Int
ern
at
i
onal
Journ
al of
El
e
ctrical
an
d
Co
mput
er
En
gin
eeri
ng
(IJ
E
C
E)
Vo
l.
8
,
No.
6
,
D
ece
m
ber
201
8
, pp.
4310
~
43
20
IS
S
N:
20
88
-
8708
,
DOI: 10
.11
591/
ijece
.
v
8
i
6
.
pp
4310
-
43
20
4310
Journ
al h
om
e
page
:
http:
//
ia
es
core
.c
om/
journa
ls
/i
ndex.
ph
p/IJECE
Novel Fr
amewor
k for Na
vigati
on u
sing Enhanced
Fuzzy
Approach
with Slid
ing M
ode Con
trolle
r
T. Somesw
ari
1
,
An
il
Kum
ar
Ti
w
ari
2
, Nagaraj
R
.
3
1,3
Dep
ar
t
m
ent
of
Elec
tr
ical & Electr
oni
cs
Engg
,
T
he
Oxford Co
ll
e
ge
of Engin
ee
rin
g
,
Indi
a
2
Am
ity
Schoo
l
o
f
Engi
n
ee
ring
,
Am
ity
Scho
ol
of
E
ngine
er
ing, India
Art
ic
le
In
f
o
ABSTR
A
CT
Art
ic
le
history:
Re
cei
ved
A
pr
2
, 2
01
8
Re
vised
Ju
l
18
,
201
8
Accepte
d
Aug 1
, 2
01
8
The
re
li
ab
il
i
t
y
of
an
y
embedd
ed
navi
g
at
or
in
adva
nc
ed
veh
i
cul
ar
s
y
stem
depe
nds
upon
c
orre
ct
and
pre
c
i
se
informati
on
o
f
navi
g
at
ion
al
d
at
a
ca
p
ture
d
and
proc
essed
to
offe
r
trustworth
y
pa
th.
Aft
er
rev
ie
wing
the
exi
sti
ng
s
y
stem,
a
signifi
c
ant
trade
-
off
is
expl
o
r
ed
bet
wee
n
th
e
exi
sting
navi
g
ati
onal
s
y
stem
and
pre
sen
t
st
at
e
of
con
trol
l
er
design
on
var
ious
c
ase
s
tudi
es
and
appl
i
ca
t
ions.
Th
e
exi
sting
desig
n
of
cont
roll
er
s
y
stem
for
navi
g
at
ion
using
err
or
-
prone
GP
S/INS
dat
a
doe
sn’t
emphasize
on
slidi
ng
m
ode
cont
rol
le
r
.
Although,
the
r
e
has
bee
n
good
num
ber
of
studie
s
in
slidi
ng
m
ode
cont
roller,
it
is
le
ss
a
tt
emp
te
d
to
opt
imize
the
n
avi
ga
ti
ona
l
per
form
ance
of
a
v
ehi
c
le.
The
ref
or
e,
thi
s
pape
r
pr
ese
nts
a
novel
op
ti
m
iz
e
d
design
of
a
sl
idi
ng
m
ode
cont
roller
tha
t
c
an
be
eff
e
ct
iv
ely
depl
o
y
ed
on
ad
vanc
ed
navi
g
at
i
onal
s
y
s
te
m
.
The
stud
y
outc
o
m
e
was
found
t
o
offe
r
h
ighe
r
spee
d,
opti
m
a
l
co
ntrol
sign
a
l
,
and
lower
err
or
occ
ura
n
ce
s
to
pr
ove
that
propose
d
s
y
stem
offe
rs
r
el
i
abl
e
and
opti
m
iz
ed
navi
g
at
ion
al
serv
ices i
n
cont
r
ast to
ex
isti
n
g
s
y
s
te
m
.
Ke
yw
or
d:
Fu
zzy
Global
posit
io
ning syst
em
In
e
rtia
l na
vig
at
ion
syst
e
m
Nav
i
gation
Sli
din
g m
od
e c
on
t
ro
ll
er
Copyright
©
201
8
Instit
ut
e
o
f Ad
vanc
ed
Engi
n
ee
r
ing
and
S
cienc
e
.
Al
l
rights re
serv
ed
.
Corres
pond
in
g
Aut
h
or
:
T. Som
eswar
i,
Dep
a
rtm
ent
of
Ele
ct
rical
&
El
ect
ronics E
ngg,
The O
xford
C
ol
le
ge
of E
ng
i
ne
erin
g,
Ba
ng
al
ore,
India
.
Em
a
il
:
so
m
eswar
io
xford@
gma
il
.co
m
1.
INTROD
U
CTION
W
it
h
the
inc
re
ase
of
a
dv
a
nce
m
ent
in
com
mu
nicat
io
n
de
vi
ces,
there
is
a
n
eq
uiv
al
e
nt
in
crease
in
a
n
e
m
bed
de
d
nav
i
gation
de
vice
too
[
1]
,
[
2].
I
n
this
re
gards,
b
ot
h
the
inerti
al
na
vig
at
io
n
syst
em
(I
NS)
an
d
G
lob
a
l
Po
sit
io
ning
S
yst
e
m
(G
PS
)
are
c
onsidere
d
as
the
pri
m
e
act
or
s
f
or
offe
rin
g
a
n
eff
ect
ive
na
vi
gationa
l
serv
ic
es
[3
]
,
[4]
.
Also
,
fou
nd
that
m
or
e
rese
arch
es
wer
e
e
xi
st
fo
r
j
oi
nt
i
m
plem
entat
ion
of
both
I
NS
a
s
well
a
s
GP
S
.
A
cl
os
er
look
i
nto
su
c
h
nav
i
gationa
l
syst
e
m
sh
ows
t
ha
t
na
vig
at
io
nal
i
nfor
m
at
ion
off
ered
by
I
NS
syst
e
m
is
po
te
ntial
ly
a
sso
ci
at
ed
with
error
s
[5
]
.
Alth
ough,
s
uch
e
rror
s
i
n
na
vig
at
i
on
al
in
f
or
m
at
i
on
of
I
NS
are
highly
sh
ort
term
ed
bu
t i
t spo
nta
neously
d
eg
rad
e
s o
ve
r
a p
eri
od of ti
m
e. O
n
the o
the
r
ha
nd, n
a
vig
at
io
nal info
rm
ation
offer
e
d
by
GPS
is
con
si
der
e
d
to
be
hi
gh
ly
sta
ble
al
ong
wit
h
bette
r
prov
isi
on
of
pr
eci
se
da
ta
relat
ed
to
sp
eed
,
locat
ion
,
a
nd
a
tt
it
ud
e
[6
]
.
