T
E
L
K
O
M
NIKA
T
elec
o
mm
un
ica
t
io
n Co
m
pu
t
i
ng
E
lect
ro
nics
a
nd
Co
ntr
o
l
Vo
l.
24
,
No
.
5
,
Octo
b
er
20
26
,
p
p
.
1
7
6
8
~
1
7
7
8
I
SS
N:
1
6
9
3
-
6
9
3
0
,
DOI
: 1
0
.
1
2
9
2
8
/
T
E
L
KOM
NI
K
A
.
v
24
i
5
.
27870
1768
J
o
ur
na
l ho
m
ep
a
g
e
:
h
ttp
:
//telko
mn
ika
.
u
a
d
.
a
c.
i
d
H
y
brid n
eura
l ne
tw
o
rk
-
a
ug
m
ent
ed
m
o
del
pred
ictive
co
ntrol for
diff
er
ential
-
d
riv
e
ro
bo
t
trac
k
ing
T
a
o
ufi
k
B
elk
e
bir¹,
H
icha
m
B
elk
eb
ir²,
Ana
s
s
M
a
ns
o
uri²
¹
L
R
S
I
,
F
a
c
u
l
t
y
o
f
S
c
i
e
n
c
e
s a
n
d
T
e
c
h
n
o
l
o
g
y
(
F
S
T
)
,
U
n
i
v
e
r
si
t
y
S
i
d
i
M
o
h
a
m
me
d
B
e
n
A
b
d
e
l
l
a
h
,
F
e
z
,
M
o
r
o
c
c
o
²
L
R
S
I
,
N
a
t
i
o
n
a
l
S
c
h
o
o
l
o
f
A
p
p
l
i
e
d
S
c
i
e
n
c
e
s (EN
S
A
)
,
S
i
d
i
M
o
h
a
me
d
B
e
n
A
b
d
e
l
l
a
h
U
n
i
v
e
r
si
t
y
,
F
e
z
,
M
o
r
o
c
c
o
Art
icle
I
nfo
AB
ST
RAC
T
A
r
ticle
his
to
r
y:
R
ec
eiv
ed
Feb
16
,
2
0
2
6
R
ev
i
s
ed
A
p
r
5
,
2
0
2
6
A
cc
ep
ted
Ma
y
25
,
2026
No
n
li
n
e
a
r
m
o
d
e
l
p
re
d
ictiv
e
c
o
n
t
ro
l
(NM
P
C)
is
w
id
e
ly
a
d
o
p
ted
f
o
r
m
o
b
il
e
ro
b
o
t
traje
c
to
ry
trac
k
in
g
,
y
e
t
th
e
in
tera
c
ti
o
n
b
e
tw
e
e
n
tran
sc
rip
ti
o
n
m
e
th
o
d
s
a
n
d
lea
rn
i
n
g
-
b
a
se
d
e
n
h
a
n
c
e
m
e
n
ts
re
m
a
in
s
u
n
st
u
d
ied
f
o
r
d
if
f
e
r
e
n
ti
a
l
-
d
riv
e
p
latf
o
rm
s.
T
h
is
p
a
p
e
r
p
re
se
n
ts
a
u
n
if
ied
Ca
sA
Di/In
terio
r
P
o
i
n
t
OP
T
im
ize
r
(IP
O
P
T
)
e
v
a
lu
a
ti
o
n
o
f
f
o
u
rth
-
o
r
d
e
r
Ru
n
g
e
-
Ku
tt
a
(RK
4
)
m
u
lt
ip
le
sh
o
o
ti
n
g
(M
S
)
a
n
d
He
rm
it
e
-
S
im
p
so
n
(HS)
d
irec
t
c
o
ll
o
c
a
ti
o
n
o
n
a
T
u
rtl
e
Bo
t3
B
u
rg
e
r
a
c
ro
ss
ten
sc
e
n
a
rio
s
sp
a
n
n
in
g
f
iv
e
traje
c
to
r
y
f
a
m
il
ies
,
m
o
d
e
l
m
is
m
a
tch
,
a
n
d
se
n
so
r
n
o
ise
.
A
-
a
m
b
ig
u
it
y
in
th
e
sta
n
d
a
rd
2
(
)
h
e
a
d
in
g
c
o
st
is
id
e
n
ti
f
ied
a
n
d
c
o
rre
c
ted
,
re
d
u
c
in
g
f
ig
u
re
-
e
ig
h
t
trac
k
in
g
e
rro
r
b
y
a
p
p
ro
x
im
a
tel
y
9
5
%
.
RK4
MS
a
c
h
iev
e
s
su
p
e
rio
r
b
a
se
li
n
e
a
c
c
u
ra
c
y
in
9
o
f
1
0
e
x
p
e
ri
m
e
n
ts
(
<
0
.
01
).
T
h
re
e
h
y
b
rid
n
e
u
ra
l
a
p
p
r
o
a
c
h
e
s
a
re
e
v
a
lu
a
ted
:
w
a
r
m
-
sta
rti
n
g
re
d
u
c
e
s
so
lv
e
ti
m
e
b
y
1
6
%
w
it
h
n
o
a
c
c
u
ra
c
y
lo
ss
;
a
d
a
p
ti
v
e
tran
sc
rip
ti
o
n
se
lec
ti
o
n
o
u
tp
e
rf
o
r
m
s
th
e
b
a
se
li
n
e
in
3
o
f
1
0
e
x
p
e
ri
m
e
n
ts;
a
n
d
n
e
u
ra
l
a
p
p
r
o
x
im
a
t
e
c
o
n
tr
o
l
a
c
h
ie
v
e
s
su
b
-
m
il
li
se
c
o
n
d
i
n
f
e
re
n
c
e
b
u
t
d
e
g
ra
d
e
s
o
n
u
n
se
e
n
traje
c
to
ries
(5
1
%
-
6
4
%
f
a
ll
b
a
c
k
r
a
te).
A
g
e
n
e
ra
li
z
a
ti
o
n
h
iera
rc
h
y
e
m
e
rg
e
s:
so
lv
e
r
-
in
-
th
e
-
lo
o
p
m
e
th
o
d
s
g
e
n
e
ra
li
z
e
w
it
h
o
u
t
m
e
a
su
ra
b
le
d
e
g
ra
d
a
ti
o
n
,
w
h
e
re
a
s
so
lv
e
r
-
re
p
la
c
e
m
e
n
t
m
e
th
o
d
s
re
q
u
ire
traje
c
to
ry
-
sp
e
c
i
f
ic
train
in
g
.
T
h
e
se
f
in
d
in
g
s
su
g
g
e
st
th
a
t
so
lv
e
r
in
v
o
lv
e
m
e
n
t
is
a
k
e
y
fa
c
to
r
in
f
lu
e
n
c
in
g
g
e
n
e
ra
li
z
a
ti
o
n
in
le
a
rn
in
g
-
a
u
g
m
e
n
ted
p
re
d
ictiv
e
c
o
n
tro
l
.
A
ll
re
su
lt
s a
re
sta
ti
stica
ll
y
v
a
li
d
a
ted
a
c
ro
ss
5
0
0
sim
u
latio
n
s.
K
ey
w
o
r
d
s
:
Dif
f
er
en
t
ial
-
d
r
i
v
e
r
o
b
o
t
Dir
ec
t c
o
llo
ca
tio
n
Mu
ltip
le
s
h
o
o
tin
g
Neu
r
al
n
et
w
o
r
k
No
n
li
n
ea
r
m
o
d
el
p
r
ed
ictiv
e
co
n
tr
o
l
T
r
a
j
ec
to
r
y
tr
ac
k
in
g
W
ar
m
-
s
tar
tin
g
T
h
is i
s
a
n
o
p
e
n
a
c
c
e
ss
a
rticle
u
n
d
e
r th
e
CC B
Y
-
SA
li
c
e
n
se
.
C
o
r
r
e
s
p
o
nd
ing
A
uth
o
r
:
T
ao
u
f
ik
B
elk
eb
ir
L
R
SI,
Facu
l
t
y
o
f
Scien
ce
s
an
d
T
ec
h
n
o
lo
g
y
(
FS
T
)
,
Un
iv
er
s
it
y
Sid
i M
o
h
a
m
m
ed
B
en
A
b
d
ellah
Fez
3
0
0
0
0
,
Mo
r
o
cc
o
E
m
ail:
b
el
k
eb
ir
.
tao
u
f
i
k
@
u
s
m
b
a.
ac
.
m
a
1.
I
NT
RO
D
UCT
I
O
N
Mo
d
el
p
r
ed
ictiv
e
co
n
tr
o
l
(
MP
C
)
is
w
id
el
y
r
ec
o
g
n
ized
a
s
a
l
ea
d
in
g
f
r
a
m
e
w
o
r
k
f
o
r
tr
aj
ec
to
r
y
tr
ac
k
i
n
g
in
m
o
b
ile
r
o
b
o
tics
,
o
w
i
n
g
to
its
s
y
s
te
m
atic
tr
ea
t
m
e
n
t
o
f
co
n
s
tr
ain
ts
a
n
d
ab
ilit
y
to
i
n
c
o
r
p
o
r
ate
n
o
n
lin
ea
r
d
y
n
a
m
ics
o
v
er
a
r
ec
ed
in
g
p
r
ed
ictio
n
h
o
r
izo
n
[
1
]
–
[
4
]
.
Fo
r
d
if
f
er
en
tia
l
-
d
r
i
v
e
p
latf
o
r
m
s
s
u
c
h
as
th
e
T
u
r
tleB
o
t3
B
u
r
g
er
w
id
el
y
u
s
ed
in
r
esear
ch
an
d
ed
u
ca
tio
n
th
e
n
o
n
h
o
lo
n
o
m
ic
k
in
e
m
at
ic
co
n
s
tr
ain
ts
an
d
lim
ited
ac
tu
at
io
n
b
an
d
w
id
t
h
(
max
=
0
.
22
m
/
s
)
d
e
m
an
d
ca
r
ef
u
l
co
n
tr
o
ller
d
esig
n
[
5
]
,
[
6
]
.
T
h
e
tr
an
s
cr
ip
tio
n
m
eth
o
d
u
s
ed
to
co
n
v
er
t
t
h
e
co
n
t
in
u
o
u
s
-
ti
m
e
o
p
tim
a
l
co
n
tr
o
l
p
r
o
b
le
m
(
OC
P
)
in
to
a
f
i
n
ite
-
d
i
m
e
n
s
io
n
al
n
o
n
li
n
ea
r
p
r
o
g
r
a
m
(
NL
P
)
s
ig
n
i
f
ica
n
tl
y
i
m
p
ac
t
s
b
o
th
s
o
lu
tio
n
ac
cu
r
ac
y
a
n
d
co
m
p
u
tatio
n
al
f
ea
s
ib
ilit
y
[
1
]
,
[
7
]
,
[
8
]
.
T
h
is
p
ap
er
ad
d
r
ess
es
th
r
ee
u
n
d
er
ex
p
lo
r
ed
asp
ec
ts
o
f
n
o
n
lin
ea
r
m
o
d
el
p
r
e
d
ictiv
e
co
n
tr
o
l
(
NM
P
C
)
f
o
r
d
if
f
er
e
n
tial
-
d
r
iv
e
r
o
b
o
ts
:
th
e
ch
o
ice
o
f
tr
an
s
cr
ip
tio
n
m
eth
o
d
,
th
e
in
teg
r
atio
n
o
f
lear
n
in
g
-
b
ased
en
h
a
n
ce
m
en
ts
,
an
d
an
o
f
te
n
-
o
v
e
r
lo
o
k
ed
h
ea
d
in
g
co
s
t a
m
b
i
g
u
i
t
y
.
W
h
ile
tr
an
s
cr
ip
tio
n
m
et
h
o
d
s
(
m
u
ltip
le
s
h
o
o
tin
g
(
MS)
,
d
ir
ec
t
co
llo
ca
tio
n
,
p
s
eu
d
o
s
p
ec
tr
al)
h
av
e
b
ee
n
ex
ten
s
i
v
el
y
b
en
c
h
m
ar
k
ed
f
o
r
au
to
n
o
m
o
u
s
v
e
h
icle
s
[
3
]
,
[
9
]
–
[
1
1
]
,
ae
r
o
s
p
ac
e
tr
a
j
ec
t
o
r
y
o
p
ti
m
izatio
n
[
7
]
,
[
8
]
,
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
n
eu
r
a
l n
etw
o
r
k
-
a
u
g
me
n
ted
mo
d
el
p
r
ed
ictive
co
n
tr
o
l
fo
r
…
(
Ta
o
u
fik
B
elke
b
ir
)
1769
an
d
au
to
n
o
m
o
u
s
m
o
b
ile
p
latf
o
r
m
s
[
5
]
,
[
1
2
]
,
th
eir
s
y
s
te
m
ati
c
co
m
p
ar
is
o
n
f
o
r
d
if
f
er
en
t
ial
-
d
r
iv
e
m
o
b
ile
r
o
b
o
ts
w
it
h
tig
h
t
ac
tu
a
tio
n
l
i
m
its
h
as
n
o
t
y
et
b
ee
n
s
y
s
te
m
atica
ll
y
s
t
u
d
ied
.
E
x
i
s
ti
n
g
i
m
p
le
m
e
n
tat
io
n
s
r
el
y
p
r
ed
o
m
in
a
n
tl
y
o
n
f
ix
ed
-
s
tep
d
is
cr
etiza
tio
n
w
it
h
o
u
t
r
ig
o
r
o
u
s
ev
al
u
atio
n
o
f
alter
n
ati
v
es
[
3
]
,
[
5
]
.
Fu
r
th
er
m
o
r
e,
lear
n
in
g
-
au
g
m
e
n
ted
MP
C
i
n
clu
d
in
g
n
eu
r
al
w
ar
m
-
s
tar
tin
g
[
1
3
]
,
n
eu
r
al
ap
p
r
o
x
i
m
ate
c
o
n
tr
o
l
w
ith
s
a
f
et
y
g
u
ar
a
n
tees
[
1
4
]
–
[
1
7
]
,
an
d
ad
a
p
tiv
e
al
g
o
r
ith
m
s
elec
tio
n
[
1
8
]
h
as b
ee
n
ex
p
lo
r
ed
p
r
i
m
ar
il
y
f
o
r
h
ig
h
-
d
i
m
e
n
s
io
n
al
s
y
s
te
m
s
.
Un
lik
e
t
h
e
q
u
ad
r
o
to
r
p
latf
o
r
m
s
s
t
u
d
ied
b
y
Sa
lz
m
a
n
n
et
a
l
.
[
1
9
]
an
d
T
o
r
r
en
te
et
a
l
.
[
2
0
]
,
o
r
th
e
r
ac
in
g
an
d
m
a
n
ip
u
la
tio
n
s
e
tu
p
s
o
f
[
2
1
]
–
[
2
4
]
,
d
if
f
er
en
tia
l
-
d
r
i
v
e
r
o
b
o
ts
o
p
er
ate
u
n
d
er
tig
h
ter
v
el
o
cit
y
an
d
ac
t
u
atio
n
co
n
s
tr
ain
ts
,
w
h
er
e
ev
e
n
s
m
all
tr
an
s
cr
ip
tio
n
d
if
f
er
e
n
ce
s
ca
n
a
f
f
ec
t
r
ea
l
-
ti
m
e
f
ea
s
ib
ilit
y
.
Mo
r
eo
v
er
,
th
e
h
ea
d
in
g
er
r
o
r
f
o
r
m
u
latio
n
a
n
o
f
te
n
-
o
v
er
lo
o
k
ed
d
esig
n
ch
o
ice
h
as
r
ec
eiv
ed
i
n
s
u
f
f
icie
n
t
atten
tio
n
d
esp
ite
its
r
o
le
i
n
n
o
n
h
o
lo
n
o
m
ic
tr
ac
k
i
n
g
s
tab
ili
t
y
:
t
h
e
w
id
el
y
-
u
s
ed
s
in
2
(
)
co
s
t
co
n
t
ain
s
a
-
a
m
b
i
g
u
it
y
t
h
at
ca
n
ca
u
s
e
s
ev
er
e
tr
ac
k
in
g
f
ail
u
r
e
o
n
h
i
g
h
-
c
u
r
v
a
t
u
r
e
tr
aj
ec
to
r
ies.
