
f
it
n
es
s
v
al
u
e.
1
1
N
i
i
M
SE
e
N
(
8
)
w
h
er
e
e
i
is
th
e
er
r
o
r
b
et
w
ee
n
d
esire
d
a
n
d
o
b
tain
ed
o
u
tp
u
ts
a
f
ter
p
r
ese
n
ti
n
g
t
h
e
it
h
d
at
u
m
to
t
h
e
n
et
w
o
r
k
,
an
d
N
is
t
h
e
n
u
m
b
er
o
f
d
ata
in
th
e
tr
ai
n
i
n
g
d
atase
t.
T
h
e
p
r
o
ce
d
u
r
e
o
f
th
e
P
SOB
P
ca
n
b
e
d
escr
ib
ed
as
f
o
llo
w
s
wh
er
e
a
p
ar
ticle
s
’
p
o
s
itio
n
r
ep
r
esen
t
s
t
h
e
v
alu
e
s
f
o
r
t
h
e
n
et
w
o
r
k
s
’
w
ei
g
h
ts
a
n
d
b
iases
.
S
tep
1
:
I
n
itialize
th
e
p
o
s
itio
n
s
an
d
v
elo
citie
s
o
f
a
g
r
o
u
p
o
f
p
a
r
ticles r
an
d
o
m
l
y
i
n
th
e
r
an
g
e
o
f
[
0
,
1
]
.
S
tep
2
:
E
v
alu
ate
ea
ch
i
n
itializ
ed
p
ar
ticles’
f
it
n
ess
v
alu
e
a
n
d
P
b
is
s
et
as
th
e
p
o
s
itio
n
s
o
f
th
e
cu
r
r
en
t
p
ar
ticles
,
w
h
ile
P
g
is
s
et
as t
h
e
b
est p
o
s
i
tio
n
o
f
t
h
e
in
i
tialized
p
ar
ticles.
-
+
ref
q
i
PSO
N
N
s
p
e
e
d
c
o
n
t
ro
l
l
e
r
PM
SM
m
o
d
e
l
T
D
L
T
D
L
m
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
694
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
,
Vo
l.
9
,
No
.
3
,
Sep
tem
b
er
2
0
1
8
:
14
12
–
14
2
2
1416
S
tep
3
:
C
h
an
g
e
th
e
v
elo
cit
y
o
f
th
e
p
ar
ticle
ac
co
r
d
in
g
to
(
5
)
an
d
u
p
d
ate
p
ar
ticle
p
o
s
itio
n
b
y
ad
d
in
g
th
e
ca
lcu
lated
v
elo
cit
y
v
alu
e
to
t
h
e
cu
r
r
en
t p
o
s
itio
n
v
al
u
e
ac
co
r
d
in
g
to
(
6
)
.
S
tep
4
:
C
o
m
p
ar
e
th
e
c
u
r
r
en
t
v
alu
e
o
f
t
h
e
MSE
w
i
th
t
h
e
p
ar
ticals’
p
r
ev
io
u
s
b
e
s
t
v
al
u
e
(
P
ib
)
.
I
f
th
e
c
u
r
r
en
t
f
it
n
es
s
v
al
u
e
is
le
s
s
,
t
h
en
u
p
d
atin
g
P
ib
.
S
tep
5
:
Fin
d
in
g
t
h
e
cu
r
r
en
t
g
l
o
b
al
m
i
n
i
m
u
m
o
f
MSE
an
d
c
o
m
p
ar
e
it
w
it
h
th
e
p
r
ev
io
u
s
g
l
o
b
al
m
i
n
i
m
u
m
(
P
g
)
.
I
f
th
e
c
u
r
r
en
t g
lo
b
al
m
i
n
i
m
u
m
is
b
etter
th
an
P
g
,
t
h
en
u
p
d
atin
g
P
g
.
S
tep
6
:
I
f
m
ax
i
m
u
m
n
u
m
b
er
o
f
iter
atio
n
s
is
r
ea
c
h
ed
o
r
th
e
f
it
n
es
s
v
a
lu
e
s
ar
e
m
e
t,
s
to
p
th
e
iter
atio
n
,
an
d
t
h
e
p
o
s
itio
n
s
o
f
p
ar
ticles ar
e
t
h
e
o
p
ti
m
al
b
est s
o
l
u
tio
n
.
Ot
h
er
w
i
s
e,
th
e
p
r
o
ce
s
s
is
r
ep
ea
ted
f
r
o
m
s
tep
3
.
S
tep
7
:
T
ak
in
g
th
e
w
ei
g
h
ts
a
n
d
b
iases
v
alu
e
s
w
h
ic
h
o
p
ti
m
ized
b
y
P
SO
a
s
t
h
e
in
i
tial
p
ar
am
eter
s
,
t
h
e
B
P
n
et
w
o
r
k
m
ak
e
s
au
to
n
o
m
o
u
s
le
ar
n
in
g
.
3
.
3
.