It
was
al
so
f
ound
that
info
rm
at
i
on
r
el
at
ed
to
al
l
these
ph
ysi
cal
and
kin
et
ic
entit
ie
s
are
prov
i
ded by
I
NS w
hile i
nfo
rm
ation
relat
ed
to
only
loca
ti
on
a
nd s
peed
is give
n by
GPS.
Ther
e
is
al
so
a
po
te
ntial
de
pende
ncy
of
both
I
NS
as
G
PS
tog
et
her
as
an
eff
ect
i
ve
and
reli
able
nav
i
gationa
l
inf
or
m
at
ion
ca
nnot
be
offer
e
d
by
I
NS
sta
nd
al
on
e
an
d
the
m
ai
n
reas
on
beh
i
nd
it
is
m
or
e
occura
nces
of
errors
ass
ocia
te
d
with
inerti
al
sens
or
s
.
It
has
al
s
o
been
seen
t
hat
a
c
on
ti
nu
ous
stre
a
m
of
nav
i
gationa
l
da
ta
can
be
offe
red
by
I
NS
w
hi
le
data
strea
m
is
al
ways
interm
i
tt
ent
fo
r
G
PS
owin
g
to
va
rio
us
form
s
of
sig
na
l
interfere
nce
c
ausin
g
ou
ta
ge
[7
]
,
[8
]
.
The
refor
e
,
95%
of
th
e
existi
ng
rese
arch
-
ba
sed
w
ork
ha
s
consi
der
e
d
bot
h
INS
a
nd
G
P
S
f
or
e
vo
l
ving
up
with
a
bette
r
f
orm
of
nav
i
gationa
l
syst
em
.
Usag
e
of
suc
h
j
oi
nt
nav
i
gationa
l
syst
e
m
can
be
seen
in
dif
fer
e
nt
fo
rm
s
of
vehi
cl
e
m
ov
ing
on
la
nd
[
9],
wat
er
[10],
an
d
ai
r
[11].
Ther
e a
re also vari
ou
s
f
or
m
s o
f wea
pons
(
e
.
g.
m
issi
le
s)
whic
h
us
es s
uc
h j
oin
t na
vig
at
i
onal
syst
e
m
[
12
]
. Th
ere
is
no
de
nyin
g
t
he
fact
that
va
rio
us
resea
rc
he
s
wer
e
a
vaila
bl
e
reg
a
rd
i
ng
t
his
con
c
er
n,
bu
t
at
the
sam
e
tim
e,
i
t
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Novel
Fr
am
ew
or
k f
or
N
avig
at
ion
us
in
g
E
nh
an
ce
d F
uzzy
A
pp
r
oac
h
wi
th
Sl
iding
Mo
de
…
(
T.
So
meswa
ri
)
4311
cannot
be
ig
nored
that
existi
ng
resea
rch
a
ppro
ac
hes
we
re
not
seem
to
su
ccessf
ully
ben
c
hm
ark
or
cl
ai
m
ed
to
offer
s
uperi
or
nav
i
gationa
l
syst
e
m
[1
3].
E
xi
sti
ng
resea
rc
he
rs
are
al
s
o
at
t
e
m
pting
to
c
onstr
uct
cost
e
f
fecti
ve
INS
-
GP
S
na
vig
at
ion
al
syst
e
m
wh
ere
t
he
prom
inent
chall
eng
e
is
t
o
prec
ise
ly
init
ializ
e
the
ta
rg
et
a
ng
l
e
ow
i
ng
to
de
grade
d
w
orkin
g
process
of
gyros
or
sim
il
ar
fo
rm
s
of
oth
e
r
se
nsors
.
This
is
e
xtre
m
el
y
essenti
al
prob
le
m
s
to
be
ad
dress
ed
as
there
is
an
increasi
ng
trend
of
devel
op
in
g
rem
otely
dr
iven
ve
hicle
s.
He
nce,
wro
ng
inf
or
m
at
ion
relat
ed
to
nav
i
gation
syst
em
wil
l
render
a
c
ollat
eral
dam
age
on
s
uch
form
s
of
sm
art
veh
ic
le
.
O
ne
of
the
ef
fecti
ve
m
echa
nis
m
s
to
a
ddress
this
pro
blem
cou
ld
be
by
us
in
g
sli
ding
m
od
e
c
on
trolle
r
[14]
whic
h
is
the
hea
rt
of
ev
ery
ad
van
ce
co
ntr
ol
syst
e
m
.
Fo
r
e
ver
y
c
om
plex
an
d
a
dv
a
nc
ed
ve
hic
ular
s
yst
e
m
co
m
es
with
a
sm
art n
avigati
on that is ca
pa
ble of
us
in
g b
ot
h
G
PS
a
nd
INS r
el
at
ed
sig
na
ls.
Ther
e
has
bee
n
sig
nificant
num
ber
of
rese
arch
wor
k
bei
ng
ca
rr
ie
d
out
towa
rd
s
de
velop
i
ng
sli
ding
m
od
e
con
tr
ol
syst
e
m
;
ho
we
ve
r,
m
ajo
rity
of
them
are
asso
ci
at
ed
with
certai
n
inh
e
re
nt
pro
blem
s
of
de
al
ing
with
dynam
ic
syst
e
m
.
It
was
obser
ve
d
that
exi
sti
ng
a
ppr
oa
ches
do
e
sn’t
cat
er
up
the
un
certai
nty
dem
a
nd
s
of
the d
ynam
ic
sy
stem
esp
eci
al
ly in
transm
issi
on
n
et
work.
W
it
h
m
ajo
rity
o
f
t
he
w
ork
la
cki
ng b
e
nc
hm
ark
in
g
an
d
non
-
inclusi
on
of
c
on
si
der
i
ng
the
case
stud
y
of
s
olv
in
g
na
vi
gational
prob
le
m
,
the
sli
din
g
m
od
e
con
tr
oller
has
beco
m
e
one
of
the
le
ss
e
xplore
d
s
olu
ti
on
s
to
ward
e
nh
a
ncin
g
nav
i
gation
syst
em
us
ing
G
PS
a
nd
IN
S
.
Ther
e
f
or
e,
a
novel
te
c
hniq
ue
is
introd
uced
t
o
s
olv
e
the
na
vi
gational
pro
ble
m
s
in
G
PS
a
nd
IN
S
.
T
he
re
vi
ew
of
existi
ng
re
sea
rch
es
a
nd
c
orres
po
ndin
g
r
esearch
pro
ble
m
s
is
discuss
ed
in
Sect
i
on
-
2
an
d
Se
ct
ion
-
3
resp
ect
ively
.
S
ect
ion
-
4
hi
gh
li
gh
ts
a
bout
ad
opte
d
m
et
ho
dolog
y
as
s
olu
ti
on
al
ong
with
il
lustrati
on
of
al
gorithm
desig
n
is
giv
e
n
in
Sect
io
n
-
5.