T
an
g
et
a
l
.
[
3
]
p
r
o
p
o
s
ed
g
eo
m
etr
ic
MP
C
f
o
r
w
h
ee
led
r
o
b
o
ts
w
it
h
c
u
r
v
at
u
r
e
-
a
w
ar
e
co
s
t
d
esig
n
,
b
u
t
d
id
n
o
t
co
m
p
ar
e
tr
an
s
cr
ip
tio
n
m
e
th
o
d
s
o
r
in
te
g
r
ate
lear
n
in
g
-
b
ased
a
u
g
m
e
n
tatio
n
.
Nas
ci
m
e
n
to
et
a
l
.
[
5
]
d
ev
elo
p
ed
m
o
d
i
f
ied
NM
P
C
f
o
r
n
o
n
h
o
lo
n
o
m
ic
m
o
b
ile
r
o
b
o
ts
,
th
o
u
g
h
o
n
l
y
E
u
ler
d
is
cr
etiza
t
io
n
w
as
e
v
al
u
ated
.
R
o
s
en
f
eld
er
et
a
l
.
[
2
5
]
em
p
l
o
y
ed
Ko
o
p
m
an
-
b
ased
p
r
ed
ictiv
e
co
n
tr
o
l
f
o
r
n
o
n
h
o
lo
n
o
m
i
c
s
y
s
te
m
s
,
w
it
h
o
u
t
ex
a
m
in
i
n
g
h
ea
d
in
g
co
s
t
f
o
r
m
u
latio
n
.
On
th
e
tr
an
s
cr
ip
tio
n
s
id
e,
Kelly
[
8
]
s
u
r
v
e
y
ed
d
ir
ec
t
co
llo
ca
tio
n
m
et
h
o
d
s
an
d
Ver
s
c
h
u
er
e
n
et
a
l
.
[
7
]
d
ev
elo
p
ed
th
e
ac
ad
o
s
f
r
a
m
e
w
o
r
k
,
y
e
t
n
e
ith
er
ad
d
r
ess
ed
d
i
f
f
er
en
tial
-
d
r
iv
e
-
s
p
ec
i
f
ic
tr
ad
e
-
o
f
f
s
.
Fo
r
lear
n
in
g
-
a
u
g
m
en
ted
MP
C
,
Her
tn
ec
k
et
a
l
.
[
1
4
]
estab
lis
h
ed
ap
p
r
o
x
i
m
ate
MP
C
w
i
th
s
a
f
et
y
g
u
ar
a
n
tees
f
o
r
g
e
n
er
al
n
o
n
li
n
ea
r
s
y
s
te
m
s
;
Kar
g
a
n
d
L
u
cia
[
1
5
]
d
em
o
n
s
tr
ated
e
f
f
ici
en
t
d
ee
p
-
lear
n
in
g
ap
p
r
o
x
im
a
tio
n
o
f
MP
C
la
w
s
;
an
d
Salz
m
an
n
et
a
l
.
[
1
9
]
ac
h
iev
ed
r
ea
l
-
ti
m
e
n
e
u
r
al
MP
C
f
o
r
q
u
ad
r
o
to
r
s
.
Ho
w
e
v
er
,
n
o
n
e
o
f
t
h
ese
w
o
r
k
s
co
m
p
ar
ed
m
u
ltip
le
h
y
b
r
id
s
t
r
ateg
ies
o
n
th
e
s
a
m
e
p
latf
o
r
m
o
r
ev
alu
ated
t
h
eir
g
en
er
aliza
tio
n
to
u
n
s
ee
n
tr
aj
ec
to
r
ies
a
g
ap
th
is
p
ap
er
ad
d
r
ess
es.
A
d
d
iti
o
n
all
y
,
R
o
s
o
li
a
an
d
B
o
r
r
elli
[
1
3
]
d
ev
elo
p
ed
lear
n
in
g
MP
C
f
o
r
iter
ativ
e
tas
k
s
,
C
a
u
li
g
i
et
a
l
.
[
2
6
]
ap
p
lied
s
u
p
e
r
v
is
ed
lear
n
in
g
f
o
r
m
i
x
ed
-
i
n
te
g
er
MP
C
,
Dr
g
o
ň
a
et
a
l
.
[
1
6
]
p
r
o
p
o
s
ed
d
if
f
er
e
n
tiab
le
p
r
ed
ictiv
e
co
n
tr
o
l,
an
d
He
w
i
n
g
et
a
l
.
[
2
]
s
u
r
v
e
y
ed
t
h
e
f
iel
d
b
r
o
ad
ly
,
all
tar
g
et
in
g
s
y
s
te
m
s
w
i
th
h
i
g
h
er
-
d
i
m
en
s
io
n
a
l
s
tate
s
p
ac
es.
T
h
is
w
o
r
k
m
ak
es
f
o
u
r
p
r
in
cip
al
co
n
tr
ib
u
tio
n
s
:
−
Hea
d
in
g
co
s
t
co
r
r
ec
tio
n
:
t
h
e
s
t
an
d
ar
d
s
in
2
(
)
h
ea
d
in
g
co
s
t
is
s
h
o
w
n
to
h
av
e
t
w
o
g
lo
b
al
m
i
n
i
m
a
(
=
0
an
d
=
)
.
A
co
r
r
ec
ted
(
1
−
co
s
)
+
s
in
2
f
o
r
m
u
l
atio
n
is
p
r
o
p
o
s
ed
an
d
p
r
o
v
en
to
h
av
e
a
u
n
iq
u
e
g
lo
b
al
m
in
i
m
u
m
(
T
h
eo
r
em
1
)
,
r
ed
u
cin
g
f
i
g
u
r
e
-
ei
g
h
t
tr
ac
k
i
n
g
er
r
o
r
b
y
ap
p
r
o
x
i
m
atel
y
9
5
%.
No
tab
ly
,
t
h
is
a
m
b
i
g
u
it
y
is
d
if
f
ic
u
lt
to
d
etec
t
in
p
r
ac
tic
e
as
it
m
an
i
f
e
s
ts
o
n
l
y
o
n
h
ig
h
-
c
u
r
v
at
u
r
e
tr
aj
ec
to
r
ies an
d
h
as n
o
t b
ee
n
p
r
ev
io
u
s
l
y
id
en
ti
f
ied
in
t
h
e
d
if
f
er
en
tial
-
d
r
iv
e
NM
P
C
liter
at
u
r
e.
−
S
y
s
te
m
a
tic
b
aseli
n
e
ev
al
u
atio
n
:
a
u
n
i
f
ied
C
a
s
A
Di/I
n
ter
io
r
P
o
in
t
OPT
im
izer
(
I
P
OPT
)
c
o
m
p
ar
is
o
n
o
f
R
K4
MS
v
er
s
u
s
Her
m
ite
-
Si
m
p
s
o
n
d
ir
ec
t
co
llo
ca
tio
n
ac
r
o
s
s
ten
s
ce
n
ar
io
s
,
es
tab
lis
h
in
g
R
K4
MS
s
u
p
er
io
r
it
y
i
n
9
o
f
1
0
ex
p
er
i
m
en
ts
(
<
0
.
01
).
−
T
h
r
ee
h
y
b
r
id
n
e
u
r
al
ap
p
r
o
ac
h
es:
(
i
)
n
eu
r
al
n
et
w
o
rk
(
NN
)
w
ar
m
-
s
tar
ti
n
g
w
it
h
1
6
%
s
o
lv
e
-
t
i
m
e
r
ed
u
ctio
n
;
(
ii
)
ad
ap
tiv
e
m
eth
o
d
s
elec
t
io
n
th
at
o
u
tp
er
f
o
r
m
s
t
h
e
b
aseli
n
e
i
n
3
o
f
1
0
ex
p
er
i
m
en
t
s
w
it
h
b
o
u
n
d
ed
r
eg
r
et;
an
d
(
iii
)
n
eu
r
al
ap
p
r
o
x
im
a
te
m
o
d
el
p
r
ed
ictiv
e
co
n
tr
o
l
(
NA
-
MP
C
)
w
i
th
s
u
b
-
m
ill
is
ec
o
n
d
in
f
er
en
ce
an
d
s
af
et
y
-
co
n
s
tr
ain
ed
f
a
llb
ac
k
.
−
Gen
er
aliza
tio
n
h
ier
ar
ch
y
:
a
c
o
n
tr
o
lled
ev
al
u
atio
n
r
ev
ea
li
n
g
th
at
s
o
lv
er
-
in
-
th
e
-
lo
o
p
m
e
th
o
d
s
g
en
er
alize
w
it
h
o
u
t
m
ea
s
u
r
ab
le
d
eg
r
ad
a
tio
n
,
w
h
ile
s
o
lv
er
-
r
ep
lace
m
e
n
t
m
e
th
o
d
s
s
h
o
w
s
tat
is
tical
l
y
s
i
g
n
if
ica
n
t
ac
cu
r
ac
y
lo
s
s
(
<
0
.
01
)
.
T
h
i
s
r
esu
lt
ad
d
r
ess
es
a
g
ap
in
th
e
liter
atu
r
e,
w
h
er
e
p
r
io
r
w
o
r
k
[
1
4
]
,
[
1
5
]
,
[
1
9
]
ev
alu
a
ted
h
y
b
r
id
m
e
th
o
d
s
i
n
d
iv
id
u
all
y
r
ath
er
t
h
an
co
m
p
a
r
ativ
el
y
ac
r
o
s
s
g
e
n
er
aliza
tio
n
r
eg
i
m
e
s
o
n
a
co
m
m
o
n
p
latf
o
r
m
.
T
h
e
r
em
ain
d
er
o
f
th
i
s
p
ap
er
is
o
r
g
an
ized
as
f
o
llo
w
s
.
Se
ctio
n
2
p
r
esen
ts
th
e
r
ese
ar
ch
m
et
h
o
d
,
in
cl
u
d
in
g
t
h
e
r
o
b
o
t
m
o
d
el,
NM
P
C
f
o
r
m
u
latio
n
,
h
ea
d
in
g
c
o
s
t
co
r
r
ec
tio
n
,
tr
an
s
cr
ip
tio
n
m
et
h
o
d
s
,
an
d
h
y
b
r
id
n
eu
r
al
ap
p
r
o
ac
h
es
.
S
ec
tio
n
3
r
ep
o
r
ts
th
e
ex
p
er
i
m
e
n
tal
r
es
u
lt
s
an
d
d
is
cu
s
s
io
n
,
a
n
d
s
ec
tio
n
4
co
n
clu
d
es
.
2.
M
E
T
H
O
D
Fig
u
r
e
1
p
r
esen
ts
th
e
o
v
er
a
ll
co
n
tr
o
l
ar
ch
itec
tu
r
e.
T
h
e
r
e
f
er
en
ce
tr
aj
ec
to
r
y
i
s
f
ed
to
t
h
e
NM
P
C
co
n
tr
o
ller
,
w
h
ich
f
o
r
m
u
late
s
an
OC
P
w
it
h
t
h
e
co
r
r
ec
ted
h
ea
d
in
g
co
s
t
(
T
h
eo
r
e
m
1
)
.
T
h
e
NL
P
is
s
o
l
v
ed
b
y
C
as
A
D
i/IP
OPT
an
d
th
e
r
esu
ltin
g
co
n
tr
o
ls
ar
e
ap
p
lied
t
o
t
h
e
T
u
r
tleB
o
t3
in
a
r
ec
ed
in
g
-
h
o
r
izo
n
lo
o
p
.
T
h
r
ee
h
y
b
r
id
n
e
u
r
al
au
g
m
en
tatio
n
s
ar
e
ev
alu
ated
:
M1
w
ar
m
-
s
tar
tin
g
th
e
N
L
P
w
it
h
a
lear
n
ed
in
itial
g
u
ess
,
M2
ad
ap
tiv
e
s
elec
tio
n
o
f
t
h
e
tr
a
n
s
cr
ip
tio
n
m
et
h
o
d
,
an
d
M3
d
ir
ec
t
p
o
licy
ap
p
r
o
x
i
m
a
tio
n
t
h
at
r
ep
lace
s
o
n
li
n
e
o
p
tim
izatio
n
w
i
th
n
eu
r
al
i
n
f
er
en
ce
(
w
ith
s
af
e
t
y
f
allb
ac
k
)
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
5
,
Octo
b
e
r
20
26
:
1
7
6
8
-
1778
1770
Fig
u
r
e
1
.
C
o
n
tr
o
l
ar
ch
itect
u
r
e
T
h
e
NM
P
C
p
ip
elin
e
(
to
p
r
o
w
)
s
o
lv
es
a
co
r
r
ec
ted
-
h
ea
d
in
g
OC
P
at
ea
ch
co
n
tr
o
l
s
tep
.
T
h
r
ee
h
y
b
r
id
n
eu
r
al
a
u
g
m
e
n
tatio
n
s
(
d
ash
e
d
b
o
x
)
a
r
e
ev
alu
ated
:
M1
p
r
o
v
id
es
a
lear
n
ed
w
ar
m
-
s
tar
t
to
ac
ce
ler
ate
co
n
v
er
g
e
n
ce
,
M2
s
elec
t
s
th
e
tr
an
s
cr
ip
tio
n
m
et
h
o
d
ad
ap
ti
v
el
y
,
a
n
d
M3
r
ep
lace
s
o
n
lin
e
o
p
tim
iza
tio
n
w
it
h
d
ir
ec
t p
o
licy
i
n
f
er
e
n
ce
u
n
d
er
s
af
et
y
-
co
n
s
tr
ai
n
ed
f
allb
ac
k
.
2
.
1
.
Dif
f
er
ent
ia
l
-
driv
e
ro
bo
t
m
o
del
T
h
e
T
u
r
tleB
o
t3
B
u
r
g
er
is
m
o
d
eled
as
a
u
n
ic
y
cle
-
t
y
p
e
d
i
f
f
er
en
tial
-
d
r
iv
e
r
o
b
o
t
[
5
]
,
[
6
]
w
i
th
s
tate
=
[
,
,
]
⊤
an
d
co
n
tr
o
l in
p
u
t
s
=
[
,
]
⊤
.
T
h
e
co
n
tin
u
o
u
s
-
ti
m
e
k
in
e
m
atic
s
ar
e:
̇
=
(
,
)
=
[
co
s
s
in
]
(
1
)
T
h
is
m
o
d
el
ca
p
tu
r
es
th
e
ess
e
n
tial
n
o
n
h
o
lo
n
o
m
ic
co
n
s
tr
ai
n
t:
th
e
r
o
b
o
t
ca
n
n
o
t
m
o
v
e
later
all
y
,
w
ith
all
m
o
tio
n
ar
is
in
g
f
r
o
m
t
h
e
in
ter
p
la
y
o
f
lin
ea
r
v
e
lo
cit
y
an
d
an
g
u
l
ar
v
elo
cit
y
.
T
h
e
p
h
y
s
ical
co
n
s
tr
ain
ts
ar
e:
∈
[
0
,
0
.
22
]
m
/s
(
f
o
r
w
ar
d
-
o
n
l
y
)
,
∈
[
−
2
.
84
,
2
.
84
]
r
ad
/s
,
w
h
ee
l
s
ep
ar
atio
n
=
0
.
160
m
,
an
d
w
h
ee
l
r
ad
iu
s
=
0
.
033
m
[
6
]
.
2
.
2
.