NN
Co
ntr
o
ller
P
a
ra
m
et
er
s
Adj
us
t
m
e
nt
T
h
e
NN
s
p
ee
d
c
o
n
tr
o
ller
w
e
i
g
h
t
s
,
o
p
ti
m
ized
b
y
th
e
P
SO,
w
il
l
b
e
u
s
ed
to
s
et
in
itial
w
ei
g
h
t
s
in
t
h
e
co
n
tr
o
ller
.
W
h
en
th
e
s
y
s
te
m
w
o
r
k
s
it
u
atio
n
c
h
an
g
e
s
d
eg
r
ad
atio
n
in
t
h
e
co
n
tr
o
l
p
er
f
o
r
m
a
n
ce
s
ca
n
b
e
ac
cr
u
ed
.
T
o
g
et
b
etter
co
n
tr
o
l
ef
f
ec
t
an
d
clo
s
e
q
u
ick
l
y
to
co
n
tr
o
l
o
b
ject
v
alu
e
s
,
at
ea
ch
co
n
tr
o
l
c
y
c
le,
w
ei
g
h
ts
m
u
s
t
b
e
ad
j
u
s
ted
ac
co
r
d
in
g
to
t
h
e
er
r
o
r
.
T
h
e
g
r
ad
ien
t
d
esce
n
t
m
et
h
o
d
ca
n
b
e
u
s
ed
to
m
in
i
m
iz
e
th
e
s
y
s
te
m
er
r
o
r
b
et
w
ee
n
th
e
o
u
tp
u
t o
f
t
h
e
P
MSM
an
d
th
e
d
esire
d
o
u
tp
u
t.
Ob
j
ec
tiv
e
f
u
n
ctio
n
is
d
ef
i
n
ed
as
f
o
llo
w
s
:
2
1
(
)
(
)
(
)
2
r
e
f
E
k
k
k
(
9
)
w
h
er
e
ω
is
th
e
ac
t
u
al
o
u
tp
u
t a
n
d
ω
ref
is
th
e
d
esire
d
o
u
tp
u
t.
T
h
e
w
ei
g
h
ts
u
p
d
atin
g
is
ca
lcu
lated
u
s
in
g
t
h
e
g
r
ad
ien
t
m
eth
o
d
,
th
e
e
x
p
r
ess
io
n
s
(
1
0
)
an
d
(
1
1
)
d
ef
in
e
th
e
f
o
r
m
u
la.
()
(
1
)
(
)
ij
ij
ij
Ek
w
k
w
k
w
(
1
0
)
()
(
1
)
(
)
jj
j
Ek
w
k
w
k
w
(
1
1
)
W
h
er
e
is
t
h
e
lear
n
in
g
r
ate
an
d
w
ij
ar
e
t
h
e
w
ei
g
h
t
s
f
r
o
m
th
e
in
p
u
t
la
y
er
to
t
h
e
h
id
d
en
la
y
e
r
an
d
w
j
ar
e
th
e
w
e
ig
h
ts
f
r
o
m
t
h
e
h
id
d
en
la
y
er
to
th
e
o
u
tp
u
t la
y
er
.
I
n
ea
c
h
s
a
m
p
li
n
g
p
er
io
d
,
to
ad
ap
t
v
ar
y
in
g
p
ar
a
m
eter
s
i
n
th
e
co
n
tr
o
l
s
y
s
te
m
a
ll
w
e
ig
h
ts
c
h
a
r
ac
ter
ized
th
e
NN
co
n
tr
o
ller
ar
e
u
p
d
ated
ac
co
r
d
in
g
to
th
e
er
r
o
r
s
o
f
clo
s
ed
lo
o
p
s
y
s
te
m
.
As
th
e
h
id
d
en
a
n
d
o
u
tp
u
t
n
eu
r
o
n
f
u
n
c
tio
n
s
w
er
e
d
ef
i
n
e
d
b
y
th
e
lo
g
i
s
tic
s
i
g
m
o
id
f
u
n
ctio
n
,
t
h
e
q
u
an
tit
ies i
n
(
1
0
)
an
d
(
1
1
)
ar
e
r
esp
ec
tiv
el
y
ca
lcu
la
ted
b
y
t
h
e
ex
p
r
ess
io
n
s
(
1
2
)
an
d
(
1
3
)
.
()
(
)
(
)
(
)
(
)
(
)
(
)
(
1
(
)
)
(
1
)
(
)
(
)
(
)
(
)
(
)
q
q
q
j
j
j
i
i
j
q
i
j
q
ik
E
k
E
k
k
k
E
k
i
k
i
k
w
O
O
O
w
k
k
i
k
w
k
i
k
(
1
2
)
()
(
)
(
)
(
)
(
)
(
)
(
)
(
1
(
)
)
(
)
(
)
(
)
(
)
(
)
q
q
q
j
j
q
j
q
ik
E
k
E
k
k
k
E
k
i
k
i
k
O
w
k
k
i
k
w
k
i
k
(
1
3
)
O
j
is
th
e
o
u
tp
u
t o
f
j
th
n
e
u
r
o
n
in
th
e
h
id
d
en
la
y
er
.
T
h
e
v
alu
e
o
f
t
h
e
p
ar
tial d
er
iv
a
tiv
e
q
i
ca
n
b
e
r
ep
lace
d
b
y
1
(
1
)
q
sign
i
.