The
perf
or
m
ance
analy
sis
is
di
scusse
d
in
Sec
ti
on
-
6
w
hile
S
ect
ion
-
7
outl
ines
the
su
m
m
ary of
th
e p
a
per as c
on
c
lusio
n.
1.1
.
B
ackgr
ound
At
pr
ese
nt,
the
re
are
va
rio
us
li
te
ratur
es
f
oc
usi
ng
o
n
desig
nin
g
a
n
eff
ect
i
ve
nav
igati
onal
s
yst
e
m
us
ing
IN
S
as
well
as
GP
S,
wit
h
un
iq
ue
m
et
h
od
s
a
nd
m
ec
han
ism
.
Existi
ng
stu
dies
ar
e
m
or
e
fo
c
use
d
on
inco
rpor
at
in
g
bette
r
preci
sio
n
w
her
e
the
st
ud
y
le
d
by
Da
cheng
et
al
[15]
prov
e
d
e
nhanced
se
ns
it
ivi
ty
by
m
ini
m
iz
ing
D
oppler
f
reque
nc
y
when
na
viga
ti
on
al
syst
em
of
INS
is
i
ntegr
at
e
d
with
G
PS.
A
doptio
n
strat
egy
of
filt
ers
is
on
e
un
i
qu
e
a
ppr
oa
ch
in
this
resea
rch
directi
on.
Accor
ding
to
Cho
et
al
.
[
16]
,
an
en
ha
nced
Kalm
an
filt
er
would
i
m
pr
ov
e
the
na
vig
at
io
nal
pe
rfor
m
ance
esp
eci
al
ly
fo
cusin
g
on
c
os
t
-
e
ff
e
ct
iveness
dep
l
oym
ent
view
po
i
nt.
De
plo
ym
ent o
f
sim
il
ar K
alm
an
filt
er o
f
exten
de
d
ty
pe
was
see
n
in the work
of Duon
g
an
d
N
gu
ye
n
[17]
for
an
ef
f
ect
ive
est
i
m
a
tio
n
of
physi
cal
nav
i
gationa
l
par
am
et
ers
e.g
.
at
ti
tud
e,
sp
ee
d,
locat
ion
,
et
c.
Usage
of
Kalm
an
filter
was
al
s
o
se
en
in
w
ork
dis
cusse
d
by
Fa
ng
an
d
G
ong
[18]
wh
e
re
the
pri
m
e
ob
j
e
ct
ive
was
to
inco
rp
or
at
e
a
predict
ive s
che
m
e o
f
a
n
i
nteg
r
at
ed
na
viga
ti
onal
ser
vice.
The
stu
dy
was
te
ste
d
ov
er
re
al
-
tim
e
senso
rs
m
ou
nted
on
a
ircraft
to
fin
d
that
offer
s
bette
r
accuracy
.
Ther
e
was
al
s
o
researc
h
w
ork
w
hich
say
s
that
dev
el
op
i
ng
na
vi
gationa
l
syst
e
m
fo
r
unde
r
wate
r
ve
hi
cl
e
is
reall
y
a
chall
eng
i
ng
one.
S
uc
h
r
esearc
h
is
f
ound
in
Lee
et
al
[19]
w
hich
is
m
eant
fo
r
m
i
nim
iz
ing
the
er
rors
i
n
nav
i
gationa
l
syst
e
m
especially
with
the
G
P
S
data.
T
he
w
ork
car
ried
out
by
Li
an
d
S
un
[
20]
has
al
s
o
use
d
Kalm
an
filt
er
of
e
xten
de
d
f
o
rm
fo
r
im
pr
ov
i
ng
t
he
acc
ur
ac
y
associat
ed
w
it
h
na
vig
at
io
na
l
data.
T
he
w
orks
of
Wu
et
al
[
21
]
ha
ve
f
oc
us
ed
on
de
velo
pi
ng
a
pr
e
dicti
ve
al
gorithm
us
ing
Kalm
an
filt
er.
T
he
stu
dy
al
so
form
ulate
s
a
n
oise
m
od
el
fo
r
com
pen
sat
ing
the
sign
al
s.
Yan
et
al
.[22
]
hav
e
al
s
o
ca
r
ried
out
w
ork
us
in
g
Kalm
an
filt
er .
Ca
se
stud
y
of
fly
ing
ob
j
ect
and
it
s
im
pr
ov
e
m
ent
of
na
vig
a
ti
on
data
wa
s
r
esearche
d
by
Nak
a
nis
hi
et
al
[2
3].
T
he
auth
or
s
ha
ve
addresse
d
the
prob
le
m
associat
ed
with
reli
abili
ty
of
data
from
GP
S
by
i
m
ple
m
enting
a
centra
li
zed
m
echan
ism
of
sensor
integ
rati
on
with
an
ai
d
of
updates
obta
ined
in
asy
nc
hr
onous
m
od
e
of
com
m
un
ic
at
ion
.
Ca
se
stud
y
of
na
vig
at
io
n
of
la
nd
veh
ic
le
has
been
co
ns
ide
r
ed
in
Qin
et
al
.
[2
4],
wh
e
re
the
Kal
m
an
filt
er
as
well
as
sta
nda
rd
fu
zzy
i
nf
e
r
ence
syst
em
f
or
m
ini
m
izing
error
s
i
n
navi
gation
syst
e
m
.
Anothe
r
a
uthor
viz.
Sun
et
al
.
[
25
]
hav
e
car
ried
out
an
ex
pe
rim
e
ntal
base
d
a
nal
ysi
s
wh
e
re
a
G
PSINS
base
d
na
vig
at
i
on
syst
em
. I
t was also
f
ound that co
up
li
ng a
m
on
g
the n
a
vi
gational
par
am
et
ers
sign
i
fican
t affect
th
e
detect
io
n
of
sig
nal.
T
his
i
ssu
e
has
bee
n
researc
he
d
by
Jam
a
l
[2
6]
by
recti
fyi
ng
t
he
e
rror
s
associat
e
d
wit
h
sensing
duri
ng
hig
h
s
pee
d
of
data
captur
i
ng.
Accor
ding
to
author,
ti
ght
couplin
g
res
ults
in
addressin
g
these
flaws.
Anothe
r
fo
rm
of
ap
pro
ach
was
th
e
usa
ge
of
m
achine
le
arn
in
g
sche
m
es
in
nav
igat
ion
al
syst
e
m
.