N
M
P
C
O
CP
A
t e
ac
h
s
a
m
p
li
n
g
ti
m
e
,
th
e
f
o
llo
w
i
n
g
OC
P
is
s
o
l
v
ed
o
v
er
th
e
p
r
ed
ictio
n
h
o
r
izo
n
[
1
]
,
[
7
]
:
min
(
⋅
)
,
(
⋅
)
∫
ℓ
0
(
,
,
)
+
(
(
)
)
(
2
)
s
u
b
j
ec
t
to
d
y
n
a
m
ics
(
1
)
,
in
iti
al
co
n
d
itio
n
(
0
)
=
,
an
d
co
n
tr
o
l
b
o
u
n
d
s
u
∈
.
T
h
e
o
b
j
ec
tiv
e
b
alan
ce
s
tr
ac
k
in
g
ac
c
u
r
ac
y
(
s
tag
e
co
s
t
ℓ
)
ag
ai
n
s
t
ter
m
i
n
al
d
ev
iatio
n
(
)
,
d
r
iv
in
g
th
e
r
o
b
o
t
to
f
o
llo
w
th
e
r
e
f
er
en
c
e
tr
aj
ec
to
r
y
w
h
ile
s
ati
s
f
y
i
n
g
p
h
y
s
ical
co
n
s
tr
ai
n
t
s
.
T
h
e
s
tag
e
co
s
t
w
i
th
t
h
e
co
r
r
ec
ted
h
ea
d
in
g
f
o
r
m
u
la
tio
n
(
s
ec
tio
n
2
.
3
)
is
:
ℓ
=
∥
−
∥
2
+
1
(
1
−
co
s
)
+
2
s
in
2
+
2
+
2
+
(
+
1
−
)
2
(
3
)
w
h
er
e
=
−
d
es
an
d
d
es
=
a
ta
n
2
(
̇
r
ef
,
̇
r
e
f
)
is
co
m
p
u
ted
an
al
y
ticall
y
f
r
o
m
tr
aj
ec
to
r
y
d
er
iv
ativ
es
[
3
]
.
T
h
e
f
ir
s
t
t
w
o
ter
m
s
p
en
a
lize
p
o
s
itio
n
a
n
d
h
ea
d
i
n
g
er
r
o
r
s
,
w
h
ile
th
e
r
e
m
ai
n
i
n
g
ter
m
s
p
r
o
m
o
te
s
m
o
o
t
h
,
e
n
er
g
y
-
ef
f
icien
t
co
n
tr
o
l
i
n
p
u
t
s
.
T
h
e
w
ei
g
h
ts
ar
e:
=
50
,
1
=
10
,
2
=
5
,
=
0
.
1
,
=
0
.
5
,
=
5
,
w
it
h
ter
m
i
n
al
w
ei
g
h
t
=
100
.
T
h
e
co
n
tr
o
ller
u
s
es
=
20
in
ter
v
al
s
,
=
2
.
0
s
h
o
r
izo
n
,
an
d
=
0
.
1
s
co
n
tr
o
l
p
er
io
d
(
1
0
Hz
)
,
s
o
lv
ed
v
ia
I
P
OPT
th
r
o
u
g
h
C
as
ADi
[
2
7
]
.
Ke
y
d
esi
g
n
ch
o
ices
in
cl
u
d
e
:
(
i)
p
o
s
itiv
e
-
o
n
l
y
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
n
eu
r
a
l n
etw
o
r
k
-
a
u
g
me
n
ted
mo
d
el
p
r
ed
ictive
co
n
tr
o
l
fo
r
…
(
Ta
o
u
fik
B
elke
b
ir
)
1771
v
elo
cit
y
to
p
r
ev
en
t
r
e
v
er
s
e
-
m
o
tio
n
ch
atter
i
n
g
[
2
5
]
;
(
ii)
tig
h
t
er
an
g
u
lar
r
ate
b
o
u
n
d
s
(
∈
[
−
1
.
5
,
1
.
5
]
r
ad
/s
)
f
o
r
s
m
o
o
th
er
tr
aj
ec
to
r
ies;
(
iii)
co
r
r
ec
ted
h
ea
d
in
g
co
s
t
(
s
ec
tio
n
2
.
3
)
;
an
d
(
iv
)
an
g
u
lar
r
ate
s
m
o
o
th
n
e
s
s
p
en
al
t
y
to
p
r
ev
en
t c
h
a
tter
in
g
.
2
.
3
.
H
ea
din
g
co
s
t
co
rr
ec
t
i
o
n:
re
s
o
lv
ing
t
he
-
a
m
big
uity
T
h
e
h
ea
d
in
g
co
s
t
d
eser
v
es
d
ed
icate
d
tr
ea
t
m
en
t
b
ec
au
s
e,
as
s
h
o
w
n
b
elo
w
,
a
s
u
b
tle
s
tr
u
ct
u
r
al
f
la
w
i
n
th
e
s
tan
d
ar
d
f
o
r
m
u
latio
n
ca
n
ca
u
s
e
s
e
v
er
e
tr
ac
k
in
g
f
ail
u
r
e
th
at
o
n
l
y
m
an
i
f
e
s
ts
o
n
h
i
g
h
-
c
u
r
v
atu
r
e
tr
aj
ec
to
r
ies
p
r
ec
is
el
y
t
h
e
co
n
d
itio
n
s
u
n
d
er
w
h
ic
h
d
i
f
f
er
e
n
tial
-
d
r
iv
e
r
o
b
o
ts
ar
e
m
o
s
t
c
h
alle
n
g
ed
.
T
h
e
h
ea
d
in
g
co
s
t
ℓ
s
t
d
=
s
in
2
(
)
is
co
m
m
o
n
l
y
u
s
ed
in
n
o
n
h
o
l
o
n
o
m
ic
MP
C
f
o
r
m
u
latio
n
s
[
3
]
,
[
5
]
d
u
e
to
its
s
m
o
o
th
n
e
s
s
a
n
d
i
m
p
licit a
n
g
le
w
r
ap
p
in
g
.
Ho
wev
er
,
th
is
f
o
r
m
u
latio
n
co
n
tai
n
s
a
s
tr
u
ctu
r
al
a
m
b
i
g
u
i
t
y
.
−
L
e
m
m
a
1
(
-
pe
r
io
d
icity
)
T
h
e
f
u
n
ctio
n
s
in
2
(
)
s
atis
f
ies
s
in
2
(
)
=
s
in
2
(
+
)
,
h
av
i
n
g
t
w
o
g
lo
b
al
m
in
i
m
a
p
er
2
-
p
er
io
d
:
at
=
0
(
co
r
r
ec
t
h
ea
d
in
g
)
an
d
=
(
b
ac
k
w
ar
d
h
ea
d
in
g
)
.
On
h
ig
h
-
c
u
r
v
at
u
r
e
tr
aj
ec
to
r
ies
(
e.
g
.
,
f
ig
u
r
e
-
eig
h
t
cr
o
s
s
o
v
er
p
o
in
ts
)
,
th
e
N
L
P
s
o
lv
er
ca
n
co
n
v
er
g
e
to
d
es
+
,
c
au
s
i
n
g
th
e
r
o
b
o
t
to
t
r
ac
k
th
e
tr
aj
ec
to
r
y
w
h
ile
p
o
i
n
tin
g
b
ac
k
w
ar
d
an
u
n
ac
ce
p
tab
le
f
ai
lu
r
e
f
o
r
f
o
r
w
ar
d
-
o
n
l
y
d
if
f
er
e
n
tial
-
d
r
iv
e
p
lat
f
o
r
m
s
.
T
h
e
f
o
llo
w
i
n
g
co
r
r
ec
ted
h
ea
d
in
g
co
s
t is p
r
o
p
o
s
ed
:
ℓ
(
)
=
(
1
−
co
s
)
+
s
in
2
(
4
)
−
T
h
eo
r
em
1
(
u
n
iq
u
e
g
lo
b
al
m
in
i
m
u
m
)
T
h
e
h
ea
d
in
g
co
s
t
(
4
)
s
atis
f
ies
:
(
i)
ℓ
θ
(
Δθ
)
≥
0
w
ith
eq
u
ali
t
y
i
f
Δθ
=
0
;
(
ii)
ℓ
θ
h
as
a
u
n
iq
u
e
g
lo
b
al
m
i
n
i
m
u
m
at
Δθ
=
0
; (
iii)
ℓ
θ
(
Δθ
)
=
3
2
Δ
θ
2
+
O
(
Δ
θ
4
)
n
ea
r
th
e
m
i
n
i
m
u
m
.
P
r
o
o
f
.
Usi
n
g
s
in
2
=
1
−
co
s
2
:
ℓ
=
(
1
−
co
s
)
(
2
+
co
s
)
(
5
)
Sin
ce
co
s
∈
[
−
1
,
1
]
,
w
e
h
a
v
e
(
1
−
co
s
)
≥
0
an
d
(
2
+
co
s
)
≥
1
>
0
.
T
h
eir
p
r
o
d
u
ct
is
ze
r
o
i
f
co
s
=
1
,
i.e
.
,
=
0
.
T
h
e
d
er
iv
ativ
e
ℓ
′
=
s
in
(
1
+
2
co
s
)
y
ield
s
cr
itical
p
o
in
t
s
at
∈
{
0
,
±
,
±
2
/
3
}
in
[
−
,
]
.
E
v
alu
ati
n
g
:
ℓ
(
0
)
=
0
(
g
lo
b
al
m
i
n
i
m
u
m
)
,
ℓ
(
±
)
=
2
(
lo
ca
l
m
i
n
i
m
u
m
)
,
ℓ
(
±
2
/
3
)
=
9
/
4
(
lo
ca
l
m
ax
i
m
a)
.
T
h
e
s
ec
o
n
d
d
er
iv
ati
v
e
co
n
f
ir
m
s
ℓ
″
(
±
2
/
3
)
<
0
(
m
ax
i
m
a)
an
d
ℓ
″
(
0
)
=
3
>
0
(
m
in
i
m
u
m
)
.
T
h
e
c
o
r
r
e
c
t
e
d
f
o
r
m
u
l
a
ti
o
n
c
o
m
b
in
es
th
e
d
i
s
am
b
i
g
u
at
i
o
n
o
f
(
1
−
co
s
)
w
i
th
th
e
l
o
ca
l
s
h
a
p
in
g
o
f
s
in
2
,
y
ie
l
d
in
g
th
e
s
t
e
e
p
e
s
t
g
r
a
d
i
en
t
(
3
×
th
e
c
o
s
in
e
-
o
n
ly
v
a
r
i
an
t
)
n
e
a
r
t
h
e
m
in
im
u
m
w
h
i
l
e
r
em
a
in
in
g
s
m
o
o
th
a
n
d
2
-
p
e
r
i
o
d
ic
.
T
h
e
f
u
n
ct
i
o
n
ℓ
a
l
s
o
s
a
t
is
f
i
es
L
y
a
p
u
n
o
v
c
a
n
d
i
d
a
t
e
p
r
o
p
e
r
t
i
e
s
(
p
o
s
i
tiv
e
d
e
f
in
it
e
,
z
e
r
o
at
e
q
u
i
li
b
r
iu
m
)
,
m
ak
in
g
i
t
s
u
i
t
a
b
le
f
o
r
N
M
P
C
s
t
a
b
i
li
ty
an
a
ly
s
is
u
n
d
e
r
s
t
a
n
d
a
r
d
t
e
r
m
in
a
l
c
o
s
t
ass
u
m
p
t
i
o
n
s
[
1
]
,
[
2
]
.
2
.
4
.
T
ra
ns
cr
iptio
n
m
et
ho
d
s
T
w
o
tr
an
s
cr
ip
tio
n
m
et
h
o
d
s
w
it
h
m
atc
h
i
n
g
(
ℎ
4
)
lo
ca
l
tr
u
n
ca
t
io
n
er
r
o
r
ar
e
co
m
p
ar
ed
:
Her
m
ite
-
Si
m
p
s
o
n
d
ir
ec
t
co
llo
ca
tio
n
a
n
d
R
K4
MS
[
7
]
,
[
8
]
,
[
2
8
]
.
Her
m
ite
-
Si
m
p
s
o
n
d
ir
ec
t
co
ll
o
ca
tio
n
u
s
es
c
u
b
ic
Her
m
ite
i
n
ter
p
o
latio
n
w
it
h
Si
m
p
s
o
n
’
s
r
u
le
q
u
ad
r
atu
r
e
[
8
]
,
[
2
7
]
.
Fo
r
ea
ch
in
ter
v
al
[
,
+
1
]
w
it
h
s
tep
ℎ
=
/
,
th
e
m
id
p
o
in
t
s
ta
te
is
i
n
te
r
p
o
lated
as
x
=
1
2
(
x
+
x
+
1
)
+
ℎ
8
(
−
+
1
)
,
an
d
th
e
Si
m
p
s
o
n
q
u
ad
r
atu
r
e
co
n
s
tr
ain
t
is
x
+
1
=
x
+
ℎ
6
(
+
4
+
+
1
)
.
T
h
e
m
id
p
o
in
t
s
ta
tes
ar
e
ad
d
i
tio
n
al
d
ec
is
io
n
v
ar
iab
les,
y
ield
i
n
g
(
2
+
1
)
+
(
+
1
)
=
163
v
ar
iab
les
an
d
6
=
120
co
n
s
tr
ain
t
s
at
=
20
.
R
K4
MS
d
iv
id
es
th
e
h
o
r
izo
n
in
to
in
ter
v
a
ls
an
d
u
s
es
ex
p
lici
t
4
th
-
o
r
d
er
R
u
n
g
e
-
K
u
tta
i
n
te
g
r
ati
o
n
w
it
h
i
n
ea
ch
[
7
]
,
[
2
8
]
.
C
o
n
tin
u
it
y
i
s
en
f
o
r
ce
d
as
NL
P
co
n
s
tr
ai
n
t
s
(
s
h
o
o
ti
n
g
g
ap
s
)
.
T
h
is
y
ield
s
(
+
1
)
+
=
103
v
ar
iab
les
an
d
3
=
60
co
n
s
tr
ai
n
ts
a
3
7
%
s
m
aller
p
r
o
b
le
m
.
T
h
e
s
i
m
p
ler
N
L
P
s
tr
u
ct
u
r
e
(
f
e
w
er
v
ar
iab
les,
s
p
ar
s
er
J
ac
o
b
ian
,
n
o
i
m
p
licit
m
id
p
o
in
t
co
u
p
li
n
g
)
co
n
tr
ib
u
te
s
to
m
o
r
e
r
o
b
u
s
t
co
n
v
er
g
e
n
ce
,
p
ar
ticu
lar
l
y
u
n
d
er
m
o
d
el
m
i
s
m
a
tch
.
2
.
5
.
H
y
brid
NN
m
et
ho
ds
L
ea
r
n
i
n
g
-
au
g
m
en
ted
MP
C
m
eth
o
d
s
ca
n
b
e
ca
te
g
o
r
ized
b
y
t
h
eir
d
eg
r
ee
o
f
s
o
l
v
er
i
n
v
o
lv
e
m
e
n
t.
So
lv
er
-
in
-
t
h
e
-
lo
o
p
m
et
h
o
d
s
r
etain
t
h
e
f
u
ll
o
p
ti
m
izatio
n
s
o
l
v
er
an
d
u
s
e
t
h
e
NN
o
n
l
y
to
ac
c
eler
ate
co
n
v
er
g
en
ce
(
e.
g
.
,
w
ar
m
-
s
tar
ti
n
g
)
.
So
l
v
er
-
s
elec
tio
n
m
et
h
o
d
s
u
s
e
a
lear
n
ed
class
i
f
ier
to
c
h
o
o
s
e
a
m
o
n
g
s
o
lv
er
s
b
u
t
s
till
ex
ec
u
te
tr
ad
itio
n
a
l
o
p
ti
m
izatio
n
.
So
lv
er
-
r
ep
lace
m
en
t
m
et
h
o
d
s
r
ep
lace
o
n
lin
e
o
p
tim
izatio
n
en
tire
l
y
w
i
th
d
ir
ec
t
p
o
licy
i
n
f
er
en
ce
,
u
s
i
n
g
s
a
f
et
y
m
ec
h
a
n
is
m
s
to
f
all
b
ac
k
t
o
th
e
s
o
lv
er
w
h
e
n
n
ee
d
ed
.