4.
ST
AB
I
L
I
T
Y
ANA
L
YS
I
S
T
h
e
o
n
-
li
n
e
tr
ain
i
n
g
alg
o
r
it
h
m
s
f
o
r
th
e
s
p
ee
d
co
n
tr
o
ller
ca
lls
f
o
r
a
p
r
o
p
er
ch
o
ice
o
f
lear
n
in
g
r
ate
η
.
L
et
L
y
ap
u
n
o
v
f
u
n
ctio
n
b
e
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2
0
8
8
-
8
694
A
n
A
d
a
p
tive
N
eu
r
a
l Netw
o
r
k
C
o
n
tr
o
ller
B
a
s
ed
o
n
P
S
O
a
n
d
Gra
d
ien
t D
escen
t Meth
o
d
fo
r
…
(
Za
in
eb
F
r
ijet
)
1417
2
1
(
)
(
)
2
V
k
e
k
(
1
4
)
w
h
er
e
(
)
(
)
(
)
r
e
f
e
k
k
k
(
1
5
)
T
h
e
ch
an
g
e
o
f
L
y
ap
u
n
o
v
f
u
n
c
tio
n
is
22
11
(
)
(
1
)
(
)
(
(
1
)
(
)
)
(
)
(
)
(
)
22
V
k
V
k
V
k
e
k
e
k
e
k
e
k
e
k
(
1
6
)
T
h
e
er
r
o
r
d
i
f
f
er
en
ce
ca
n
b
e
r
ep
r
esen
ted
as
()
(
1
)
(
)
(
)
(
)
(
)
()
T
ek
e
k
e
k
e
k
e
k
W
k
Wk
(
1
7
)
()
Wk
is
th
e
n
et
w
o
r
k
s
w
eig
h
t
v
ec
to
r
s
.
T
h
e
ex
p
r
ess
io
n
o
f
()
Wk
is
g
i
v
en
b
y
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
(
)
q
qq
i
q
i
k
i
k
e
k
e
k
k
W
k
e
k
k
W
k
k
i
k
W
k
W
k
(
1
8
)
B
ased
o
n
E
q
u
atio
n
(
1
7
)
an
d
E
q
u
atio
n
(
1
8
)
w
e
ca
n
o
b
tain
ed
()
()
(
)
(
)
(
)
(
)
(
)
q
T
q
i
ik
ek
e
k
e
k
k
W
k
W
k
(
1
9
)
Su
b
s
ti
tu
t
in
g
t
h
e
ex
p
r
ess
io
n
o
f
()
ek
in
()
Vk
,
w
e
h
a
v
e
()
()
(
)
(
)
(
)
(
)
(
)
()
1
(
)
(
)
(
)
(
)
2
(
)
(
)
q
q
T
q
i
T
q
i
ik
ek
V
k
e
k
k
W
k
W
k
ik
ek
e
k
e
k
k
W
k
W
k
(
2
0
)
Sin
ce
(
)
(
)
()
(
)
(
)
q
i
e
k
k
k
W
k
w
k
,
s
u
b
s
tit
u
ti
n
g
th
i
s
i
n
E
q
u
atio
n
(
2
0
)
,
w
e
h
av
e,
2
2
2
2
2
4
(
)
(
)
(
)
(
)
1
(
)
(
)
(
(
)
)
(
)
(
(
)
)
(
)
(
)
2
(
)
(
)
qq
TT
q
q
q
q
ii
i
k
i
k
i
k
i
k
V
k
e
k
k
e
k
k
W
k
W
k
W
k
W
k
(
2
1
)
I
n
o
u
r
ca
s
e,
th
e
v
al
u
e
o
f
th
e
p
ar
tial
d
er
iv
ativ
e
q
i
q
i
is
r
ep
lace
d
b
y
1
(
1
)
q
sign
i
.
W
e
o
b
tain
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
694
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
,
Vo
l.
9
,
No
.
3
,
Sep
tem
b
er
2
0
1
8
:
14
12
–
14
2
2
1418
24
2
2
2
22
2
(
)
(
)
1
(
)
(
)
(
)
(
)
2
(
)
(
)
(
)
1
2
(
)
2
(
)
(
)
qq
qq
i
k
i
k
V
k
e
k
e
k
W
k
w
k
i
k
i
k
ek
W
k
W
k
(
2
2
)
A
cc
o
r
d
in
g
to
th
e
L
y
ap
u
n
o
v
T
h
eo
r
e
m
[
2
5
]
,
th
e
clo
s
ed
lo
o
p
s
y
s
te
m
is
s
tab
le
if
(
)
0
Vk
(
2
3
)
So
2
()
20
()
q
ik
Wk
(
2
4
)
Fin
all
y
,
w
h
en
w
e
d
e
f
in
e
t
h
e
m
atr
ix
n
o
r
m
2
by
2
m
a
x
(
)
T
M
M
M
(
2
5
)
th
e
th
eo
r
e
m
r
esu
lt is
es
tab
lis
h
ed
.