Usage
of
ne
ur
al
netw
ork
was
witne
ssed
i
n
the
w
ork
of
Ja
rad
at
a
nd
Ha
fez
[27]
wh
e
re
a
re
gr
es
sion
-
ba
sed
ap
proa
c
h
was
im
ple
m
ented
f
or
a
ddres
sing
the
com
m
un
ic
at
ion
del
ay
in
nav
igati
on
syst
em
.
Existi
ng
syst
e
m
has
al
so
witnesse
d
st
udy
towa
rd
s
ide
nt
ify
ing
a
nd
ad
dressi
ng
fau
lt
s
i
n
nav
i
gatio
nal
syst
e
m
.
The
w
ork
of
Xi
n
et
a
l.
[
28
]
has use
d
sta
ti
sti
cal
-
based a
ppr
oach f
or f
a
ult i
den
ti
ficat
io
n.
Althou
gh,
ther
e
are
var
i
ou
s
stud
ie
s
to
ward
s
i
m
pr
ov
i
ng
na
vig
at
io
nal
perform
ance,
bu
t
an
eff
ic
ie
nt
con
t
ro
ll
er
can
be
m
od
el
ed
us
in
g
sli
ding
m
od
e
design
appr
oach.
The
re
hav
e
been
var
i
ou
s
re
sear
ch
-
base
d
appr
oach
es
to
wards
sli
ding
m
od
e
con
tr
oller
[29]
-
[
28]
towa
rd
s
a
ddressi
ng
dif
fer
e
nt
pha
ses
of
rese
arc
h
-
base
d
pro
blem
s.
Usag
e
of
f
uzzy
l
ogic
on
sli
ding
m
od
e
co
ntro
ll
e
r
desig
n
is
car
ried
out
by
Ch
en
[3
0
]
c
on
si
de
rin
g
pro
blem
associ
at
ed
with
im
age
sta
bili
zat
ion
.
The
w
ork
of
R
ao
et
al
[
31
]
in
tro
du
ce
d
a
sli
di
ng
m
od
e
co
ntr
oller
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
20
88
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
4310
-
4320
4312
m
et
ho
d
by
co
ncen
t
rati
ng
on
the
perf
or
m
ance
enh
a
ncem
e
nt
of
di
strib
ute
d
po
wer
flo
w
con
t
ro
ll
er.
Sim
il
arly
,
Bou
a
rro
udj
et
al
[32]
an
d
Othm
an
et
al
[
33
]
,
hav
e
al
so
a
dopte
d
sli
ding
m
od
e
con
tr
olle
r
m
od
el
for
non
-
li
nea
r
syst
e
m
s
and
electro
-
hydr
o
ll
ic
act
uator
syst
em
to
achieve
bette
r
con
tr
ol
spe
ci
ficat
ion
s.
A
ccordin
g
to
D
o
[
34
]
identific
at
ion
of
non
-
li
near
di
sturb
a
nce
c
ould
i
m
pr
ove
the
con
tr
oller
des
ign
.
T
he
a
utho
rs
ha
ve
use
d
s
i
m
i
la
r
fu
zzy
-
base
d
sli
ding
m
od
e
co
nt
ro
ll
er
f
or
ad
dr
essing
t
he
c
hatte
ring
e
ff
ect
.
S
i
m
i
la
r
fu
zzy
-
ba
sed
a
ppr
oach
was
al
so
see
n for e
nh
a
n
ci
ng
powe
r qu
al
it
y.
The wo
rk of El
gam
m
al
an
d
El
-
na
ggar
[
35
]
has use
d
this a
ppr
oac
h
f
or
de
ve
lop
in
g
act
ive
powe
r
filt
er.
Si
m
il
ar
directi
on
of
work
ha
s
bee
n
see
n
i
n
the
work
of
F
ei
et
al
.
[
36
]
wh
ic
h
witness
ed
a
doptio
n
of
ne
ur
al
netw
ork
as
w
el
l
as
fu
zzy
based
sli
ding
appr
oach
w
as
us
e
d
for
m
inim
iz
ing
the
chatt
ering
ef
fect.
Stud
y
towa
rd
s
c
os
t
optim
iz
at
ion
us
ing
fuzzy
-
ba
sed
con
t
ro
ll
er
desi
gn
has
bee
n
pr
esented
by
Li
et
al
[
37
]
w
he
r
e
the
auth
or
has
ty
pi
cal
ly
us
ed
interval
ty
pe
-
2
f
uzzy
ap
proac
h
for
ad
dr
es
sin
g
delay
in
sy
stem
.
Si
m
i
la
r
fu
zzy
appr
oach
was
al
so
see
n
in
de
velo
ping
sli
di
ng
wind
ow
f
or
const
ruct
ing
hybri
d
c
on
t
ro
ll
er
desi
gn
with
m
or
e
e
m
ph
asi
s
on
powe
r
m
anag
e
m
ent
issues.
T
he
perform
ance
of
co
ntr
oller
desig
n
us
in
g
fu
zzy
-
base
d
sl
iding
window
was
pro
ve
n
to
be
e
n
han
ce
d
c
on
si
de
rin
g
case
stu
dy
of
ve
hicle
and
dr
i
ving
pr
ob
le
m
s.
The
work
of
Sh
e
n
et
al
[
38
]
has
use
d
a
da
ptive
m
echani
s
m
of
de
velo
ping
a
c
on
t
ro
l
le
r
syst
e
m
us
ing
sam
e
fu
zzy
-
bas
e
d
appr
oa
ch
on t
he
top of
gr
a
ph t
heory.
The
w
ork
ad
dresses
the
pro
blem
s
associat
ed
with
the
m
ulti
-
age
nt
syst
e
m
.
The
incorporati
on
of
adap
ti
ve
ness
i
s
al
so
seen
in
the
wor
k
of
Wang
an
d
Fei
[
39
]
w
here
th
e
el
i
m
inati
on
of
c
hatte
rin
g
ha
s
been
carried
out
al
on
g
with
ada
ptive
featu
re
of
sta
bili
ty
.
W
e
n
et
al
.
[4
0
]
,
[
41
]
hav
e
de
velo
pe
d
a
m
echan
is
m
wh
e
r
e
the
dynam
ic
featur
es
ha
ve
be
en
i
ncorpo
rated
as
well
a
s
th
ey
ha
ve
al
s
o
f
ocused
on
de
ve
lop
in
g
fa
ult
tolera
nt
act
uator
desi
gn
with
e
noug
h
a
dap
ti
ve
feat
ur
e
s.
A
nothe
r
un
i
qu
e
f
or
m
of
im
plem
entat
ion
ha
s
bee
n
obser
ve
d
i
n
the
w
ork
of
Yu
et
al
[
42
]
wh
e
re
a
b
io
-
i
ns
pi
red
al
gorit
hm
has
bee
n
desig
ne
d
f
or
be
tt
er
physi
cal
con
t
ro
l
syst
e
m
.