T
h
r
ee
r
e
p
r
esen
tativ
e
ap
p
r
o
ac
h
es
s
p
an
n
in
g
th
i
s
s
p
ec
tr
u
m
ar
e
ev
alu
a
ted
,
tr
ain
ed
o
n
4
,
6
5
0
s
am
p
les
f
r
o
m
t
w
o
tr
aj
ec
to
r
y
f
a
m
i
lie
s
(
cir
cle,
f
ig
u
r
e
-
ei
g
h
t
)
u
n
d
er
d
iv
er
s
e
co
n
d
itio
n
s
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
5
,
Octo
b
e
r
20
26
:
1
7
6
8
-
1778
1772
−
Me
th
o
d
1
:
NN
w
ar
m
-
s
tar
t (
N
N
-
W
S)
A
f
u
ll
y
co
n
n
ec
ted
n
et
w
o
r
k
(
8
→
128
→
128
→
64
→
103,
r
ec
tif
ied
lin
ea
r
u
n
it
(
R
e
L
U
)
ac
ti
v
atio
n
s
)
p
r
ed
icts
n
ea
r
-
o
p
ti
m
al
N
L
P
d
ec
is
io
n
v
ar
iab
les
f
r
o
m
t
h
e
r
o
b
o
t
s
tate
an
d
r
ef
er
en
ce
[
1
3
]
.
I
n
p
u
t
f
ea
t
u
r
es
ar
e
[
,
,
,
r
e
f
,
r
e
f
,
̇
r
ef
,
̇
r
ef
,
]
.
T
h
e
p
r
ed
ictio
n
in
itialize
s
I
P
OPT
v
ia
.
_
(
)
,
r
etain
in
g
f
u
ll
s
o
lv
e
r
f
ea
s
ib
ili
t
y
a
n
d
o
p
ti
m
alit
y
g
u
a
r
an
tees.
T
r
ain
in
g
u
s
es
A
d
a
m
o
p
tim
izer
,
5
0
0
ep
o
ch
s
(
f
i
n
al
m
ea
n
s
q
u
ar
ed
er
r
o
r
(
)
:
8
.
68
×
10
−
4
).
−
Me
th
o
d
2
:
a
d
ap
tiv
e
m
et
h
o
d
s
elec
to
r
(
A
MS)
A
b
in
ar
y
cla
s
s
i
f
ier
(
9
→
64
→
32
→
1
,
s
ig
m
o
id
o
u
tp
u
t)
p
r
ed
icts
w
h
i
ch
tr
an
s
cr
ip
tio
n
m
et
h
o
d
ac
h
iev
es
lo
w
er
er
r
o
r
at
ea
ch
co
n
tr
o
l
s
te
p
[
1
8
]
.
A
n
ad
d
itio
n
al
i
n
p
u
t
f
ea
tu
r
e
n
o
i
s
e
en
co
d
es
t
h
e
n
o
is
e
le
v
el.
T
r
ain
in
g
u
s
es
b
in
ar
y
cr
o
s
s
-
e
n
tr
o
p
y
,
3
0
0
e
p
o
ch
s
(
ac
cu
r
ac
y
:
8
9
.
7
%).
B
o
th
s
o
lv
er
s
ar
e
d
is
p
atch
ed
n
o
r
m
all
y
o
n
l
y
th
e
s
elec
tio
n
is
lear
n
ed
.
−
Me
th
o
d
3
: N
A
-
MP
C
A
p
o
lic
y
n
et
w
o
r
k
(
9
→
128
→
64
→
32
→
2
,
tan
h
o
u
tp
u
t
s
ca
led
to
c
o
n
tr
o
l
b
o
u
n
d
s
)
d
ir
ec
tl
y
p
r
ed
icts
[
∗
,
∗
]
,
r
ep
lacin
g
o
n
lin
e
o
p
ti
m
izatio
n
w
it
h
d
ir
ec
t
p
o
lic
y
i
n
f
er
en
ce
[
1
4
]
,
[
1
6
]
.
A
s
a
f
et
y
m
ec
h
a
n
i
s
m
m
o
n
ito
r
s
tr
ac
k
in
g
er
r
o
r
:
if
∥
−
∥
>
=
2
c
m
,
th
e
c
o
n
tr
o
ller
f
alls
b
ac
k
to
f
u
ll
I
P
OPT
(
R
K4
MS)
[
1
7
]
.
T
r
ain
in
g
u
s
es M
SE
lo
s
s
,
5
0
0
ep
o
ch
s
(
f
i
n
al
MSE
: 0
.
1
9
2
)
.
3.
RE
SU
L
T
S AN
D
D
I
SCU
SS
I
O
N
3
.
1
.
E
x
peri
m
ent
a
l set
up
A
ll
m
et
h
o
d
s
ar
e
ev
al
u
ated
o
n
f
iv
e
tr
aj
ec
to
r
y
f
a
m
ilie
s
:
(
i
)
s
m
o
o
t
h
cir
cu
lar
(
=
0
.
3
m
)
,
(
ii
)
s
m
o
o
th
f
i
g
u
r
e
-
ei
g
h
t
(
=
0
.
25
m
)
,
(
iii
)
lem
n
i
s
ca
te
o
f
B
er
n
o
u
lli
(
=
0
.
35
m
)
,
(
iv
)
S
-
c
u
r
v
e
(
Fo
u
r
ier
-
b
ased
)
,
an
d
(
v
)
m
u
lti
-
w
a
y
p
o
in
t
p
ath
(
6
p
o
in
ts
,
cu
b
ic
s
p
lin
e)
.
T
r
a
j
ec
to
r
ies
s
p
an
cu
r
v
at
u
r
e
p
r
o
f
iles
f
r
o
m
m
ax
=
3
.
33
to
68
.
2
m
−
1
.
T
r
a
j
ec
to
r
ies
1
–
2
a
r
e
u
s
ed
f
o
r
NN
tr
ain
in
g
;
tr
aj
ec
to
r
ies
3
–
5
ar
e
h
eld
o
u
t
f
o
r
g
e
n
er
aliza
tio
n
test
in
g
.
Fig
u
r
e
2
illu
s
tr
ates t
h
e
tr
aj
ec
to
r
y
l
ib
r
ar
y
.
Fig
u
r
e
2
.
Fiv
e
tr
aj
ec
to
r
y
f
a
m
il
ies u
s
ed
f
o
r
ev
alu
a
tio
n
C
ir
cle
an
d
f
ig
u
r
e
-
eig
h
t
(
s
o
lid
l
in
es)
s
er
v
e
as
t
h
e
NN
tr
ai
n
i
n
g
s
et;
le
m
n
i
s
ca
te,
S
-
c
u
r
v
e,
an
d
w
a
y
p
o
in
t
s
(
d
ash
ed
lin
e
s
)
ar
e
h
eld
o
u
t
f
o
r
g
en
er
aliza
tio
n
tes
tin
g
.
C
u
r
v
atu
r
e
r
an
g
es
s
p
an
m
ax
=
3
.
33
to
68
.
2
m
−
1
.
T
e
n
e
x
p
e
r
im
en
t
s
ev
a
lu
a
te
r
o
b
u
s
tn
es
s
an
d
g
en
e
r
a
li
z
a
ti
o
n
:
n
o
m
in
al
(
E
x
p
.
1
)
,
m
o
d
e
l
m
is
m
a
t
ch
±
1
0
%
(
E
x
p
s
.
2
–
3
)
,
G
a
u
s
s
i
an
n
o
is
e
∈
{
0
.
5
,
1
,
2
}
cm
(
E
x
p
s
.
4
–
6
)
,
f
ig
u
r
e
-
e
ig
h
t
(
E
x
p
.
7
)
,
a
n
d
t
h
r
e
e
h
e
l
d
-
o
u
t
t
r
a
je
c
t
o
r
i
es
(
E
x
p
s
.
8
–
1
0
)
.
A
l
l
ex
p
e
r
im
e
n
t
s
u
s
e
=
20
,
=
2
.
0
s
,
=
0
.
1
s
,
r
e
p
e
a
t
e
d
w
ith
1
0
r
an
d
o
m
s
e
ed
s
(
4
2
–
5
1
)
,
t
o
t
a
l
in
g
5
0
0
s
im
u
l
at
i
o
n
s
(
5
m
e
th
o
d
s
×
1
0
e
x
p
e
r
im
en
ts
×
1
0
s
e
e
d
s
)
.
S
im
u
l
at
i
o
n
s
r
u
n
o
n
I
n
t
e
l
X
e
o
n
(
4
C
PU
c
o
r
e
s
,
3
0
G
B
R
A
M
)
.
S
t
at
is
t
i
c
al
s
ig
n
if
i
c
an
c
e
i
s
as
s
ess
e
d
v
i
a
th
e
W
il
c
o
x
o
n
s
ig
n
e
d
-
r
a
n
k
t
e
s
t
w
ith
=
0
.
01
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
n
eu
r
a
l n
etw
o
r
k
-
a
u
g
me
n
ted
mo
d
el
p
r
ed
ictive
co
n
tr
o
l
fo
r
…
(
Ta
o
u
fik
B
elke
b
ir
)
1773
3
.
1
.
1
.
B
a
s
eline
re
s
ults
T
ab
le
1
p
r
esen
ts
t
h
e
b
asel
in
e
co
m
p
ar
is
o
n
b
et
w
ee
n
R
K4
MS
an
d
Her
m
ite
-
Si
m
p
s
o
n
d
ir
ec
t
co
llo
ca
tio
n
.
R
K4
MS
ac
h
ie
v
e
s
lo
w
er
o
r
eq
u
al
tr
ac
k
in
g
er
r
o
r
in
9
o
f
1
0
ex
p
er
im
en
ts
(
s
i
g
n
if
ica
n
tl
y
b
etter
i
n
7
,
n
o
n
-
s
i
g
n
i
f
ican
t
i
n
2
)
.
T
h
e
o
n
l
y
ex
ce
p
tio
n
is
th
e
le
m
n
i
s
c
ate
(
E
x
p
.
8
)
,
w
h
er
e
H
-
S
ac
h
iev
es
a
s
tati
s
ticall
y
s
ig
n
i
f
ica
n
t
b
u
t
m
ar
g
in
a
l
0
.
0
0
1
c
m
ad
v
a
n
tag
e.
B
o
th
m
et
h
o
d
s
m
a
i
n
tai
n
s
u
b
-
ce
n
ti
m
eter
n
o
m
i
n
al
ac
cu
r
ac
y
(
0
.
5
7
c
m
v
s
.
0
.
7
1
c
m
)
.
U
n
d
er
h
i
g
h
n
o
i
s
e
(
≥
1
c
m
,
E
x
p
s
.
5
–
6
)
,
th
e
d
i
f
f
er
en
ce
b
ec
o
m
e
s
n
o
n
-
s
i
g
n
if
ica
n
t,
in
d
icati
n
g
t
h
at
m
ea
s
u
r
e
m
e
n
t
n
o
is
e
d
o
m
in
a
tes
tr
an
s
cr
ip
tio
n
ef
f
ec
t
s
.
Av
er
ag
e
s
o
l
v
e
ti
m
e
s
ar
e
co
m
p
ar
ab
le:
1
4
2
m
s
(
R
K4
MS)
v
s
.
1
5
2
m
s
(
H
-
S).
No
tab
l
y
,
t
h
e
co
r
r
ec
ted
h
ea
d
in
g
co
s
t
(
T
h
eo
r
em
1
)
r
ed
u
ce
s
f
ig
u
r
e
-
e
ig
h
t
er
r
o
r
f
r
o
m
6
.
0
9
c
m
to
0
.
2
9
c
m
(
9
5
% r
ed
u
ctio
n
)
.
T
ab
le
1
.
B
aselin
e
co
m
p
ar
is
o
n
:
R
K4
MS
v
s
.
Her
m
ite
-
Si
m
p
s
o
n
d
ir
ec
t c
o
llo
ca
tio
n
.
Me
an
tr
ac
k
in
g
er
r
o
r
(
cm
)
o
v
er
1
0
s
ee
d
s
.
Sig
n
i
f
ica
n
ce
:
⋆
⋆
<
0
.
01
,
n
s
=
n
o
t sig
n
i
f
ica
n
t
#
Ex
p
e
r
i
me
n
t
R
K
4
M
S
H
-
S
(
cm
)
W
i
n
n
e
r
1
N
o
mi
n
a
l
0
.
5
6
7
0
.
7
1
4
0
.
1
4
7
0
.
0
0
2
⋆
⋆
R
K
4
M
S
2
M
o
d
e
l
−
1
0
%
1
.
0
0
6
1
.
1
6
3
0
.
1
5
7
0
.
0
0
2
⋆
⋆
R
K
4
M
S
3
M
o
d
e
l
+
1
0
%
0
.
2
2
7
0
.
3
5
7
0
.
1
3
0
0
.
0
0
2
⋆
⋆
R
K
4
M
S
4
N
o
i
se
=
0
.
5
cm
0
.
7
0
4
0
.
7
7
2
0
.
0
6
8
0
.
0
0
2
⋆
⋆
R
K
4
M
S
5
N
o
i
se
=
1
cm
0
.
9
3
7
0
.
9
2
7
−
0
.
0
1
0
0
.
6
9
5
—
6
N
o
i
se
=
2
cm
1
.
2
3
8
1
.
2
9
2
0
.
0
5
4
0
.
3
7
5
—
7
F
i
g
u
r
e
-
e
i
g
h
t
0
.
2
9
2
0
.
3
1
8
0
.
0
2
6
0
.
0
0
2
⋆
⋆
R
K
4
M
S
8
L
e
mn
i
sca
t
e
0
.
4
6
2
0
.
4
6
1
−
0
.
0
0
1
0
.
0
0
2
⋆
⋆
H
-
S
9
S
-
c
u
r
v
e
0
.
3
4
8
0
.
4
7
7
0
.
1
2
9
0
.
0
0
2
⋆
⋆
R
K
4
M
S
10
W
a
y
p
o
i
n
t
s
0
.
3
2
0
0
.
3
2
4
0
.
0
0
5
0
.
0
0
2
⋆
⋆
R
K
4
M
S
3
.
1
.
2
.
H
y
brid
m
et
ho
d r
es
ults
T
ab
le
2
p
r
esen
ts
t
h
e
f
iv
e
-
m
et
h
o
d
co
m
p
ar
is
o
n
.
Fi
g
u
r
e
3
v
i
s
u
alize
s
t
h
e
tr
ac
k
in
g
er
r
o
r
d
is
tr
ib
u
tio
n
s
ac
r
o
s
s
all
ten
ex
p
er
i
m
en
ts
,
ill
u
s
tr
ati
n
g
th
at
n
o
s
i
n
g
le
m
et
h
o
d
d
o
m
in
a
tes
all
s
ce
n
ar
io
s
.
Ke
y
in
s
i
g
h
t
:
NN
-
W
S
m
atc
h
es
R
K4
MS
ac
cu
r
ac
y
o
n
all
e
x
p
er
i
m
e
n
t
s
(
in
c
lu
d
i
n
g
h
eld
-
o
u
t
tr
aj
ec
to
r
ies)
w
h
i
le
r
ed
u
cin
g
s
o
lv
e
ti
m
e
b
y
1
6
%.
A
MS
ac
h
iev
e
s
t
h
e
lo
w
e
s
t e
r
r
o
r
in
n
o
is
e
an
d
f
i
g
u
r
e
-
ei
g
h
t scen
ar
io
s
,
w
h
er
ea
s
N
A
-
MP
C
s
h
o
w
s
s
u
b
s
ta
n
tia
l
d
eg
r
ad
atio
n
o
n
h
eld
-
o
u
t tr
aj
ec
to
r
ies (
E
x
p
s
.
8
–
1
0
)
.
T
ab
le
2
.