T
h
e
s
tab
ilit
y
co
n
d
it
io
n
(
)
0
Vk
is
s
atis
f
ied
o
n
l
y
if
2
m
a
x
m
a
x
2
()
()
()
2
w
h
e
r
e
()
(
)
(
)
qr
ik
k
Wk
and
k
kk
(
2
6
)
5.
SI
M
UL
AT
I
O
N
S RE
SU
L
T
S
T
o
v
er
if
y
t
h
e
ef
f
ec
ti
v
e
n
es
s
o
f
th
e
p
r
o
p
o
s
ed
h
ig
h
s
p
ee
d
P
MSM
co
n
tr
o
l a
lg
o
r
it
h
m
,
a
d
i
g
ital
s
i
m
u
lat
io
n
b
ased
o
n
t
h
e
Ma
t
lab
/Si
m
u
lin
k
s
o
f
t
w
ar
e
p
ac
k
a
g
e
h
as
b
ee
n
ca
r
r
ied
o
u
t.
T
h
e
P
MSM
c
h
ar
ac
ter
is
tics
ar
e
g
i
v
e
n
i
n
T
ab
le
1
.
T
ab
le
1
.
Ma
ch
in
e
P
ar
am
eter
s
P
a
r
a
me
t
e
r
V
a
l
u
e
R
a
t
e
d
v
o
l
t
a
g
e
2
2
0
/
3
8
0
V
S
t
a
t
o
r
r
e
si
st
a
n
c
e
(
R
s
)
1
.
4
Ω
I
n
d
u
c
t
a
n
c
e
(
L
d
=
L
q
)
6
.
6
mH
M
a
g
n
e
t
i
c
f
l
u
x
c
o
n
s
t
a
n
t
(
Φ
)
0
.
1
5
4
6
W
b
M
o
t
o
r
i
n
e
r
t
i
a
(
J
)
0
.
0
0
1
7
6
N
.
m.s
2
/
r
a
d
F
r
i
c
t
i
o
n
c
o
e
f
f
i
c
i
e
n
t
(
f
)
0
.
0
0
0
3
8
8
N
.
m
.
s/
r
a
d
P
o
l
e
p
a
i
r
s
n
u
m
b
e
r
s (
p
)
3
T
h
e
f
u
n
ct
io
n
o
f
P
SO
in
NN
s
p
ee
d
co
n
tr
o
ller
is
to
g
et
th
e
b
est
s
et
o
f
w
ei
g
h
t
s
(
p
ar
ticles’
p
o
s
itio
n
s
)
w
h
er
e
s
e
v
er
al
p
ar
ticles
ar
e
m
o
v
in
g
to
h
a
v
e
t
h
e
b
est
s
o
l
u
ti
o
n
.
T
h
e
NN
is
tr
ain
ed
u
s
i
n
g
th
e
tr
ain
in
g
s
y
s
te
m
d
escr
ib
ed
in
s
ec
tio
n
3
an
d
Fig
u
r
e
2
.
I
n
o
u
r
p
r
o
p
o
s
ed
m
et
h
o
d
,
th
e
P
SO
is
ad
o
p
ted
to
s
elec
t
in
itial
w
ei
g
h
ts
.
T
h
e
d
i
m
en
s
io
n
o
f
ea
ch
p
ar
ticle
i
s
t
h
e
to
tal
n
u
m
b
er
o
f
w
ei
g
h
ts
a
n
d
b
iases
.
Du
r
i
n
g
th
e
p
r
o
ce
s
s
o
f
w
ei
g
h
ts
i
n
itia
tio
n
b
y
P
SO,
n
u
m
b
er
s
o
f
p
o
p
u
latio
n
s
i
s
s
p
ec
if
ied
as 2
1
,
an
d
iter
atio
n
n
u
m
b
er
is
s
p
ec
if
ied
as 1
0
0
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2
0
8
8
-
8
694
A
n
A
d
a
p
tive
N
eu
r
a
l Netw
o
r
k
C
o
n
tr
o
ller
B
a
s
ed
o
n
P
S
O
a
n
d
Gra
d
ien
t D
escen
t Meth
o
d
fo
r
…
(
Za
in
eb
F
r
ijet
)
1419
I
n
t
h
e
tr
ai
n
in
g
p
h
ase,
th
e
s
etp
o
in
t
s
eq
u
en
ce
co
n
s
is
ti
n
g
o
f
p
u
ls
e
s
o
f
r
an
d
o
m
a
m
p
li
tu
d
e
i
n
th
e
r
a
n
g
e
[
0
;
3
0
0
r
ad
/s
ec
]
,
an
d
w
it
h
a
d
u
r
atio
n
o
f
2
0
s
a
m
p
li
n
g
p
er
io
d
s
.
T
h
e
m
ax
i
m
u
m
a
m
p
lit
u
d
es
ar
e
ch
o
s
en
s
o
as
to
f
u
ll
y
e
x
p
lo
r
e
th
e
d
esire
d
s
p
ee
d
a
m
p
litu
d
e
r
a
n
g
e.