The
w
ork
car
ried
out
by
Zha
o
et
al
[
43
]
has
desig
ned
f
uz
zy
con
trolle
r
s
yst
e
m
con
side
r
ing
the
case
stud
y
of
cl
ose
d
l
oop
syst
e
m
associat
ed
wit
h
m
ulti
-
inp
ut
a
nd
m
ulti
-
ou
tp
ut
syst
e
m
.
Ap
a
r
t
from
this
the
re
a
re
al
so
ot
her
opti
m
iz
at
ion
-
base
d
appr
oach
e
s
to
wards
sli
di
ng
m
od
e
con
tr
oller
desi
gn
e.
g.
[
44
]
,
[
45
].
He
nce
,
ther
e
hav
e
bee
n
va
ri
ou
s
a
rc
hiv
es
of
li
te
ratur
es,
w
her
e
fu
zzy
-
bas
ed
inf
ere
nce
s
yst
e
m
has
been
f
ound
to
be
act
ively
us
e
d
in
enh
a
nc
ing
the
pe
rform
ance
of
sli
din
g
m
od
e
co
nverter
as
well
as
there
are
al
so
good
num
ber
of
w
or
k
bein
g
in
vestiga
te
d
to
wards
na
vig
at
io
nal
a
ppr
oach
e
s
to
o
i
n
m
os
t
recent
ti
m
es.
The
belo
w
sect
io
n
pro
vi
des
the
pro
blem
s ev
olv
ed
in
e
xisti
ng
syst
e
m
.
1.2
.
Th
e Pr
ob
l
em
Fr
om
the
pe
rspect
ive
of
t
he
a
dv
a
nce
d
na
vi
ga
ti
on
syst
em
,
bo
th I
N
S
& G
P
S
is
us
e
d
in p
r
esent
syst
em
.
At
pr
ese
nt
m
ajo
rity
of
the
ex
ist
ing
nav
i
gational
syst
em
has
dep
e
nd
e
ncy
on
GP
S
ho
we
ver
the
re
are
va
rio
us
flaws
in
us
in
g
GP
S
e.
g.
i)
abno
rm
al
ps
eu
do
-
ra
nge
er
ro
r
gen
e
rati
on
oc
curred
due
to
troposp
her
ic
delay
,
ionos
ph
e
res
de
la
y
any
m
any
m
or
e.
At
t
he
sam
e
t
i
m
e,
us
a
ge
of
I
NS
is
al
so
ass
oci
at
ed
with
e
rro
r
-
pro
ne
read
i
ng.
Fo
ll
owin
g
are
the
evide
nces
to
pr
ov
e
this
fact
about
issues
of
bo
t
h
INS
an
d
GP
S
.
The
w
ork
carri
e
d
ou
t by
C
hian
g
et
al
[
46
]
, Y
in et
al
[
47
]
,
Fil
ya
shkin
et
al
[
48
]
,
a
nd
Ha
n
et
al
[
49
]
are
so
m
e
of
the
stu
dies
w
here
nav
i
gationa
l
s
yst
e
m
has
bee
n
cl
os
el
y
stu
died,
how
eve
r,
there
are
s
om
e
sign
ific
a
nt
pro
blem
s
associat
ed
wit
h
al
l t
hese s
tu
die
s.
S
om
e o
f
the
po
te
ntial
p
r
oblem
s ar
e as
fo
ll
ow
s:
a.
Althou
gh
there
are
good
num
ber
of
w
ork
be
ing
ca
rr
ie
d
out
in
na
vig
at
io
nal
syst
e
m
as
well
as
f
uzzy
-
base
d
sli
din
g
m
od
e
con
t
ro
ll
er,
but
they
wer
e
ne
ver
bein
g
j
oi
ntly
stud
ie
d.
Fuzz
y
-
base
d
sli
di
ng
a
ppr
oac
h
ha
s
been ap
plied
on v
ari
ou
s
case
stud
ie
s
but n
ot
assessed
w
it
h r
espect t
o na
viga
ti
on
al
pr
ob
le
m
s.
b.
Existi
ng
na
viga
ti
on
al
-
base
d
appr
oach
do
e
s
n’
t
co
ns
i
der
c
om
plete
pr
oblem
s
of
na
vig
a
ti
on
al
data
th
at
cou
l
d
be
po
s
sibly
arr
i
ved
f
rom
e
it
her
I
NS
or
GPS.
It
is
not
cl
early
m
entioned
in
m
uch
of
t
he
e
xisti
ng
stud
ie
s
that
av
ai
la
bili
ty
of
G
PS
is
nev
e
r
confirm
ed
and
in
su
ch
sit
uati
on
us
er/
dev
ic
e
will
hav
e
on
l
y
i
m
par
ti
al
or
va
gu
e
nav
i
gationa
l
data
t
hat
wil
l
le
ad
to
e
rror
pro
ne
directi
on
al
vector
s.
E
xisti
ng
ap
proac
he
s
towa
rd
s
enha
nc
ing
preci
sio
n pro
blem
s d
oes
n’
t a
ddress
this
issue.
c.
Existi
ng
st
ud
i
es
towards
f
uz
zy
-
base
d
sli
di
ng
m
od
e
co
ntr
ollers
ha
ve
us
ed
sim
i
la
r
arch
it
ect
ur
e
o
f
dep
l
oym
ent
on
diff
ere
nt
case
stud
ie
s.
Howe
ver,
it
is
no
t
co
ns
ide
red
that
there
is
a
fair
chan
ce of
di
ff
e
r
ent
app
li
cat
io
ns
to
possess
di
ff
e
r
ent
re
quirem
ents
of
f
uzzy
str
uc
ture.
Mo
reov
er,
with
ou
t
im
pro
vem
ent
of
t
he
fu
zzy
c
ontrolle
r,
t
he
r
ule
-
base
d
a
ppr
oach can
no
t
offe
r bett
er
covera
ge
t
o
sc
al
abili
ty
f
act
or
.
d.
The
sce
na
rios
consi
der
e
d
in
e
xisti
ng
ap
proa
ches
a
re
near
ly
well
-
def
i
ne
d
with
in
volvem
ent
of
uncertai
nty
of
inf
orm
ation
from
IN
S
.
Althou
gh,
it
is
a
r
are
case
,
bu
t
it
will
le
ad
to
m
al
-
functi
on
of
t
he
na
vig
at
io
na
l
syst
e
m
that
is
diff
ic
ult
to
be
identifie
d.
He
nc
e,
e
xisti
ng
stu
dies
a
re
f
ound
no
t
t
o
fo
c
us
on
reli
abili
ty
of
the
inp
uts c
on
si
dered
for navi
gational
data.