Fiv
e
-
m
eth
o
d
co
m
p
ar
i
s
o
n
:
m
ea
n
tr
ac
k
in
g
er
r
o
r
(
cm
)
o
v
er
1
0
s
ee
d
s
.
Star
s
(
⋆
)
in
d
icate
th
e
s
tr
ict
w
i
n
n
er
p
er
ex
p
er
im
e
n
t; b
o
ld
f
ac
e
i
n
d
icate
s
th
e
b
est r
es
u
lt.
NN
-
W
S ti
es R
K4
MS
o
n
all
d
eter
m
in
i
s
ti
c
ex
p
er
i
m
e
n
ts
#
Ex
p
e
r
i
me
n
t
R
K
4
M
S
H
-
S
NN
-
WS
A
M
S
NA
-
M
P
C
W
i
n
n
e
r
1
N
o
mi
n
a
l
0
.
5
6
7
0
.
7
1
4
0
.
5
6
7
0
.
5
6
7
0
.
6
5
6
R
K
4
=
W
S
2
M
o
d
e
l
-
1
0
%
1
.
0
0
6
1
.
1
6
3
1
.
0
0
6
1
.
0
0
6
1
.
6
1
7
R
K
4
=
W
S
3
M
o
d
e
l
+
1
0
%
0
.
2
2
7
0
.
3
5
7
0
.
2
2
7
0
.
2
2
7
0
.
3
3
1
R
K
4
=
W
S
4
N
o
i
se
0
.
5
c
m
0
.
7
0
4
0
.
7
7
2
0
.
7
1
3
0
.
7
0
3
⋆
0
.
7
8
3
A
M
S
5
N
o
i
se
1
cm
0
.
9
3
7
0
.
9
2
7
0
.
9
2
2
0
.
8
8
5
⋆
0
.
9
1
3
A
M
S
6
N
o
i
se
2
cm
1
.
2
3
8
1
.
2
9
2
1
.
2
9
0
1
.
2
5
5
1
.
1
9
5
⋆
NA
-
M
P
C
7
F
i
g
u
r
e
-
e
i
g
h
t
0
.
2
9
2
0
.
3
1
8
0
.
2
9
2
0
.
2
8
9
⋆
1
.
4
0
1
A
M
S
8
L
e
mn
i
sca
t
e
0
.
4
6
2
0
.
4
6
1
⋆
0
.
4
6
2
0
.
4
6
6
2
.
0
6
4
H
-
S
9
S
-
C
u
r
v
e
0
.
3
4
8
0
.
4
7
7
0
.
3
4
8
0
.
3
9
0
2
.
0
5
9
R
K
4
=
W
S
10
W
a
y
p
o
i
n
t
s
0
.
3
2
0
0
.
3
2
4
0
.
3
2
0
0
.
3
2
2
1
.
9
4
7
R
K
4
=
W
S
No
s
in
g
le
m
et
h
o
d
d
o
m
in
ate
s
all
s
ce
n
ar
io
s
.
Strict
ac
cu
r
ac
y
w
i
n
s
ar
e
d
is
tr
ib
u
ted
b
et
w
ee
n
AM
S
(
3
/1
0
)
,
NA
-
MP
C
(
1
/1
0
)
,
an
d
H
-
S
(
1
/1
0
)
,
w
h
ile
NN
-
W
S
ties
R
K4
MS
o
n
5
/1
0
ex
p
er
im
e
n
t
s
in
clu
d
in
g
h
eld
-
o
u
t
tr
aj
ec
to
r
ies.
NN
w
ar
m
-
s
tar
t
p
r
eser
v
es
al
l
I
P
OPT
g
u
ar
an
tee
s
,
as
t
h
e
N
N
m
o
d
i
f
ies
o
n
l
y
t
h
e
i
n
itial
g
u
es
s
.
I
t
ac
h
iev
e
s
id
en
t
ical
ac
cu
r
ac
y
t
o
R
K4
MS
o
n
all
d
eter
m
i
n
i
s
tic
ex
p
er
i
m
e
n
t
s
(
1
–
3
,
7
–
1
0
)
in
cl
u
d
in
g
h
e
ld
-
o
u
t
tr
aj
ec
to
r
ies
w
h
i
le
r
ed
u
ci
n
g
a
v
er
ag
e
s
o
lv
e
t
i
m
e
f
r
o
m
1
4
2
m
s
to
1
2
1
m
s
(
1
6
%
r
ed
u
ct
io
n
)
[
1
3
]
.
A
d
ap
tiv
e
m
et
h
o
d
s
elec
to
r
ac
h
ie
v
es
8
9
.
7
%
class
i
f
icat
io
n
ac
c
u
r
ac
y
a
n
d
o
u
tp
er
f
o
r
m
s
t
h
e
b
ase
lin
e
i
n
3
o
f
1
0
s
ce
n
ar
io
s
,
in
cl
u
d
in
g
th
e
lo
w
es
t
er
r
o
r
o
n
f
ig
u
r
e
-
e
ig
h
t
(
0
.
2
8
9
c
m
)
.
I
ts
m
is
c
lass
if
ica
tio
n
r
is
k
is
b
o
u
n
d
ed
:
f
o
r
ac
cu
r
ac
y
an
d
m
et
h
o
d
g
ap
=
|
RK
4
(
)
−
H
-
S
(
)
|
,
th
e
ex
p
ec
te
d
ex
ce
s
s
er
r
o
r
s
atis
f
ies
[
]
≤
∗
+
(
1
−
)
[
1
8
]
.
W
ith
=
0
.
897
an
d
m
ax
=
0
.
157
c
m
,
th
e
w
o
r
s
t
-
ca
s
e
e
x
c
ess
is
0
.
0
1
6
cm
n
eg
lig
ib
le.
Neu
r
al
ap
p
r
o
x
i
m
ate
MP
C
ac
h
iev
e
s
a
s
u
b
s
ta
n
tial
late
n
c
y
r
ed
u
ctio
n
o
n
tr
ain
in
g
tr
aj
ec
to
r
ies
(
1
4
2
m
s
→
0
.
0
3
m
s
,
ap
p
r
o
x
i
m
atel
y
×
4
7
0
0
)
b
u
t
p
er
f
o
r
m
a
n
ce
d
iv
er
g
es
o
n
h
eld
-
o
u
t
tr
aj
ec
to
r
ies:
tr
ac
k
in
g
er
r
o
r
in
cr
ea
s
es
to
1
.
9
5
–
2
.
0
6
c
m
(
4
.
5
–
5
.
9
×
w
o
r
s
e
th
an
R
K4
MS)
,
an
d
s
af
et
y
f
allb
ac
k
r
ate
s
r
is
e
to
5
1
–
6
4
%.
T
h
e
s
af
et
y
m
ec
h
a
n
is
m
[
1
4
]
,
[
1
7
]
p
r
ev
en
ts
ca
tast
r
o
p
h
ic
f
ai
lu
r
e:
e
v
en
at
6
4
.
3
% f
allb
ac
k
(
le
m
n
i
s
ca
te)
,
tr
ac
k
in
g
er
r
o
r
r
em
ai
n
s
b
o
u
n
d
ed
at
2
.
0
6
c
m
.
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
5
,
Octo
b
e
r
20
26
:
1
7
6
8
-
1778
1774
Fig
u
r
e
3
.
Fiv
e
-
m
e
th
o
d
co
m
p
ar
is
o
n
ac
r
o
s
s
te
n
ex
p
er
i
m
en
t
s
3
.
1
.
3
.
G
ener
a
liza
t
io
n a
na
ly
s
is
T
ab
le
3
p
r
esen
ts
th
e
co
n
tr
o
lled
ev
alu
at
io
n
o
f
g
e
n
er
aliza
t
io
n
to
tr
aj
ec
to
r
ies
o
u
ts
id
e
th
e
tr
ain
i
n
g
d
is
tr
ib
u
tio
n
.
T
h
r
ee
d
is
tin
ct
g
en
er
aliza
tio
n
b
eh
a
v
io
r
s
e
m
er
g
e,
f
o
r
m
i
n
g
a
h
ier
ar
ch
y
d
ir
ec
tl
y
r
ef
lec
tin
g
t
h
e
d
eg
r
ee
o
f
s
o
lv
er
in
v
o
l
v
e
m
e
n
t:
−
NN
-
W
S
g
e
n
er
alize
s
w
i
th
o
u
t
m
ea
s
u
r
ab
le
d
eg
r
ad
atio
n
.
T
r
ac
k
in
g
er
r
o
r
is
id
en
tical
to
R
K4
MS
o
n
all
h
eld
-
o
u
t
tr
aj
ec
to
r
ies.
T
h
e
s
o
lv
er
-
in
-
th
e
-
lo
o
p
ar
ch
itect
u
r
e
e
n
s
u
r
es
th
i
s
:
I
P
OP
T
co
n
v
er
g
es
to
o
p
ti
m
alit
y
r
eg
ar
d
less
o
f
i
n
itial
g
u
ess
q
u
al
it
y
[
1
3
]
.
−
A
M
S
g
en
er
al
izes
w
ell.
Ma
x
i
m
u
m
d
eg
r
ad
atio
n
r
elati
v
e
to
R
K4
MS
is
1
2
%
(
S
-
cu
r
v
e:
0
.
3
9
0
v
s
.
0
.
3
4
8
c
m
)
.
T
h
e
class
i
f
ier
co
r
r
ec
tl
y
s
elec
ts
R
K4
MS
f
o
r
m
o
s
t
h
eld
-
o
u
t c
o
n
d
itio
n
s
.
−
NA
-
MP
C
s
h
o
w
s
s
tati
s
tical
l
y
s
ig
n
if
ican
t
d
eg
r
ad
atio
n
(
<
0
.
01
)
.
E
r
r
o
r
in
cr
ea
s
es
f
r
o
m
0
.
6
6
–
1
.
4
0
c
m
(
tr
ain
in
g
)
to
1
.
9
5
–
2
.
0
6
c
m
(
h
eld
-
o
u
t)
,
w
i
th
f
a
llb
ac
k
r
ates
r
is
in
g
f
r
o
m
0
–
1
6
%
to
5
1
–
6
4
%.
T
h
e
s
af
et
y
m
ec
h
a
n
i
s
m
co
r
r
ec
tl
y
d
etec
t
s
o
u
t
-
of
-
d
is
tr
ib
u
tio
n
b
eh
a
v
io
r
[
1
4
]
,
[
1
7
]
.
T
ab
le
3
.
Gen
er
aliza
tio
n
r
esu
lt
s
: tr
ain
i
n
g
v
s
.
h
e
ld
-
o
u
t tr
aj
ec
to
r
ies.
Me
an
tr
ac
k
in
g
er
r
o
r
(
cm
)
an
d
NA
-
MP
C
f
allb
ac
k
r
ate
(
%
)
G
r
o
u
p
T
r
a
j
e
c
t
o
r
y
R
K
4
M
S
NN
-
WS
A
M
S
NA
-
M
P
C
F
a
l
l
b
a
c
k
T
r
a
i
n
i
n
g
C
i
r
c
l
e
0
.
5
6
7
0
.
5
6
7
0
.
5
6
7
0
.
6
5
6
0
.
0
%
T
r
a
i
n
i
n
g
f
i
g
u
r
e
-
e
i
g
h
t
0
.
2
9
2
0
.
2
9
2
0
.
2
8
9
1
.
4
0
1
1
6
.
0
%
H
e
l
d
-
o
u
t
L
e
mn
i
sca
t
e
0
.
4
6
2
0
.
4
6
2
0
.
4
6
6
2
.
0
6
4
6
4
.
3
%
H
e
l
d
-
o
u
t
S
-
C
u
r
v
e
0
.
3
4
8
0
.
3
4
8
0
.
3
9
0
2
.
0
5
9
5
0
.
8
%
H
e
l
d
-
o
u
t
W
a
y
p
o
i
n
t
s
0
.
3
2
0
0
.
3
2
0
0
.
3
2
2
1
.
9
4
7
6
2
.
7
%
T
h
is
h
ier
ar
ch
y
s
o
lv
er
-
in
-
t
h
e
-
lo
o
p
>
s
o
lv
er
-
s
elec
tio
n
>
s
o
lv
er
-
r
ep
lace
m
e
n
t
s
u
g
g
e
s
ts
a
d
esig
n
p
r
in
cip
le:
th
e
d
eg
r
ee
o
f
s
o
lv
er
in
v
o
l
v
e
m
e
n
t
i
n
th
e
co
n
tr
o
l
p
ip
elin
e
is
a
k
e
y
f
ac
to
r
in
f
lu
e
n
c
in
g
g
en
er
aliza
tio
n
ca
p
ab
ilit
y
f
o
r
lear
n
in
g
-
au
g
m
e
n
ted
MP
C
,
co
n
s
is
te
n
t
w
it
h
f
i
n
d
in
g
s
o
f
C
au
lig
i
et
a
l
.
[
2
6
]
an
d
Nu
b
er
t
et
a
l
.
[
2
3
]
.
Fro
m
a
t
h
eo
r
etica
l
p
er
s
p
ec
ti
v
e,
t
h
is
h
ier
ar
ch
y
ca
n
b
e
u
n
d
er
s
to
o
d
th
r
o
u
g
h
t
h
e
le
n
s
o
f
a
p
p
r
o
x
i
m
atio
n
an
d
g
en
er
aliza
tio
n
th
eo
r
y
.
So
l
v
er
-
in
-
th
e
-
lo
o
p
m
et
h
o
d
s
(
NN
-
W
S
)
u
s
e
th
e
NN
o
n
l
y
to
p
r
o
v
id
e
an
in
it
ial
g
u
es
s
;
th
e
s
o
lv
er
t
h
en
co
r
r
ec
ts
an
y
ap
p
r
o
x
i
m
atio
n
er
r
o
r
,
ef
f
ec
ti
v
el
y
ac
t
in
g
as
a
r
eg
u
lar
izer
th
a
t
b
o
u
n
d
s
th
e
o
u
tp
u
t
to
th
e
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
n
eu
r
a
l n
etw
o
r
k
-
a
u
g
me
n
ted
mo
d
el
p
r
ed
ictive
co
n
tr
o
l
fo
r
…
(
Ta
o
u
fik
B
elke
b
ir
)
1775
f
ea
s
ib
le
a
n
d
o
p
ti
m
al
s
et.
So
l
v
er
-
s
elec
tio
n
m
e
th
o
d
s
(
A
MS)
r
ed
u
ce
th
e
p
r
o
b
le
m
to
b
in
ar
y
c
lass
i
f
icatio
n
,
w
h
ic
h
h
as
lo
w
er
s
a
m
p
le
co
m
p
le
x
it
y
t
h
an
f
u
ll
p
o
lic
y
r
e
g
r
ess
io
n
.
So
lv
er
-
r
ep
lace
m
en
t
m
eth
o
d
s
(
N
A
-
MP
C
)
m
u
s
t
ap
p
r
o
x
im
a
te
th
e
e
n
tire
i
n
p
u
t
-
to
-
co
n
tr
o
l
m
ap
p
in
g
,
a
h
i
g
h
e
r
-
d
i
m
e
n
s
io
n
al
f
u
n
ctio
n
th
at
i
s
in
h
er
en
t
l
y
m
o
r
e
s
en
s
iti
v
e
to
d
is
tr
ib
u
t
io
n
s
h
i
f
t
[
2
]
,
[
1
5
]
.
T
h
is
an
al
y
s
is
p
r
o
v
id
es
a
p
r
in
cip
led
b
asis
f
o
r
th
e
o
b
s
er
v
ed
h
ier
ar
ch
y
b
ey
o
n
d
th
e
e
m
p
ir
ica
l e
v
id
en
ce
.
3
.
2
.