T
o
illu
s
tr
ate
th
e
ad
v
a
n
ta
g
es
o
f
o
u
r
co
n
tr
o
ller
,
ex
p
er
im
en
ts
w
er
e
ex
ec
u
ted
u
s
i
n
g
t
w
o
s
p
ee
d
co
n
tr
o
ller
s
s
u
ch
a
s
:
t
h
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
an
d
a
co
n
v
e
n
tio
n
al
ad
ap
tiv
e
NN
co
n
tr
o
ll
er
w
it
h
a
r
an
d
o
m
in
itial
izatio
n
.
F
ig
u
r
e
3
s
h
o
w
s
P
MSM
m
o
d
el
r
esp
o
n
s
e
u
s
i
n
g
NN
C
o
p
ti
m
ized
b
y
P
SO
f
o
r
r
an
d
o
m
s
etp
o
in
t.
(
(
d
ash
-
d
o
tted
)
s
etp
o
in
t ω
ref
; (
s
o
lid
)
P
MSM
m
o
d
el
r
esp
o
n
s
e
ω
m
)
.
Fig
u
r
e
3
.
P
MSM
m
o
d
el
r
esp
o
n
s
e
u
s
i
n
g
NNC o
p
ti
m
ized
b
y
P
SO f
o
r
r
an
d
o
m
s
etp
o
in
t.
(
(
d
ash
-
d
o
tted
)
s
etp
o
in
t
ω
ref
; (
s
o
lid
)
P
MSM
m
o
d
el
r
esp
o
n
s
e
ω
m
)
Fig
u
r
e
4
s
h
o
w
s
th
e
s
p
ee
d
tr
ac
k
in
g
p
er
f
o
r
m
a
n
ce
test
o
f
b
o
th
th
e
co
n
tr
o
ller
s
.
I
t
is
clea
r
th
at
th
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
h
as
th
e
ab
ilit
y
to
k
ee
p
g
o
o
d
co
n
tr
o
l
p
er
f
o
r
m
an
ce
in
th
e
en
tire
p
r
o
ce
s
s
w
h
er
e
t
h
e
o
v
er
s
h
o
o
t,
s
ettli
n
g
ti
m
e
an
d
r
i
s
e
ti
m
e
ar
e
r
ed
u
ce
d
.
Fig
u
r
e
4
.
Sp
ee
d
tr
ac
k
in
g
r
esp
o
n
s
e
s
.
(
(
d
ash
-
d
o
tted
)
s
etp
o
in
t; (
s
o
lid
)
w
it
h
ad
ap
tiv
e
NN
co
n
tr
o
ller
; (
d
o
tted
)
w
it
h
th
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
)
Fig
u
r
e
5
s
h
o
w
s
t
h
e
s
tato
r
c
u
r
r
en
t
q
-
co
m
p
o
n
e
n
ts
.
E
ac
h
ch
a
n
g
e
i
n
th
e
s
p
ee
d
r
ef
er
e
n
ce
i
s
ac
co
m
p
a
n
ied
b
y
a
p
ea
k
i
n
t
h
eir
s
h
ap
es
.
I
t
ca
n
b
e
o
b
s
er
v
ed
t
h
at
at
e
v
er
y
ch
a
n
g
in
g
o
p
er
atin
g
co
n
d
itio
n
t
h
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
d
em
o
n
s
tr
ate
s
a
f
as
t c
u
r
r
en
t c
o
n
tr
o
l r
esp
o
n
s
e.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
2
0
8
8
-
8
694
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
,
Vo
l.
9
,
No
.
3
,
Sep
tem
b
er
2
0
1
8
:
14
12
–
14
2
2
1420
Fig
u
r
e
5
.
Stato
r
cu
r
r
en
t q
-
a
x
is
-
co
m
p
o
n
e
n
ts
.
(
(
s
o
lid
)
w
it
h
ad
ap
tiv
e
NN
co
n
tr
o
ller
; (
d
o
tted
)
w
it
h
t
h
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
)
T
h
e
p
u
r
p
o
s
e
o
f
a
s
ec
o
n
d
test
is
to
ev
alu
ate
t
h
e
d
is
t
u
r
b
an
c
es
r
ej
ec
tio
n
ef
f
ec
ti
v
e
n
ess
i
n
p
r
esen
ce
o
f
s
tep
lo
ad
to
r
q
u
e
d
is
tu
r
b
an
ce
.
T
h
e
s
y
s
te
m
i
s
test
ed
w
it
h
v
al
u
e
o
f
5
N
m
lo
ad
at
5
.
0
s
w
it
h
1
0
0
r
ad
/s
ec
s
p
ee
d
r
ef
er
en
ce
.
Fro
m
Fi
g
u
r
e
6
,
it
is
o
b
v
io
u
s
th
at
th
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
d
r
o
p
th
e
s
p
ee
d
f
r
o
m
1
0
0
r
ad
/s
ec
t
o
9
0
r
ad
/s
ec
w
i
th
s
h
o
r
t
s
ettl
in
g
ti
m
e.