Ther
e
f
or
e,
the
pri
m
e
research
pro
blem
is
that
in
a
bs
en
ce
of
GP
S
,
I
NS
sho
uld
of
fer
accu
rate
nav
i
gationa
l
s
erv
ic
es
with
a
n
e
ff
ect
ive
op
tim
iz
ed
ve
rsion
of
f
uzzy
-
ba
sed
sli
di
ng
m
od
e
co
ntr
oller.
This
pro
blem
is
direct
rep
re
s
entat
ion
of
dynam
ic
un
ce
rtai
nty
th
at
sh
ould
be
be
tt
er
so
lve
d
by
fu
zzy
lo
gic
in
vo
l
ve
d
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Novel
Fr
am
ew
or
k f
or
N
avig
at
ion
us
in
g
E
nh
an
ce
d F
uzzy
A
pp
r
oac
h
wi
th
Sl
iding
Mo
de
…
(
T.
So
meswa
ri
)
4313
in
sli
ding
m
od
e
c
on
t
ro
ll
er
in
ve
hicle
na
vig
at
io
nal
syst
e
m
.
The
rese
arch
m
et
hodo
l
og
y
im
ple
m
ented
t
o
ov
e
rc
om
e these proble
m
s is exp
la
ine
d wit
h
belo
w
sect
io
n.
2.
RESEA
R
CH MET
H
O
DOL
OGY
The
m
ai
n
inte
ntion
of
this
work
is
t
o
des
ign
a
fr
am
ework
of
a
novel
veh
ic
le
na
vig
a
ti
on
al
syst
em
us
in
g
fu
zzy
l
og
ic
an
d
sli
di
ng
m
od
e
co
ntro
ll
er.
Howe
ver, di
ff
ere
nt
f
r
om
e
xisti
ng
ap
proa
ch,
t
he
propose
d
stu
dy
will
choose
t
o
init
ia
ll
y
op
ti
m
iz
e
the
error
-
pro
ne
i
nputs
of
na
vig
at
io
na
l
data
f
ollow
e
d
by
ap
plyi
ng
fu
zzy
sli
din
g
m
od
e
con
t
ro
ll
er.
T
he
stud
y
fo
c
us
es
on
de
velo
ping
a
novel
coupled
I
NS
/G
PS
nav
igati
onal
syst
e
m
al
ong
with
m
i
nim
iz
at
ion
of
chatt
erin
g
pro
bl
e
m
wh
il
e
us
ing
sli
ding
m
od
e
con
tr
oller.
T
he
propos
e
d
syst
e
m
is
perform
ed
by a
naly
ti
cal
r
esear
ch
m
et
ho
dolo
gy
w
it
h fo
ll
owi
ng sc
hem
e as
sh
ow
n
in
Fig
ure
1.
P
r
o
b
l
e
m
I
d
e
n
t
i
f
i
c
a
t
i
o
n
d
y
n
a
m
i
c
u
n
c
e
r
t
a
i
n
t
y
i
n
I
N
S
p
r
o
b
l
e
m
f
o
r
m
u
l
a
t
i
o
n
E
n
h
a
n
c
e
m
e
n
t
F
u
z
z
y
L
o
g
i
c
A
p
p
l
y
L
e
a
r
n
i
n
g
P
r
o
c
e
s
s
A
d
d
S
e
l
f
-
A
d
a
p
t
i
v
e
F
e
a
t
u
r
e
N
o
v
e
l
F
u
z
z
y
C
o
n
t
r
o
l
l
e
r
D
e
s
i
g
n
S
l
i
d
i
n
g
M
o
d
e
C
o
n
t
r
o
l
l
e
r
Figure
1
.
Sc
he
m
at
ic
diagr
am
of 1
st m
et
ho
do
log
y
The
stu
dy
con
s
iders
dy
nam
ic
un
ce
rtai
nty
of
tim
e
interval
of
IN
S
w
he
n
inf
o
rm
at
ion
relat
ed
to
GPS
i
s
ei
ther
incom
plete
or
va
gu
e
.
T
his
dynam
ic
un
certai
nty
will
a
lso
giv
e
rise
to
m
axi
m
u
m
no
n
-
li
near
e
ven
t, wh
ic
h
sh
oul
d
be
s
up
pr
ess
ed
.
The
i
niti
al
par
t
of
the
stu
dy
will
fo
c
us
on
e
volvin
g
up
with
a
pro
blem
fo
rm
ula
ti
on
relat
ed
to
exis
ti
ng
nav
i
gatio
nal
syst
em
fo
ll
ow
e
d
by
de
sign
i
ng
an
en
ha
nced
f
uzzy
lo
gic
syst
em
in
order
t
o
handle
m
or
e
a
m
ou
nts
of
unc
ertai
nties.
T
his
pro
blem
is
m
or
e
or
le
ss
will
be
giv
e
n
the
sh
a
pe
of
c
ons
trai
nt
sat
isfact
ion
pr
ob
le
m
The
pro
po
s
ed
syst
em
dev
el
ops
a
n
e
nhance
d
f
uzzy
con
t
ro
ll
er
syst
e
m
,
wh
e
re
a
pa
rt
f
ro
m
conve
ntion
al
c
on
ce
pt
of
f
uzz
y
log
ic
,
it
will
inclu
de
desi
gnin
g
new
ca
pa
bili
ti
es.
These
capa
bili
ti
es
w
il
l
be
form
ed
m
a
them
at
ic
ally
con
s
iderin
g
sim
ple
op
ti
m
iz
at
ion
process
that
will
include
app
ly
in
g
the
le
arn
i
ng
proce
ss
m
entio
ne
d
by
T
hangar
a
j
[
5
0
]
.
T
he
stud
y
al
so
de
velo
p
f
ur
the
r
enh
a
nce
of
th
is
le
arn
ing
pro
cess
by
add
i
ng m
or
e se
lf
-
ada
ptive
f
eat
ur
e
that assist
s
in d
eci
si
on m
a
king.
The
pro
posed
s
tud
y
al
s
o
a
pp
li
es
sta
bili
ty
fun
ct
ion
in
or
der
t
o
asses
s
the
sta
bili
ty
.
The
ou
t
com
e
of
the
stud
y
is
ver
ifie
d
th
rou
gh
er
rors
in
t
he
sig
nal
ge
ner
at
e
d
by
I
NS
im
ple
m
entin
g
the
pro
po
se
d
al
gorithm
as
well
as
sp
ee
d
of
the
veh
ic
le
.