Dis
cus
s
io
n
T
ab
le
3
an
d
Fig
u
r
e
4
p
r
o
v
id
es
an
o
v
er
v
ie
w
o
f
th
e
b
aseli
n
e
an
d
h
y
b
r
id
m
eth
o
d
r
esu
lt
s
ac
r
o
s
s
all
ten
ex
p
er
i
m
e
n
tal
s
ce
n
ar
io
s
,
co
n
s
o
lid
atin
g
th
e
f
i
n
d
in
g
s
f
r
o
m
t
h
e
p
r
ec
ed
in
g
s
u
b
s
ec
tio
n
s
.
He
ad
in
g
co
s
t
i
m
p
ac
t
:
a
m
o
n
g
t
h
e
f
ac
to
r
s
e
v
al
u
ated
,
h
ea
d
in
g
co
s
t
f
o
r
m
u
latio
n
h
as
t
h
e
lar
g
e
s
t
m
ea
s
u
r
ed
im
p
ac
t
o
n
co
m
p
le
x
tr
aj
ec
to
r
ies.
T
h
e
f
i
g
u
r
e
-
ei
g
h
t
r
esu
lt
s
(
T
ab
le
1
)
co
n
f
ir
m
t
h
at
t
h
e
-
a
m
b
ig
u
it
y
i
n
s
in
2
(
)
ca
u
s
es
s
e
v
er
e
p
r
ac
tical
f
ailu
r
e
o
n
h
i
g
h
-
cu
r
v
atu
r
e
tr
aj
ec
to
r
ies,
r
ed
u
cin
g
tr
ac
k
in
g
er
r
o
r
b
y
ap
p
r
o
x
im
atel
y
9
5
%
w
h
en
co
r
r
ec
ted
,
as
s
h
o
w
n
i
n
Fi
g
u
r
e
4
(
a)
.
T
h
is
f
i
n
d
in
g
is
ap
p
lic
ab
le
to
o
th
er
n
o
n
h
o
lo
n
o
m
ic
MP
C
f
o
r
m
u
latio
n
s
e
m
p
lo
y
i
n
g
s
in
2
-
b
ased
h
ea
d
i
n
g
co
s
ts
[
3
]
,
[
5
]
.
P
r
ac
tical
g
u
id
eli
n
es
:
f
o
r
g
en
er
al
d
ep
lo
y
m
e
n
t,
R
K
4
MS
w
it
h
N
N
w
ar
m
-
s
tar
t
o
f
f
er
s
th
e
m
o
s
t
f
av
o
r
ab
le
s
p
ee
d
/acc
u
r
ac
y
tr
ad
e
-
o
f
f
,
as
s
u
m
m
ar
ized
in
Fig
u
r
e
4
(
b
)
.
A
MS
is
p
r
ef
er
r
ed
w
h
e
n
o
p
er
atin
g
co
n
d
itio
n
s
v
ar
y
s
ig
n
i
f
ica
n
tl
y
an
d
tr
aj
ec
to
r
y
-
d
ep
en
d
en
t
p
er
f
o
r
m
an
ce
g
ap
s
e
m
er
g
e.
NA
-
MP
C
s
u
its
co
m
p
u
te
-
co
n
s
tr
ain
ed
p
lat
f
o
r
m
s
o
p
er
atin
g
w
ith
in
th
e
tr
ain
i
n
g
d
is
tr
ib
u
tio
n
,
t
h
o
u
g
h
s
a
f
et
y
f
allb
ac
k
to
f
u
l
l
I
P
OPT
is
m
a
n
d
ato
r
y
,
g
iv
e
n
t
h
e
ele
v
ated
f
allb
ac
k
r
ates
o
n
h
eld
-
o
u
t
tr
aj
ec
to
r
ies
s
h
o
w
n
i
n
Fig
u
r
e
4
(
c
)
[
1
4
]
,
[
1
6
]
,
[
1
7
]
.
T
h
e
g
e
n
er
aliza
tio
n
h
ier
ar
ch
y
d
escr
ib
ed
in
s
ec
tio
n
3
.
1
.
3
,
illu
s
tr
ated
in
Fi
g
u
r
e
4
(
d
),
p
r
o
v
id
es a
p
r
in
cip
led
b
asis
f
o
r
th
is
m
et
h
o
d
s
elec
tio
n
.
Statis
t
ical
r
ig
o
r
:
a
ll
clai
m
s
ar
e
s
u
p
p
o
r
ted
b
y
W
ilco
x
o
n
s
i
g
n
ed
-
r
an
k
te
s
ts
(
=
10
s
ee
d
s
)
.
NN
-
W
S
s
h
o
w
s
n
o
s
ig
n
i
f
ica
n
t
d
if
f
er
e
n
ce
f
r
o
m
R
K4
MS
i
n
an
y
ex
p
er
i
m
en
t,
co
n
f
ir
m
i
n
g
ac
c
u
r
ac
y
p
r
eser
v
atio
n
.
N
A
-
MP
C
s
h
o
w
s
s
tat
is
tical
l
y
s
i
g
n
i
f
ican
t
d
if
f
er
en
ce
s
(
<
0
.
01
)
in
8
o
f
1
0
ex
p
er
i
m
e
n
ts
,
co
n
f
ir
m
i
n
g
th
at
its
ac
cu
r
ac
y
d
eg
r
ad
atio
n
i
s
n
o
t
a
s
a
m
p
li
n
g
ar
ti
f
ac
t.
L
i
m
itat
i
o
n
s
an
d
f
u
t
u
r
e
w
o
r
k
:
t
h
e
cu
r
r
en
t
ev
al
u
atio
n
i
s
s
i
m
u
lat
io
n
-
o
n
l
y
;
h
a
r
d
w
ar
e
d
ep
lo
y
m
e
n
t
o
n
a
p
h
y
s
ica
l
T
u
r
tleB
o
t3
B
u
r
g
er
w
it
h
r
ea
l
-
ti
m
e
R
OS2
in
te
g
r
atio
n
i
s
p
lan
n
ed
as
f
u
t
u
r
e
w
o
r
k
.
N
A
-
MP
C
ac
cu
r
ac
y
d
e
g
r
ad
es
o
n
h
eld
-
o
u
t
tr
aj
ec
to
r
ies,
r
eq
u
ir
in
g
ex
p
an
d
ed
tr
ain
i
n
g
d
ata
o
r
o
n
lin
e
ad
ap
tatio
n
f
o
r
b
r
o
ad
e
r
d
ep
l
o
y
m
en
t.
Fu
t
u
r
e
d
ir
ec
tio
n
s
in
c
lu
d
e
o
n
li
n
e
lear
n
i
n
g
f
o
r
ad
ap
tiv
e
NN
r
ef
in
e
m
e
n
t
[
2
]
,
d
ata
-
d
r
iv
e
n
d
y
n
a
m
ics
v
ia
Ko
o
p
m
a
n
o
p
er
ato
r
s
[
2
5
]
o
r
p
h
y
s
ics
-
i
n
f
o
r
m
ed
NN
s
[
2
9
]
,
in
teg
r
atio
n
w
it
h
s
a
f
et
y
-
co
n
s
tr
ai
n
ed
n
eu
r
al
co
n
tr
o
ller
s
[
2
3
]
,
an
d
co
m
p
ar
is
o
n
w
it
h
ac
ad
o
s
[
7
]
an
d
d
is
tr
ib
u
ted
MP
C
ar
ch
itect
u
r
es
[
3
0
]
f
o
r
em
b
e
d
d
ed
r
ea
l
-
ti
m
e
d
ep
lo
y
m
e
n
t.
Key
ta
k
ea
w
a
y
:
NN
-
W
S
(
s
o
lv
er
-
in
-
t
h
e
-
lo
o
p
)
co
n
s
is
ten
tl
y
m
atc
h
es
o
r
i
m
p
r
o
v
es
u
p
o
n
R
K4
MS,
w
h
ile
NA
-
MP
C
(
s
o
l
v
er
-
r
ep
lace
m
e
n
t)
s
h
o
w
s
m
ar
k
ed
d
eg
r
ad
atio
n
o
n
h
eld
-
o
u
t tr
aj
ec
to
r
ies (
E
x
p
s
.
8
–
10)
(
a)
(
b
)
(
c)
(
d
)
Fig
u
r
e
4
.
Su
m
m
ar
y
o
f
b
aseli
n
e
an
d
h
y
b
r
id
m
et
h
o
d
r
esu
lt
s
ac
r
o
s
s
all
ten
s
ce
n
ar
io
s
: (
a)
f
i
g
u
r
e
-
ei
g
h
t tr
ac
k
i
n
g
er
r
o
r
b
ef
o
r
e
an
d
af
ter
th
e
co
r
r
ec
ted
h
ea
d
in
g
co
s
t f
o
r
m
u
latio
n
,
(
b
)
av
er
ag
e
s
o
lv
e
ti
m
e
ac
r
o
s
s
m
et
h
o
d
s
,
(
c)
I
P
OPT
f
allb
ac
k
r
ate
o
n
tr
ain
in
g
v
er
s
u
s
h
e
ld
-
o
u
t tr
aj
ec
to
r
i
es,
an
d
(
d
)
m
ea
n
tr
ac
k
i
n
g
er
r
o
r
o
n
tr
ain
in
g
v
er
s
u
s
h
e
ld
-
o
u
t tr
aj
ec
to
r
ies b
y
m
eth
o
d
Evaluation Warning : The document was created with Spire.PDF for Python.
I
SS
N
:
1
6
9
3
-
6930
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
,
Vo
l.
24
,
No
.
5
,
Octo
b
e
r
20
26
:
1
7
6
8
-
1778
1776
4.
CO
NCLU
SI
O
N
T
h
is
p
ap
er
p
r
esen
ted
a
u
n
if
i
ed
ev
alu
atio
n
o
f
NM
P
C
tr
an
s
cr
ip
tio
n
m
eth
o
d
s
an
d
th
r
e
e
lear
n
in
g
-
au
g
m
e
n
ted
h
y
b
r
id
ap
p
r
o
ac
h
es
f
o
r
d
i
f
f
er
e
n
tial
-
d
r
iv
e
tr
aj
ec
to
r
y
tr
ac
k
i
n
g
o
n
a
T
u
r
tleB
o
t3
B
u
r
g
er
.
T
h
e
m
ai
n
f
i
n
d
in
g
s
ar
e:
(
i)
t
h
e
co
r
r
ec
ted
h
ea
d
in
g
co
s
t
(
T
h
eo
r
e
m
1
)
el
i
m
i
n
ates
a
π
-
a
m
b
i
g
u
it
y
in
th
e
s
tan
d
ar
d
2
(
⋅
)
f
o
r
m
u
latio
n
,
y
ield
i
n
g
ap
p
r
o
x
i
m
atel
y
9
5
%
er
r
o
r
r
ed
u
ctio
n
o
n
h
i
g
h
-
c
u
r
v
at
u
r
e
t
r
aj
ec
to
r
ies;
(
ii)
R
K4
MS
p
r
o
v
id
es
th
e
m
o
s
t
r
o
b
u
s
t
b
as
elin
e
ac
r
o
s
s
d
i
v
er
s
e
co
n
d
itio
n
s
;
(
iii)
N
N
w
ar
m
-
s
tar
tin
g
i
s
th
e
r
ec
o
m
m
e
n
d
ed
au
g
m
e
n
tatio
n
,
co
m
b
i
n
i
n
g
s
o
lv
er
-
g
u
ar
a
n
teed
ac
cu
r
ac
y
w
it
h
1
6
%
s
o
lv
e
-
ti
m
e
r
ed
u
ctio
n
;
an
d
(
iv
)
th
e
g
en
er
aliza
tio
n
h
ier
ar
ch
y
r
ev
e
aled
b
y
t
h
e
r
es
u
lt
s
co
r
r
elate
s
d
ir
ec
tl
y
w
it
h
t
h
e
d
eg
r
ee
o
f
s
o
lv
er
i
n
v
o
l
v
e
m
e
n
t.
T
h
is
h
ier
ar
ch
y
o
f
f
er
s
a
p
r
ac
tic
al
d
esig
n
p
r
in
cip
le:
w
h
e
n
s
ele
ctin
g
lear
n
i
n
g
-
a
u
g
m
en
ted
MP
C
ar
ch
itect
u
r
es,
t
h
e
d
eg
r
ee
o
f
s
o
lv
er
in
v
o
lv
e
m
e
n
t
s
h
o
u
ld
b
e
m
atch
ed
to
th
e
ex
p
ec
ted
v
ar
iab
ili
t
y
o
f
th
e
d
ep
lo
y
m
en
t
en
v
ir
o
n
m
e
n
t.
T
h
ese
r
esu
lts
p
o
in
t
to
w
ar
d
a
b
r
o
ad
e
r
o
p
p
o
r
tu
n
it
y
f
o
r
r
ea
l
-
t
i
m
e
NM
P
C
d
ep
lo
y
m
e
n
t
o
n
r
e
s
o
u
r
ce
-
co
n
s
tr
ai
n
ed
e
m
b
ed
d
ed
p
latf
o
r
m
s
,
w
h
er
e
co
m
b
i
n
i
n
g
cla
s
s
ical
o
p
ti
m
iza
tio
n
g
u
ar
an
tee
s
w
i
th
NN
e
f
f
icien
c
y
ca
n
en
ab
le
r
o
b
u
s
t a
n
d
co
m
p
u
tat
io
n
all
y
tr
a
ctab
le
co
n
tr
o
l.
F
UNDIN
G
I
NF
O
RM
AT
I
O
N
Au
t
h
o
r
s
s
tate
n
o
f
u
n
d
i
n
g
i
n
v
o
l
v
ed
.
AUTHO
R
CO
NT
RIB
UT
I
O
NS ST
A
T
E
M
E
NT
T
h
is
j
o
u
r
n
al
u
s
e
s
th
e
C
o
n
tr
ib
u
to
r
R
o
les
T
ax
o
n
o
m
y
(
C
R
ed
iT
)
to
r
ec
o
g
n
ize
in
d
i
v
id
u
al
au
th
o
r
co
n
tr
ib
u
tio
n
s
,
r
ed
u
ce
au
t
h
o
r
s
h
ip
d
is
p
u
tes,
an
d
f
ac
ilit
ate
co
lla
b
o
r
atio
n
.
Na
m
e
o
f
Aut
ho
r
C
M
So
Va
Fo
I
R
D
O
E
Vi
Su
P
Fu
T
ao
u
f
ik
B
elk
eb
ir
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
✓
Hich
a
m
B
elk
eb
ir
✓
✓
✓
✓
✓
An
as
s
Ma
n
s
o
u
r
i
✓
✓
✓
✓
✓
C
:
C
o
n
c
e
p
t
u
a
l
i
z
a
t
i
o
n
M
:
M
e
t
h
o
d
o
l
o
g
y
So
:
So
f
t
w
a
r
e
Va
:
Va
l
i
d
a
t
i
o
n
Fo
:
Fo
r
mal
a
n
a
l
y
si
s
I
:
I
n
v
e
st
i
g
a
t
i
o
n
R
:
R
e
so
u
r
c
e
s
D
:
D
a
t
a
C
u
r
a
t
i
o
n
O
:
W
r
i
t
i
n
g
-
O
r
i
g
i
n
a
l
D
r
a
f
t
E
:
W
r
i
t
i
n
g
-
R
e
v
i
e
w
&
E
d
i
t
i
n
g
Vi
:
Vi
su
a
l
i
z
a
t
i
o
n
Su
:
Su
p
e
r
v
i
si
o
n
P
:
P
r
o
j
e
c
t
a
d
mi
n
i
st
r
a
t
i
o
n
Fu
:
Fu
n
d
i
n
g
a
c
q
u
i
si
t
i
o
n
CO
NF
L
I
C
T
O
F
I
N
T
E
R
E
S
T
ST
A
T
E
M
E
NT
Au
t
h
o
r
s
s
tate
n
o
co
n
f
lic
t o
f
i
n
t
er
est.