W
h
ile,
ad
ap
tiv
e
NN
co
n
tr
o
ller
w
it
h
a
r
an
d
o
m
i
n
itializat
io
n
d
r
o
p
th
e
s
p
ee
d
f
r
o
m
1
0
0
r
ad
/s
ec
to
6
3
r
ad
/s
ec
w
it
h
lar
g
e
s
t
s
et
tli
n
g
ti
m
e.
T
h
e
p
r
o
p
o
s
ed
s
p
ee
d
c
o
n
tr
o
ller
p
r
o
d
u
ce
s
ig
n
i
f
ica
n
t
i
m
p
r
o
v
e
m
en
t
co
n
tr
o
l
p
er
f
o
r
m
an
ce
co
m
p
ar
e
to
t
h
e
ad
ap
tiv
e
NN
co
n
tr
o
ller
s
u
c
h
as
r
ed
u
ce
o
v
er
s
h
o
o
t,
s
ett
lin
g
ti
m
e
a
n
d
r
is
e
ti
m
e
s
p
ee
d
r
esp
o
n
s
e.
Fig
u
r
e
6
.
Dis
tu
r
b
an
ce
r
ej
ec
tio
n
r
esp
o
n
s
e
s
.
(
(
d
ash
-
d
o
tted
)
s
et
p
o
in
t; (
s
o
lid
)
w
it
h
ad
ap
tiv
e
N
N
co
n
tr
o
ller
;
(
d
o
tted
)
w
it
h
th
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
)
T
h
e
s
y
s
te
m
r
o
b
u
s
tn
e
s
s
a
g
ain
s
t
th
e
m
o
to
r
in
er
tia
ch
a
n
g
es
is
g
iv
e
n
in
F
ig
u
r
e
7
.
A
s
ca
n
b
e
s
ee
n
,
th
e
in
cr
e
m
e
n
t o
f
th
e
m
o
m
en
t o
f
i
n
er
tia
d
o
es n
o
t i
m
p
o
s
e
a
n
y
s
i
g
n
if
ica
n
t e
f
f
ec
t o
n
t
h
e
p
er
f
o
r
m
a
n
ce
o
f
th
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
tech
n
iq
u
e
b
u
t
o
n
l
y
af
f
ec
ts
th
e
r
is
e
ti
m
e.
F
u
r
th
er
m
o
r
e,
w
h
en
ca
r
ef
u
ll
y
s
t
u
d
y
F
ig
u
r
e
7
ac
co
r
d
in
g
to
th
e
s
ettli
n
g
ti
m
e
an
d
o
v
er
s
h
o
o
t,
th
e
b
est
p
er
f
o
r
m
a
n
ce
b
elo
n
g
s
to
t
h
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
.
T
h
is
m
ea
n
s
t
h
at
th
e
p
r
o
p
o
s
ed
c
o
n
tr
o
ller
is
in
s
en
s
i
tiv
e
to
p
ar
a
m
e
tr
ic
v
ar
iatio
n
s
an
d
a
r
o
b
u
s
t
tr
ac
k
i
n
g
p
er
f
o
r
m
an
ce
is
ac
h
ie
v
ed
i
n
p
r
esen
ce
o
f
th
e
u
n
ce
r
tain
p
ar
a
m
eter
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
n
t J
P
o
w
E
lec
&
Dr
i
S
y
s
t
I
SS
N:
2
0
8
8
-
8
694
A
n
A
d
a
p
tive
N
eu
r
a
l Netw
o
r
k
C
o
n
tr
o
ller
B
a
s
ed
o
n
P
S
O
a
n
d
Gra
d
ien
t D
escen
t Meth
o
d
fo
r
…
(
Za
in
eb
F
r
ijet
)
1421
Fig
u
r
e
7
.
R
esp
o
n
s
e
s
o
f
t
h
e
P
MSM
w
it
h
v
ar
iatio
n
1
0
0
% in
th
e
m
o
m
e
n
t o
f
i
n
er
tia
(
(
d
ash
-
d
o
tted
)
s
etp
o
in
t;
(
s
o
lid
)
w
it
h
ad
ap
tiv
e
NN
co
n
tr
o
ller
; (
d
o
tted
)
w
it
h
th
e
p
r
o
p
o
s
ed
co
n
tr
o
ller
)
6.
CO
NCLU
SI
O
N
I
n
th
i
s
p
ap
er
,
w
e
h
a
v
e
p
r
o
p
o
s
ed
a
n
o
v
el
ap
p
r
o
ac
h
o
f
s
p
ee
d
co
n
tr
o
l
f
o
r
P
MSM
co
m
b
in
i
n
g
P
SO
an
d
ad
ap
tiv
e
NN.
P
SO
al
g
o
r
ith
m
w
a
s
ad
o
p
ted
to
s
elec
t
i
n
itial
w
ei
g
h
ts
an
d
b
ias,
p
r
ese
n
ted
a
s
p
ar
ticles
p
o
s
itio
n
,
w
h
ic
h
w
er
e
u
p
d
ated
u
s
i
n
g
m
o
v
e
m
e
n
t
a
n
d
v
elo
cit
y
u
p
d
ate
eq
u
atio
n
s
to
av
o
id
t
h
e
w
ei
g
h
t
a
n
d
b
iases
b
ei
n
g
tr
ap
p
ed
in
lo
ca
l
m
in
i
m
a.