T
he
s
ign
ific
a
nt
co
n
trib
ution
of
t
he
pro
po
se
d
st
ud
y
are
i)
perform
ing
op
ti
m
iz
ation
on
nav
i
gationa
l
da
ta
befor
e
ap
pl
yi
ng
it
to
fu
zzy
sli
din
g
m
od
e
con
t
ro
ll
er,
ii
)
m
ake
the
desig
n
progressi
ve
a
nd
le
ss
it
erati
ve
by inc
orp
or
at
in
g
ada
ptive f
eat
ur
es
, iii
)
ensu
r
e b
et
te
r
res
pons
e ti
m
e
com
pu
ta
ti
on
a
ll
y i
n
or
de
r
to
assess
the
pract
ic
al
im
ple
m
entat
ion
scenari
o.
T
he
nex
t
sect
i
on
of
the
pa
pe
r
disc
us
ses
a
bout
co
re
al
gorithm
i
m
ple
m
entat
io
n
in
or
der
t
o
ac
hieve
t
he
a
bove
d
isc
us
se
d
re
s
earch
m
et
ho
dolog
y.
3.
IMPLEME
N
TATION
OF
ALGO
RITH
M
The
desig
ne
d
al
gorithm
is
util
iz
ed
to
de
vel
op
a
s
ophisti
cat
ed
an
d
ye
t
li
ght
wei
gh
t
c
ontrolle
r
syst
em
for
facil
it
at
ing
bette
r
na
vig
at
ion
al
syst
em
.
Howe
ver,
the
pro
po
se
d
al
gor
it
h
m
act
s
ver
y
dif
fer
e
ntly
f
rom
any
existi
ng
ap
pro
aches
a
s
in
ste
ad
of
dev
el
op
i
ng
t
he
sli
di
ng
m
od
e
c
on
tr
oller,
it
init
ia
ll
y
wo
r
ks
on
op
ti
m
iz
i
ng
th
e
nav
i
gationa
l i
nput.
T
he
al
go
rithm
f
or
this
pu
rpose
is
h
i
gh
li
gh
te
d
as
b
el
ow:
The
pro
po
se
d
syst
e
m
i
m
ple
ments
a
sim
ple
op
tim
iz
at
ion
te
chn
i
qu
e
w
her
e
the
m
ajo
r
ai
m
i
s
to
ob
ta
i
n
bette
r
form
of
structu
red
in
f
or
m
a
ti
on
abou
t
the
nav
igati
on
al
in
form
at
i
on.
He
nce,
be
fore
de
velo
ping
the
pro
po
se
d
form
to
ta
ke
th
e
s
hap
e
of
sli
ding
m
od
e
f
uzzy
con
t
ro
ll
er
syst
e
m
,
the
pro
posed
syst
em
per
f
or
m
s
necessa
ry
op
ti
m
iz
at
ion
.
The
first
phase
of
t
his
op
ti
m
iz
ation
process
is
al
l
about
in
it
ia
li
z
at
ion
proces
s.
Hen
ce
,
a
m
at
rix
H
a,b
is con
st
ru
ct
e
d
ra
ndom
ly
b
y ensu
ri
ng
t
hat there
is an
un
dev
ia
ti
ng
form
s o
f
da
ta
d
ist
ribu
ti
on
α
(for
offer
i
ng
m
or
e
predict
abili
ty
in
na
vig
at
io
na
l
data).
Hen
c
e,
f
or
al
l
the
n
it
erati
ons
(L
ine
-
1)
,
the
al
gorithm
ob
ta
in
s
the
m
i
nim
a
l
lim
it
s
of
the
m
at
rix
H
i.e.
H
m
in,b
wh
er
e
the
prefi
x
a
a
nd
b
re
pr
ese
nts
the
r
ow
an
d
c
olu
m
n
of
t
he
m
at
rix
(Line
-
2).
T
he
new
va
riable
ΔH
will
re
pr
e
sents
dot
product
of
an
ar
bitrary
ge
ner
at
io
n
of
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
20
88
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
4310
-
4320
4314
unde
viati
ng
f
orm
of
functi
on
and
dif
fer
e
nce
of
hig
he
r
an
d
lowe
r
bounds
of
m
at
rix
H
(Line
-
2)
.
T
he
pro
po
s
e
d
syst
e
m
u
ses pr
ob
a
bili
ty
in
this w
her
e
the
nu
m
erical
v
al
ue of
t
he unde
viati
ng
form
o
f fun
ct
ion
li
es in
the
r
an
ge
of 0 an
d 1.
Algori
th
m
for
Op
timi
z
ing Nav
i
gatio
na
l I
n
put
Inpu
t
: H
(navi
gational i
nput)
Out
p
ut
:
H
a,m
od
3
(eli
te
o
utc
ome
of
nav
i
gationa
l vecto
r)
St
ar
t
1.
For
i=
1:n
2.
H
a,b
H
m
in,
b
+ ΔH
3.
H
o
a,b
H
m
in_
m
ax
-
H
a,b
4.
If
α
1
< α
2
,
5.
H
old
H
6.
F
or
poli
cy
=1
7.
If
β
>0
8.
H
a,mod1
H
a,new
+ β
1
9.
El
se
10.
H
a,mod1
H
a,new
+ β
2
11.
End
12.
F
or
po
li
c
y=
2
13. H
a,m
od3
H
a,new
+ α
a
.c
14.
End
End
To
bri
ng
im
pr
ov
em
ent
in
the
converge
nce
s
peed
of
the
propose
d
proce
s
sing,
this
al
gor
it
h
m
fu
rthe
r
cl
assifi
es
the
nav
i
gationa
l
inputs
i
nto
cl
ust
ers.
For
t
his
pur
pose,
a
dy
nam
ic
syst
em
D
is
form
e
d
th
at
is
m
at
he
m
at
ic
a
lly represe
nted
(
1)
,
D
a,b, c+1
θ(π
D
c
),
c
=
1, 2,
3,
….
c
m
ax
(1)
In
t
he
a
bove
expressi
on,
th
e
va
riable
c
will
represe
nt
a
rec
ursive
f
un
ct
io
n
with
r
espect
to
any
trigon
om
et
ric
fu
nctio
n
θ
.
T
he
ab
ov
e
ex
press
ion
is
substi
tut
ed
f
or
unde
viati
ng
form
of
f
unct
ion
in
Δ
H
(
Line
-
2).
This
will
finall
y
le
ad
to
the
norm
al
izati
on
of
the
s
te
p
(Line
-
3)
to
re
pr
ese
nt
th
e
com
pu
te
value
of
nav
i
gationa
l
inp
ut
H
o
a,b.
The
var
ia
ble
H
m
in_
max
will
rep
rese
nts
bth
var
ia
ble
cor
re
spo
nd
i
ng
to
the
at
h
ar
ray
of
data
in
m
at
rix
with
nav
i
gatio
nal
data
.