DATA AV
AI
L
AB
I
L
I
T
Y
T
h
e
au
th
o
r
s
co
n
f
ir
m
t
h
at
th
e
d
ata
s
u
p
p
o
r
tin
g
t
h
e
f
i
n
d
in
g
s
o
f
t
h
is
s
t
u
d
y
ar
e
a
v
ailab
le
w
it
h
i
n
t
h
e
ar
ticle.
T
h
e
s
i
m
u
la
tio
n
co
d
e
is
av
ailab
le
f
r
o
m
t
h
e
co
r
r
esp
o
n
d
in
g
a
u
t
h
o
r
u
p
o
n
r
ea
s
o
n
ab
le
r
eq
u
est.
RE
F
E
R
E
NC
E
S
[
1
]
D
.
Q
.
M
a
y
n
e
,
J.
B
.
R
a
w
l
i
n
g
s,
C
.
V
.
R
a
o
,
a
n
d
P
.
O
.
M
.
S
c
o
k
a
e
r
t
,
“
C
o
n
st
r
a
i
n
e
d
mo
d
e
l
p
r
e
d
i
c
t
i
v
e
c
o
n
t
r
o
l
:
S
t
a
b
i
l
i
t
y
a
n
d
o
p
t
i
m
a
l
i
t
y
,
”
Au
t
o
m
a
t
i
c
a
,
v
o
l
.
3
6
,
n
o
.
6
,
p
p
.
7
8
9
–
8
1
4
,
J
u
n
.
2
0
0
0
,
d
o
i
:
1
0
.
1
0
1
6
/
S
0
0
0
5
-
1
0
9
8
(
9
9
)
0
0
2
1
4
-
9.
[
2
]
L
.
H
e
w
i
n
g
,
K
.
P
.
W
a
b
e
r
si
c
h
,
M
.
M
e
n
n
e
r
,
a
n
d
M
.
N
.
Z
e
i
l
i
n
g
e
r
,
“
L
e
a
r
n
i
n
g
-
B
a
se
d
M
o
d
e
l
P
r
e
d
i
c
t
i
v
e
C
o
n
t
r
o
l
:
T
o
w
a
r
d
S
a
f
e
L
e
a
r
n
i
n
g
i
n
C
o
n
t
r
o
l
,
”
A
n
n
u
a
l
R
e
v
i
e
w
o
f
C
o
n
t
ro
l
,
R
o
b
o
t
i
c
s
,
a
n
d
Au
t
o
n
o
m
o
u
s
S
y
s
t
e
m
s
,
v
o
l
.
3
,
n
o
.
1
,
p
p
.
2
6
9
–
2
9
6
,
M
a
y
2
0
2
0
,
d
o
i
:
1
0
.
1
1
4
6
/
a
n
n
u
r
e
v
-
c
o
n
t
r
o
l
-
0
9
0
4
1
9
-
0
7
5
6
2
5
.
[
3
]
J.
T
a
n
g
e
t
a
l
.
,
“
G
M
P
C
:
G
e
o
m
e
t
r
i
c
M
o
d
e
l
P
r
e
d
i
c
t
i
v
e
C
o
n
t
r
o
l
f
o
r
W
h
e
e
l
e
d
M
o
b
i
l
e
R
o
b
o
t
T
r
a
j
e
c
t
o
r
y
T
r
a
c
k
i
n
g
,
”
I
EEE
R
o
b
o
t
i
c
s
a
n
d
Au
t
o
m
a
t
i
o
n
L
e
t
t
e
rs
,
v
o
l
.
9
,
n
o
.
5
,
p
p
.
4
8
2
2
–
4
8
2
9
,
M
a
y
2
0
2
4
,
d
o
i
:
1
0
.
1
1
0
9
/
L
R
A
.
2
0
2
4
.
3
3
8
1
0
8
8
.
[
4
]
J.
B
.
R
a
w
l
i
n
g
s,
D
.
Q
.
M
a
y
n
e
,
a
n
d
M
.
M
.
D
i
e
h
l
,
M
o
d
e
l
p
re
d
i
c
t
i
v
e
c
o
n
t
ro
l
:
T
h
e
o
ry,
C
o
m
p
u
t
a
t
i
o
n
,
a
n
d
D
e
si
g
n
,
2
n
d
e
d
.
S
a
n
t
a
B
a
r
b
a
r
a
,
C
A
:
N
o
b
H
i
l
l
P
u
b
l
i
s
h
i
n
g
,
2
0
1
7
.
[
5
]
T
.
P
.
N
a
sc
i
me
n
t
o
,
C
.
E.
T
.
D
ó
r
e
a
,
a
n
d
L
.
M
.
G
.
G
o
n
ç
a
l
v
e
s,
“
N
o
n
l
i
n
e
a
r
mo
d
e
l
p
r
e
d
i
c
t
i
v
e
c
o
n
t
r
o
l
f
o
r
t
r
a
j
e
c
t
o
r
y
t
r
a
c
k
i
n
g
o
f
n
o
n
h
o
l
o
n
o
mi
c
mo
b
i
l
e
r
o
b
o
t
s:
A
mo
d
i
f
i
e
d
a
p
p
r
o
a
c
h
,
”
I
n
t
e
r
n
a
t
i
o
n
a
l
J
o
u
r
n
a
l
o
f
A
d
v
a
n
c
e
d
R
o
b
o
t
i
c
S
y
st
e
m
s
,
v
o
l
.
1
5
,
n
o
.
1
,
Ja
n
.
2
0
1
8
,
d
o
i
:
1
0
.
1
1
7
7
/
1
7
2
9
8
8
1
4
1
8
7
6
0
4
6
1
.
[
6
]
S
.
M
a
c
e
n
s
k
i
,
T
.
F
o
o
t
e
,
B
.
G
e
r
k
e
y
,
C
.
L
a
l
a
n
c
e
t
t
e
,
a
n
d
W
.
W
o
o
d
a
l
l
,
“
R
o
b
o
t
O
p
e
r
a
t
i
n
g
S
y
st
e
m
2
:
D
e
si
g
n
,
a
r
c
h
i
t
e
c
t
u
r
e
,
a
n
d
u
se
s
i
n
t
h
e
w
i
l
d
,
”
S
c
i
e
n
c
e
R
o
b
o
t
i
c
s
,
v
o
l
.
7
,
n
o
.
6
6
,
M
a
y
2
0
2
2
,
d
o
i
:
1
0
.
1
1
2
6
/
sci
r
o
b
o
t
i
c
s.
a
b
m
6
0
7
4
.
[
7
]
R
.
V
e
r
sch
u
e
r
e
n
e
t
a
l
.
,
“
A
c
a
d
o
s
—
a
M
o
d
u
l
a
r
O
p
e
n
-
S
o
u
r
c
e
F
r
a
me
w
o
r
k
f
o
r
F
a
st
Emb
e
d
d
e
d
O
p
t
i
mal
C
o
n
t
r
o
l
,
”
M
a
t
h
e
m
a
t
i
c
a
l
Pro
g
ra
m
m
i
n
g
C
o
m
p
u
t
a
t
i
o
n
,
v
o
l
.
1
4
,
n
o
.
1
,
p
p
.
1
4
7
–
1
8
3
,
M
a
r
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
0
7
/
s
1
2
5
3
2
-
0
2
1
-
0
0
2
0
8
-
8.
[
8
]
M
.
K
e
l
l
y
,
“
A
n
i
n
t
r
o
d
u
c
t
i
o
n
t
o
t
r
a
j
e
c
t
o
r
y
o
p
t
i
m
i
z
a
t
i
o
n
:
H
o
w
t
o
d
o
y
o
u
r
o
w
n
d
i
r
e
c
t
c
o
l
l
o
c
a
t
i
o
n
∗
,
”
S
I
AM
R
e
v
i
e
w
,
v
o
l
.
5
9
,
n
o
.
4
,
p
p
.
Evaluation Warning : The document was created with Spire.PDF for Python.
T
E
L
KOM
NI
K
A
T
elec
o
m
m
u
n
C
o
m
p
u
t E
l
C
o
n
tr
o
l
Hyb
r
id
n
eu
r
a
l n
etw
o
r
k
-
a
u
g
me
n
ted
mo
d
el
p
r
ed
ictive
co
n
tr
o
l
fo
r
…
(
Ta
o
u
fik
B
elke
b
ir
)
1777
8
4
9
–
9
0
4
,
J
a
n
.
2
0
1
7
,
d
o
i
:
1
0
.
1
1
3
7
/
1
6
M
1
0
6
2
5
6
9
.
[
9
]
H
.
B
e
l
k
e
b
i
r
a
n
d
T
.
B
e
l
k
e
b
i
r
,
“
En
h
a
n
c
e
d
t
r
a
j
e
c
t
o
r
y
t
r
a
c
k
i
n
g
f
o
r
a
u
t
o
n
o
mo
u
s
v
e
h
i
c
l
e
s
u
s
i
n
g
a
n
o
n
l
i
n
e
a
r
mo
d
e
l
p
r
e
d
i
c
t
i
v
e
c
o
n
t
r
o
l
l
e
r
,
”
T
u
rki
s
h
J
o
u
r
n
a
l
o
f
E
l
e
c
t
r
o
m
e
c
h
a
n
i
c
s
&
En
e
rg
y
,
v
o
l
.
9
,
n
o
.
3
,
p
p
.
1
1
4
–
1
2
3
,
2
0
2
4
.
[
1
0
]
T
.
B
e
l
k
e
b
i
r
,
H
.
B
e
l
k
e
b
i
r
,
a
n
d
A
.
M
a
n
so
u
r
i
,
“
R
e
a
l
-
T
i
me
Emb
e
d
d
e
d
N
M
P
C
f
o
r
A
u
t
o
n
o
mo
u
s
V
e
h
i
c
l
e
P
a
t
h
T
r
a
c
k
i
n
g
w
i
t
h
C
u
r
v
a
t
u
r
e
-
A
w
a
r
e
S
p
e
e
d
A
d
a
p
t
a
t
i
o
n
a
n
d
S
e
n
si
t
i
v
i
t
y
A
n
a
l
y
si
s,”
Au
t
o
m
a
t
i
o
n
,
v
o
l
.
7
,
n
o
.
2
,
2
0
2
6
,
d
o
i
:
1
0
.
3
3
9
0
/
a
u
t
o
mat
i
o
n
7
0
2
0
0
4
4
.
[
1
1
]
B
.
P
a
d
e
n
,
M
.
Č
á
p
,
S
.
Z
.
Y
o
n
g
,
D
.
Y
e
r
sh
o
v
,
a
n
d
E.
F
r
a
z
z
o
l
i
,
“
A
su
r
v
e
y
o
f
mo
t
i
o
n
p
l
a
n
n
i
n
g
a
n
d
c
o
n
t
r
o
l
t
e
c
h
n
i
q
u
e
s
f
o
r
s
e
l
f
-
d
r
i
v
i
n
g
u
r
b
a
n
v
e
h
i
c
l
e
s,”
I
E
EE
T
r
a
n
s
a
c
t
i
o
n
s
o
n
I
n
t
e
l
l
i
g
e
n
t
Ve
h
i
c
l
e
s
,
v
o
l
.
1
,
n
o
.
1
,
p
p
.
3
3
–
5
5
,
M
a
r
.
2
0
1
6
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
V
.
2
0
1
6
.
2
5
7
8
7
0
6
.
[
1
2
]
L
.
C
h
e
n
e
t
a
l
.
,
“
M
i
l
e
st
o
n
e
s
i
n
A
u
t
o
n
o
mo
u
s
D
r
i
v
i
n
g
a
n
d
I
n
t
e
l
l
i
g
e
n
t
V
e
h
i
c
l
e
s:
S
u
r
v
e
y
o
f
S
u
r
v
e
y
s,”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
I
n
t
e
l
l
i
g
e
n
t
Ve
h
i
c
l
e
s
,
v
o
l
.
8
,
n
o
.
2
,
p
p
.
1
0
4
6
–
1
0
5
6
,
F
e
b
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
TI
V
.
2
0
2
2
.
3
2
2
3
1
3
1
.
[
1
3
]
U
.
R
o
so
l
i
a
a
n
d
F
.
B
o
r
r
e
l
l
i
,
“
L
e
a
r
n
i
n
g
mo
d
e
l
p
r
e
d
i
c
t
i
v
e
c
o
n
t
r
o
l
f
o
r
i
t
e
r
a
t
i
v
e
t
a
s
k
s.
A
d
a
t
a
-
d
r
i
v
e
n
c
o
n
t
r
o
l
f
r
a
me
w
o
r
k
,
”
I
EEE
T
ra
n
s
a
c
t
i
o
n
s
o
n
A
u
t
o
m
a
t
i
c
C
o
n
t
r
o
l
,
v
o
l
.
6
3
,
n
o
.
7
,
p
p
.
1
8
8
3
–
1
8
9
6
,
Ju
l
.
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
T
A
C
.
2
0
1
7
.
2
7
5
3
4
6
0
.
[
1
4
]
M
.
H
e
r
t
n
e
c
k
,
J
.
K
o
h
l
e
r
,
S
.
T
r
i
mp
e
,
a
n
d
F
.
A
l
l
g
o
w
e
r
,
“
L
e
a
r
n
i
n
g
a
n
A
p
p
r
o
x
i
ma
t
e
M
o
d
e
l
P
r
e
d
i
c
t
i
v
e
C
o
n
t
r
o
l
l
e
r
w
i
t
h
G
u
a
r
a
n
t
e
e
s,”
I
EEE
C
o
n
t
r
o
l
S
y
st
e
m
s Le
t
t
e
rs
,
v
o
l
.
2
,
n
o
.
3
,
p
p
.
5
4
3
–
5
4
8
,
Ju
l
.
2
0
1
8
,
d
o
i
:
1
0
.
1
1
0
9
/
L
C
S
Y
S
.
2
0
1
8
.
2
8
4
3
6
8
2
.
[
1
5
]
B
.
K
a
r
g
a
n
d
S
.
L
u
c
i
a
,
“
Ef
f
i
c
i
e
n
t
R
e
p
r
e
se
n
t
a
t
i
o
n
a
n
d
A
p
p
r
o
x
i
ma
t
i
o
n
o
f
M
o
d
e
l
P
r
e
d
i
c
t
i
v
e
C
o
n
t
r
o
l
L
a
w
s
v
i
a
D
e
e
p
L
e
a
r
n
i
n
g
,
”
I
E
EE
T
ra
n
s
a
c
t
i
o
n
s
o
n
C
y
b
e
rn
e
t
i
c
s
,
v
o
l
.
5
0
,
n
o
.
9
,
p
p
.
3
8
6
6
–
3
8
7
8
,
S
e
p
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
T
C
Y
B
.
2
0
2
0
.
2
9
9
9
5
5
6
.
[
1
6
]
J.
D
r
g
o
ň
a
,
K
.
K
i
š,
A
.
T
u
o
r
,
D
.
V
r
a
b
i
e
,
a
n
d
M
.
K
l
a
u
č
o
,
“
D
i
f
f
e
r
e
n
t
i
a
b
l
e
p
r
e
d
i
c
t
i
v
e
c
o
n
t
r
o
l
:
D
e
e
p
l
e
a
r
n
i
n
g
a
l
t
e
r
n
a
t
i
v
e
t
o
e
x
p
l
i
c
it
mo
d
e
l
p
r
e
d
i
c
t
i
v
e
c
o
n
t
r
o
l
f
o
r
u
n
k
n
o
w
n
n
o
n
l
i
n
e
a
r
sy
st
e
ms,”
J
o
u
rn
a
l
o
f
P
r
o
c
e
ss
C
o
n
t
ro
l
,
v
o
l
.
1
1
6
,
p
p
.
8
0
–
9
2
,
A
u
g
.
2
0
2
2
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
p
r
o
c
o
n
t
.
2
0
2
2
.
0
6
.
0
0
1
.
[
1
7
]
K
.
P
.
W
a
b
e
r
si
c
h
a
n
d
M
.
N
.