T
h
en
,
in
ea
ch
s
a
m
p
li
n
g
p
er
io
d
,
to
ad
a
p
t
th
e
in
itiall
y
u
n
ce
r
tain
an
d
v
ar
y
in
g
p
ar
am
eter
s
in
t
h
e
co
n
tr
o
l s
y
s
te
m
th
e
g
r
ad
ien
t d
esce
n
t
m
et
h
o
d
w
as
u
s
ed
to
u
p
d
ate
th
e
w
ei
g
h
t
s
o
f
co
n
tr
o
ller
.
RE
F
E
R
E
NC
E
S
[1
]
E.
Ce
ti
n
E
a
n
d
U.
Qg
u
z
,
"
A
h
y
b
rid
c
o
n
tro
ll
e
r
f
o
r
th
e
sp
e
e
d
c
o
n
tro
l
o
f
a
p
e
rm
a
n
e
n
t
m
a
g
n
e
t
s
y
n
c
h
ro
n
o
u
s
m
o
to
r
d
riv
e
,"
Co
n
t.
En
g
.
Pr
a
c
t.
,
v
o
l
.
1
6
,
n
o
.
3
,
p
p
.
2
6
0
-
2
7
0
,
2
0
0
8
.
[2
]
M
.
Ka
ra
b
a
c
a
k
a
n
d
H.
I.
Es
k
ik
u
rt,
"
De
sig
n
,
m
o
d
e
ll
in
g
a
n
d
sim
u
latio
n
o
f
a
n
e
w
n
o
n
li
n
e
a
r
a
n
d
f
u
ll
a
d
a
p
ti
v
e
b
a
c
k
ste
p
p
in
g
sp
e
e
d
trac
k
in
g
c
o
n
t
ro
ll
e
r
f
o
r
u
n
c
e
rtain
P
M
S
M
,"
A
p
p
li
e
d
M
a
th
e
m
a
ti
c
a
l
M
o
d
e
ll
i
n
g
,
v
o
l.
3
6
,
p
p
.
5
1
9
9
-
5
2
1
3
,
2
0
1
3
.
[3
]
T
.
J.
Re
n
a
n
d
T
.
C.
Ch
e
n
,
"
Ro
b
u
st
sp
e
e
d
-
c
o
n
tro
ll
e
d
in
d
u
c
ti
o
n
m
o
to
r
d
riv
e
b
a
s
e
d
o
n
re
c
u
rre
n
t
n
e
u
ra
l
n
e
tw
o
rk
,"
Ap
p
li
e
d
M
a
t
h
e
ma
ti
c
a
l
M
o
d
e
ll
i
n
g
,
v
o
l.
7
6
,
p
p
.
1
0
6
4
-
1
0
7
4
,
2
0
0
6
.
[4
]
M
.
M
o
u
jah
e
d
e
t
a
l.
,"
Ex
ten
d
e
d
K
a
lma
n
F
il
ter f
o
r
S
e
n
so
rles
s F
a
u
lt
T
o
lera
n
t
Co
n
tr
o
l
o
f
P
M
S
M
w
it
h
S
tato
r
Re
sist
a
n
c
e
Esti
m
a
ti
o
n
,"
IJ
PE
DS
,
v
o
l.
9
,
p
p
.
5
7
9
-
5
9
0
,
2
0
1
8
.
[5
]
C.
L
in
a
n
d
R.
W
u
,"
A
d
a
p
ti
v
e
Re
c
u
rre
n
t
Ne
tw
o
rk
Ne
t
w
o
rk
Un
c
e
rtain
ty
Ob
se
r
v
e
r
Ba
s
e
d
In
teg
ra
l
Ba
c
k
ste
p
p
in
g
Co
n
tr
o
l
f
o
r
a
P
M
S
M
Driv
e
S
y
ste
m
,"
IJ
PE
DS
,
v
o
l.
2
,
p
p
.
1
2
7
-
1
3
7
,
2
0
1
2
.
[6
]
S
.
Rich
a
,
e
t
a
l.
,
"
A
n
a
d
a
p
ti
v
e
P
I
D
li
k
e
c
o
n
tro
ll
e
r
u
sin
g
m
ix
lo
c
a
ll
y
r
e
c
u
rre
n
t
n
e
u
ra
l
n
e
tw
o
rk
f
o
r
ro
b
o
ti
c
m
a
n
ip
u
lat
o
r
w
it
h
v
a
riab
le p
a
y
lo
a
d
,"
IS
A
T
r
a
n
s
a
c
ti
o
n
s
,
v
o
l.
6
2
,
p
p
.
2
5
8
-
2
6
7
,
2
0
1
6
.
[7
]
N.
M
u
h
d
e
t
a
l.
,
"
S
p
e
e
d
Co
n
tro
l
o
f
P
e
rm
a
n
e
n
t
M
a
g
n
e
t
S
y
n
c
h
ro
n
o
u
s
M
o
t
o
r
Us
in
g
F
OC
Ne
u
ra
l
Ne
tw
o
rk
,"
th
e
se
ries
Lec
tu
re
No
tes
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
p
p
.
2
9
5
-
2
0
3
,
1
9
9
6
.
(
Bo
o
k
sty
le)
[8
]
N.
M
u
h
d
,
e
t
a
l.
,
"
S
p
e
e
d
C
o
n
tr
o
l
o
f
P
e
rm
a
n
e
n
t
M
a
g
n
e
t
S
y
n
c
h
ro
n
o
u
s
M
o
t
o
r
Us
in
g
F
OC
Ne
u
ra
l
Ne
tw
o
rk
,"
th
e
se
ries
Lec
tu
re
No
tes
in
El
e
c
tri
c
a
l
En
g
in
e
e
rin
g
,
p
p
.
2
9
5
-
2
0
3
,
1
9
9
6
.
(
B
o
o
k
sty
le)
[9
]
B.
Zh
a
n
g
a
n
d
Y.
P
i,
"
En
h
a
n
c
e
d
r
o
b
u
st
f
ra
c
ti
o
n
a
l
o
r
d
e
r
p
ro
p
o
rti
o
n
a
l
-
p
lu
s
-
i
n
teg
ra
l
c
o
n
tro
ll
e
r
b
a
se
d
o
n
n
e
u
ra
l
n
e
tw
o
rk
f
o
r
v
e
lo
c
it
y
c
o
n
tro
l
o
f
p
e
rm
a
n
e
n
t
m
a
g
n
e
t
s
y
n
c
h
ro
n
o
u
s m
o
to
r
,"
IS
A
T
ra
n
sa
c
ti
o
n
s
,
v
o
l.
5
2
,
p
p
.
5
1
0
-
5
1
6
,
2
0
1
3
.
[1
0
]
K.
Vik
a
s
,
e
t
a
l
.
,
"
AN
N
b
a
se
d
se
lf
tu
n
e
d
P
ID
li
k
e
a
d
a
p
ti
v
e
c
o
n
tr
o
ll
e
r
d
e
sig
n
f
o
r
h
ig
h
p
e
rf
o
rm
a
n
c
e
P
M
S
M
p
o
siti
o
n
c
o
n
tro
l
,"
Exp
e
rt S
y
ste
ms
wit
h
Ap
p
li
c
a
ti
o
n
s
,
v
o
l.
4
1
,
p
p
.
7
9
9
5
-
8
0
0
2
,
2
0
1
4
.
[1
1
]
L
.
Ju
a
n
,
e
t
a
l.
,
"
.
A
d
a
p
ti
v
e
sp
e
e
d
c
o
n
tro
l
o
f
a
P
M
S
M
se
rv
o
s
y
ste
m
u
sin
g
a
n
RBF
N
d
istu
rb
a
n
c
e
o
b
se
rv
e
r
,"
T
ra
n
sa
c
ti
o
n
s
o
f
th
e
I
n
stit
u
te
o
f
M
e
a
su
re
me
n
t
a
n
d
C
o
n
tr
o
l
,
v
o
l
.
3
4
,
p
p
.
6
1
5
-
6
2
1
,
2
0
1
1
.
[1
2
]
H.
L
.
Ch
ih
.
No
v
e
l
a
p
p
li
c
a
ti
o
n
o
f
c
o
n
ti
n
u
o
u
sly
v
a
riab
le
tran
s
m
is
sio
n
sy
ste
m
u
sin
g
c
o
m
p
o
site
re
c
u
rre
n
t
L
a
g
u
e
rre
o
rth
o
g
o
n
a
l
p
o
ly
n
o
m
ials
m
o
d
if
ied
P
S
O NN c
o
n
tr
o
l
s
y
ste
m
,"
IS
A
T
ra
n
sa
c
ti
o
n
s
,
v
o
l.
6
4
,
p
p
.
4
0
5
-
4
1
6
,
2
0
1
6
.
[1
3
]
G
.
M
a
rc
o
a
n
d
T
.
A
lb
e
rto
.
On
t
h
e
p
ro
b
lem
o
f
lo
c
a
l
m
in
ima
in
b
a
c
k
-
p
ro
p
a
g
a
ti
o
n
,"
IEE
E
T
ra
n
s.
P
a
tt
e
rn
An
a
l.
M
a
c
h
.
In
tell
,
v
o
l
.
1
4
,
p
p
.
7
6
-
8
6
,
1
9
9
2
.
[1
4
]
A
.
A
ji
th
,
e
t
a
l
.
,
"
S
w
a
r
m
in
telli
g
e
n
c
e
:
f
o
u
n
d
a
ti
o
n
s,
p
e
rsp
e
c
ti
v
e
s
a
n
d
a
p
p
li
c
a
ti
o
n
s
,
"
S
tu
d
ies
in
c
o
mp
u
t
a
ti
o
n
a
l
in
telli
g
e
n
c
e
,
v
o
l.
2
6
,
p
p
.
3
-
2
5
,
2
0
0
6
.
Evaluation Warning : The document was created with Spire.PDF for Python.