Z
e
i
l
i
n
g
e
r
,
“
A
p
r
e
d
i
c
t
i
v
e
safe
t
y
f
i
l
t
e
r
f
o
r
l
e
a
r
n
i
n
g
-
b
a
se
d
c
o
n
t
r
o
l
o
f
c
o
n
st
r
a
i
n
e
d
n
o
n
l
i
n
e
a
r
d
y
n
a
mi
c
a
l
sy
st
e
ms,”
Au
t
o
m
a
t
i
c
a
,
v
o
l
.
1
2
9
,
p
.
1
0
9
5
9
7
,
J
u
l
.
2
0
2
1
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
a
u
t
o
m
a
t
i
c
a
.
2
0
2
1
.
1
0
9
5
9
7
.
[
1
8
]
P
.
K
e
r
sch
k
e
a
n
d
H
.
T
r
a
u
t
m
a
n
n
,
“
A
u
t
o
mat
e
d
a
l
g
o
r
i
t
h
m
se
l
e
c
t
i
o
n
o
n
c
o
n
t
i
n
u
o
u
s
b
l
a
c
k
-
b
o
x
p
r
o
b
l
e
ms
b
y
c
o
mb
i
n
i
n
g
e
x
p
l
o
r
a
t
o
r
y
l
a
n
d
s
c
a
p
e
a
n
a
l
y
si
s
a
n
d
m
a
c
h
i
n
e
l
e
a
r
n
i
n
g
,
”
E
v
o
l
u
t
i
o
n
a
ry
C
o
m
p
u
t
a
t
i
o
n
,
v
o
l
.
2
7
,
n
o
.
1
,
p
p
.
9
9
–
1
2
7
,
2
0
1
8
,
d
o
i
:
1
0
.
1
1
6
2
/
e
v
c
o
_
a
_
0
0
2
3
6
.
[
1
9
]
T
.
S
a
l
z
man
n
,
E
.
K
a
u
f
ma
n
n
,
J.
A
r
r
i
z
a
b
a
l
a
g
a
,
M
.
P
a
v
o
n
e
,
D
.
S
c
a
r
a
m
u
z
z
a
,
a
n
d
M
.
R
y
l
l
,
“
R
e
a
l
-
T
i
me
N
e
u
r
a
l
M
P
C
:
D
e
e
p
L
e
a
r
n
i
n
g
M
o
d
e
l
P
r
e
d
i
c
t
i
v
e
C
o
n
t
r
o
l
f
o
r
Q
u
a
d
r
o
t
o
r
s
a
n
d
A
g
i
l
e
R
o
b
o
t
i
c
P
l
a
t
f
o
r
ms,”
I
E
EE
R
o
b
o
t
i
c
s
a
n
d
A
u
t
o
m
a
t
i
o
n
L
e
t
t
e
rs
,
v
o
l
.
8
,
n
o
.
4
,
p
p
.
2
3
9
7
–
2
4
0
4
,
A
p
r
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
L
R
A
.
2
0
2
3
.
3
2
4
6
8
3
9
.
[
2
0
]
G
.
To
r
r
e
n
t
e
,
E.
K
a
u
f
ma
n
n
,
P
.
F
o
h
n
,
a
n
d
D
.
S
c
a
r
a
mu
z
z
a
,
“
D
a
t
a
-
D
r
i
v
e
n
M
P
C
f
o
r
Q
u
a
d
r
o
t
o
r
s,”
I
EEE
Ro
b
o
t
i
c
s
a
n
d
Au
t
o
m
a
t
i
o
n
L
e
t
t
e
rs
,
v
o
l
.
6
,
n
o
.
2
,
p
p
.
3
7
6
9
–
3
7
7
6
,
A
p
r
.
2
0
2
1
,
d
o
i
:
1
0
.
1
1
0
9
/
L
R
A
.
2
0
2
1
.
3
0
6
1
3
0
7
.
[
2
1
]
J.
K
a
b
z
a
n
,
L
.
H
e
w
i
n
g
,
A
.
L
i
n
i
g
e
r
,
a
n
d
M
.
N
.
Z
e
i
l
i
n
g
e
r
,
“
L
e
a
r
n
i
n
g
-
B
a
se
d
M
o
d
e
l
P
r
e
d
i
c
t
i
v
e
C
o
n
t
r
o
l
f
o
r
A
u
t
o
n
o
mo
u
s
R
a
c
i
n
g
,
”
I
EEE
R
o
b
o
t
i
c
s
a
n
d
Au
t
o
m
a
t
i
o
n
L
e
t
t
e
r
s
,
v
o
l
.
4
,
n
o
.
4
,
p
p
.
3
3
6
3
–
3
3
7
0
,
O
c
t
.
2
0
1
9
,
d
o
i
:
1
0
.
1
1
0
9
/
L
R
A
.
2
0
1
9
.
2
9
2
6
6
7
7
.
[
2
2
]
M
.
I
mad
,
O
.
D
o
u
k
h
i
,
D
.
J.
L
e
e
,
J.
c
h
u
l
K
i
m,
a
n
d
Y
.
J.
K
i
m,
“
D
e
e
p
L
e
a
r
n
i
n
g
-
B
a
se
d
N
M
P
C
f
o
r
L
o
c
a
l
M
o
t
i
o
n
P
l
a
n
n
i
n
g
o
f
L
a
st
-
M
i
l
e
D
e
l
i
v
e
r
y
R
o
b
o
t
,
”
S
e
n
s
o
rs
,
v
o
l
.
2
2
,
n
o
.
2
1
,
p
.
8
1
0
1
,
O
c
t
.
2
0
2
2
,
d
o
i
:
1
0
.
3
3
9
0
/
s
2
2
2
1
8
1
0
1
.
[
2
3
]
J.
N
u
b
e
r
t
,
J.
K
ö
h
l
e
r
,
V
.
B
e
r
e
n
z
,
F
.
A
l
l
g
ö
w
e
r
,
a
n
d
S
.
T
r
i
mp
e
,
“
S
a
f
e
a
n
d
F
a
st
T
r
a
c
k
i
n
g
o
n
a
R
o
b
o
t
M
a
n
i
p
u
l
a
t
o
r
:
R
o
b
u
st
M
P
C
a
n
d
N
e
u
r
a
l
N
e
t
w
o
r
k
C
o
n
t
r
o
l
,
”
I
EEE
Ro
b
o
t
i
c
s
a
n
d
Au
t
o
m
a
t
i
o
n
L
e
t
t
e
rs
,
v
o
l
.
5
,
n
o
.
2
,
p
p
.
3
0
5
0
–
3
0
5
7
,
A
p
r
.
2
0
2
0
,
d
o
i
:
1
0
.
1
1
0
9
/
L
R
A
.
2
0
2
0
.
2
9
7
5
7
2
7
.
[
2
4
]
E.
A
r
c
a
r
i
e
t
a
l
.
,
“
B
a
y
e
si
a
n
M
u
l
t
i
-
T
a
s
k
L
e
a
r
n
i
n
g
M
P
C
f
o
r
R
o
b
o
t
i
c
M
o
b
i
l
e
M
a
n
i
p
u
l
a
t
i
o
n
,
”
I
E
EE
R
o
b
o
t
i
c
s
a
n
d
A
u
t
o
m
a
t
i
o
n
L
e
t
t
e
rs
,
v
o
l
.
8
,
n
o
.
6
,
p
p
.
3
2
2
2
–
3
2
2
9
,
Ju
n
.
2
0
2
3
,
d
o
i
:
1
0
.
1
1
0
9
/
L
R
A
.
2
0
2
3
.
3
2
6
4
7
5
8
.
[
2
5
]
M
.
R
o
se
n
f
e
l
d
e
r
,
L
.
B
o
l
d
,
H
.
Es
c
h
man
n
,
P
.
Eb
e
r
h
a
r
d
,
K
.
W
o
r
t
h
man
n
,
a
n
d
H
.
Eb
e
l
,
“
D
a
t
a
-
d
r
i
v
e
n
p
r
e
d
i
c
t
i
v
e
c
o
n
t
r
o
l
o
f
n
o
n
h
o
l
o
n
o
mi
c
r
o
b
o
t
s
b
a
se
d
o
n
a
b
i
l
i
n
e
a
r
K
o
o
p
man
r
e
a
l
i
z
a
t
i
o
n
:
D
a
t
a
d
o
e
s
n
o
t
r
e
p
l
a
c
e
g
e
o
me
t
r
y
,
”
Ro
b
o
t
i
c
s
a
n
d
Au
t
o
n
o
m
o
u
s
S
y
s
t
e
m
s
,
v
o
l
.
1
9
4
,
p
.
1
0
5
1
5
6
,
D
e
c
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
r
o
b
o
t
.
2
0
2
5
.
1
0
5
1
5
6
.
[
2
6
]
A
.
C
a
u
l
i
g
i
,
P
.
C
u
l
b
e
r
t
so
n
,
E
.
S
c
h
me
r
l
i
n
g
,
M
.
S
c
h
w
a
g
e
r
,
B
.
S
t
e
l
l
a
t
o
,
a
n
d
M
.
P
a
v
o
n
e
,
“
C
o
C
o
:
O
n
l
i
n
e
M
i
x
e
d
-
I
n
t
e
g
e
r
C
o
n
t
r
o
l
V
i
a
S
u
p
e
r
v
i
se
d
L
e
a
r
n
i
n
g
,
”
I
E
EE
R
o
b
o
t
i
c
s
a
n
d
A
u
t
o
m
a
t
i
o
n
L
e
t
t
e
rs
,
v
o
l
.
7
,
n
o
.
2
,
p
p
.
1
4
4
7
–
1
4
5
4
,
A
p
r
.
2
0
2
2
,
d
o
i
:
1
0
.
1
1
0
9
/
L
R
A
.
2
0
2
1
.
3
1
3
5
9
3
1
.
[
2
7
]
J.
A
.
E.
A
n
d
e
r
sso
n
,
J.
G
i
l
l
i
s,
G
.
H
o
r
n
,
J.
B
.
R
a
w
l
i
n
g
s,
a
n
d
M
.
D
i
e
h
l
,
“
C
a
s
A
D
i
:
a
so
f
t
w
a
r
e
f
r
a
mew
o
r
k
f
o
r
n
o
n
l
i
n
e
a
r
o
p
t
i
mi
z
a
t
i
o
n
a
n
d
o
p
t
i
mal
c
o
n
t
r
o
l
,
”
Ma
t
h
e
m
a
t
i
c
a
l
Pro
g
ra
m
m
i
n
g
C
o
m
p
u
t
a
t
i
o
n
,
v
o
l
.
1
1
,
n
o
.
1
,
p
p
.
1
–
3
6
,
M
a
r
.
2
0
1
9
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
2
5
3
2
-
018
-
0
1
3
9
-
4.
[
2
8
]
R
.
Q
u
i
r
y
n
e
n
,
S
.
G
r
o
s,
B
.
H
o
u
sk
a
,
a
n
d
M
.
D
i
e
h
l
,
“
L
i
f
t
e
d
c
o
l
l
o
c
a
t
i
o
n
i
n
t
e
g
r
a
t
o
r
s
f
o
r
d
i
r
e
c
t
o
p
t
i
m
a
l
c
o
n
t
r
o
l
i
n
A
C
A
D
O
t
o
o
l
k
i
t
,
”
Ma
t
h
e
m
a
t
i
c
a
l
Pr
o
g
ra
m
m
i
n
g
C
o
m
p
u
t
a
t
i
o
n
,
v
o
l
.
9
,
n
o
.
4
,
p
p
.
5
2
7
–
5
7
1
,
D
e
c
.
2
0
1
7
,
d
o
i
:
1
0
.
1
0
0
7
/
s1
2
5
3
2
-
0
1
7
-
0
1
1
9
-
0.
[
2
9
]
M
.
R
a
i
ssi
,
P
.
P
e
r
d
i
k
a
r
i
s
,
a
n
d
G
.
E.
K
a
r
n
i
a
d
a
k
i
s,
“
P
h
y
si
c
s
-
i
n
f
o
r
me
d
n
e
u
r
a
l
n
e
t
w
o
r
k
s:
A
d
e
e
p
l
e
a
r
n
i
n
g
f
r
a
me
w
o
r
k
f
o
r
so
l
v
i
n
g
f
o
r
w
a
r
d
a
n
d
i
n
v
e
r
se
p
r
o
b
l
e
ms
i
n
v
o
l
v
i
n
g
n
o
n
l
i
n
e
a
r
p
a
r
t
i
a
l
d
i
f
f
e
r
e
n
t
i
a
l
e
q
u
a
t
i
o
n
s,”
J
o
u
r
n
a
l
o
f
C
o
m
p
u
t
a
t
i
o
n
a
l
Ph
y
si
c
s
,
v
o
l
.
3
7
8
,
p
p
.
6
8
6
–
7
0
7
,
F
e
b
.
2
0
1
9
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
j
c
p
.
2
0
1
8
.
1
0
.
0
4
5
.
[
3
0
]
T
.
S
c
h
i
z
a
n
d
H
.
Eb
e
l
,
“
R
e
d
u
c
i
n
g
t
h
e
c
o
mm
u
n
i
c
a
t
i
o
n
o
f
d
i
st
r
i
b
u
t
e
d
m
o
d
e
l
p
r
e
d
i
c
t
i
v
e
c
o
n
t
r
o
l
:
A
u
t
o
e
n
c
o
d
e
r
s
a
n
d
f
o
r
mat
i
o
n
c
o
n
t
r
o
l
,
”
C
o
n
t
ro
l
E
n
g
i
n
e
e
ri
n
g
Pr
a
c
t
i
c
e
,
v
o
l
.
1
6
5
,
p
.
1
0
6
5
6
0
,
D
e
c
.
2
0
2
5
,
d
o
i
:
1
0
.
1
0
1
6
/
j
.
c
o
n
e
n
g
p
r
a
c
.
2
0
2
5
.
1
0
6
5
6
0
.
B
I
O
G
RAP
H
I
E
S O
F
AUTH
O
RS
Ta
o
u
fi
k
B
e
l
k
e
b
ir
re
c
e
iv
e
d
th
e
E
n
g
in
e
e
rin
g
d
e
g
re
e
in
E
lec
tro
n
ics
a
n
d
I
n
d
u
strial
C
o
m
p
u
ti
n
g
f
ro
m
th
e
Na
ti
o
n
a
l
S
c
h
o
o
l
o
f
A
p
p
li
e
d
S
c
ien
c
e
s
(ENS
A
),
Ou
j
d
a
,
M
o
r
o
c
c
o
,
in
2
0
1
2
,
a
n
d
t
h
e
M
a
ste
r’s
d
e
g
re
e
in
P
ro
jec
t
M
a
n
a
g
e
m
e
n
t
in
2
0
1
6
.
He
is
c
u
rre
n
tl
y
p
u
rsu
i
n
g
t
h
e
P
h
.
D.
d
e
g
re
e
w
it
h
th
e
L
RS
I
L
a
b
o
ra
t
o
ry
,
F
a
c
u
lt
y
o
f
S
c
ien
c
e
s
a
n
d
T
e
c
h
n
o
lo
g
y
(F
S
T
),
Un
iv
e
rsit
y
S
id
i
M
o
h
a
m
m
e
d
Be
n
A
b
d
e
ll
a
h
,
F
e
z
,
M
o
ro
c
c
o
.
His
re
se
a
rc
h
in
tere
sts
in
c
lu
d
e
o
p
ti
m
a
l
c
o
n
tro
l,
m
o
d
e
l
p
re
d
ictiv
e
c
o
n
tr
o
l,
re
i
n
f
o
rc
e
m
e
n
t
lea
rn
in
g
,
a
n
d
a
u
to
n
o
m
o
u
s
m
o
b
il
e
ro
b
o
ti
c
s.
He
c
a
n
b
e
c
o
n
tac
ted
a
t
e
m
a
il
:
b
e
lk
e
b
ir.
tao
u
f
ik
@u
sm
b
a
.
a
c
.
m
a
.
Evaluation Warning : The document was created with Spire.PDF for Python.