T
he
nex
t
pa
rt
of
t
he
al
gorithm
i
m
plem
entat
ion
is
ab
out
f
or
m
ulati
ng
a
conditi
on
for
obta
inin
g
upgra
ded
data.
The
al
gorithm
us
es
a
new
vecto
r
H
old
for
perf
orm
ing
upgradat
ion
of
the
na
vi
gationa
l
data
in
eve
r
y
rou
nd.
T
he
process
of
i
niti
al
iz
at
ion
is
ve
ry
sim
il
ar
to
H
a
nd
the
pro
cess
of
upgradati
on
is carried
out
co
nsi
der
in
g
If
-
T
H
EN
wh
ic
h
is
now
te
rm
ed
as
poli
cy
of
u
pg
radi
ng
t
he
na
vig
at
ion
al
data f
or sli
ding
m
od
e f
uzzy co
ntr
oller (
Li
ne
-
4).
The
em
pirical
relat
ion
sh
i
p
be
tween
H
old
and
H
is
bound
by
(Line
-
5).
This
will
al
s
o
m
ean
that
pro
po
se
d
syst
e
m
of
fe
rs
a
buf
fer
s
pace
wh
e
r
e
the
sta
ti
c
inf
or
m
at
ion
of
th
e
nav
i
gatio
nal
data
is
retai
ne
d
unti
l
and
unle
ss
it
is
f
ound
to
be
sign
i
ficantl
y
cha
ng
e
d.
T
he
pro
po
s
ed
syst
em
perform
s
an
ar
bitrary
per
m
utati
on
of
the
num
ber
of
nav
i
gati
onal
in
pu
t
data
ass
oci
at
ed
with
H
old
in
or
der
t
o
e
di
t
the
or
der
to
al
l
the
unit
val
ue
of
nav
i
gationa
l
data
(Line
-
5)
.
A
ccordin
g
to
the
po
li
cy
fo
rm
at
i
on,
if
α
1
is
fo
und
to
be
le
ss
t
han
α
2
than
on
ly
the
upgradati
on
of
H
will
be
ta
ke
n
place
(Li
ne
-
5).
T
he
nex
t
pa
rt
of
the
stu
dy
will
furthe
r
perform
second
la
ye
r
of
op
ti
m
iz
ation
wh
e
re
a
n
ada
pt
ive
knowle
dg
e
-
acq
uisit
ion
proces
s
is
init
ia
te
d.
Acc
ordi
ng
to
this
proce
s
s,
the
pro
po
se
d
al
gor
it
h
m
will
be
fa
ci
li
ta
te
d
with
be
tt
er
form
of
dec
isi
on
.
F
or
t
hi
s
pur
pose,
the algo
rithm
fo
rm
ulate
s
two dif
fer
e
nt for
m
s o
f
poli
ci
es.
Fo
r
the
first
poli
cy
,
the
al
go
rithm
intro
duc
es
a
ne
w
var
ia
ble
β
wh
ic
h
is
com
par
ed
with
0
(
Line
-
7),
wh
e
re
the
c
omparis
on
res
ults
in
two
di
ff
e
re
nt
form
s
of
co
m
pu
ta
ti
on
of
H
m
at
rix
(Line
-
8
a
nd
Li
ne
-
10)
.
The
al
gorithm
us
es
β
1
and
β
2
to
r
epr
ese
nts
pro
duct
of
β
with
ol
d
resi
du
al
outc
om
e
of
na
viga
ti
on
al
data
(L
ine
-
8)
and
pro
duct
of
βwith
new
res
idu
al
outc
om
e
of
na
vig
at
io
nal
data
(Li
ne
-
10)
resp
ect
i
vely
.
The
stu
dy
c
onsider
s
H
a,m
od1
to
be
fir
st
el
it
e
ou
tc
om
e
on
th
e b
asi
s o
f
this po
li
cy
c
on
st
ru
ct
io
n.
F
or
the
sec
ond
po
li
cy
fo
rm
ulati
on
,
t
he
al
gorithm
i
m
p
lem
ents
an
em
pirical
exp
re
ssion
as
s
how
n
in
Line
-
13.
Accor
ding
to
this
fo
rm
ulati
on
,
the
cal
culat
ion
of
H
a,m
od3
m
at
rix
is
car
ried
out
by
add
i
ng
H
a,new
with
pro
duct
of
α
a
a
nd
c
,
w
he
re
the
var
ia
ble
c
will
represe
nt
a
re
s
idu
al
ou
tc
om
e
of
na
vig
at
io
nal
data.
I
n
order
to
ob
ta
in
resi
du
al
outc
om
e
of
na
vig
at
io
nal
data,
the
rec
ursive
f
un
ct
io
n
obta
in
s
the
el
it
e
ou
tc
om
e
and
fin
ds
it
s
aver
a
ge.
It
t
hen
s
ub
tr
act
s
t
he
el
it
e
outc
om
e
with
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Novel
Fr
am
ew
or
k f
or
N
avig
at
ion
us
in
g
E
nh
an
ce
d F
uzzy
A
pp
r
oac
h
wi
th
Sl
iding
Mo
de
…
(
T.
So
meswa
ri
)
4315
the
ave
ra
ge
sc
or
e
of
the
na
vi
gational
data
to
obta
in
resid
ual
na
vig
at
i
onal
data.
T
he
c
on
t
rib
ution
of
this
al
gorithm
is
th
at
i
m
ple
m
entation
of
first
poli
cy
resu
lt
s
in
bu
il
ding
bette
r
c
apab
il
it
ie
s
fo
r
bo
t
h
local
as
well
as
glo
bal
scal
e
of
op
ti
m
iz
a
ti
on
by
e
xtracti
ng
res
pecti
ve
ou
tc
om
es.
Howe
ver,
the
sec
ond
po
li
cy
assis
ts
in
i
m
pr
ovin
g
the
global
op
ti
m
i
zat
ion
m
or
e.
Finall
y,
the
be
st
poli
cy
can
be
obta
ined
dep
e
ndin
g
upon
the
ou
tc
om
es
ob
ta
ined,
w
hich
is
ob
se
r
ved
by
lowe
r
sc
or
e
of
er
r
or
s
i
n
na
viga
ti
on
al
data.
A
fter
this
pa
rt
of
the
op
ti
m
iz
ation
is
carried
out,
th
e
nex
t
par
t
of
the
w
ork
will
be
relat
ed
to
ap
plyi
ng
this
op
ti
m
iz
at
ion
pr
inci
ple
t
o
the
co
ntr
oller
desig
n.
T
he
m
at
hem
atical
exp
ressi
on
f
or
th
e
sli
din
g
m
ode
con
tr
oller
sy
stem
ρ
can
be
now
represe
nted
as
,
1
0
2
1
.
)
(
n
i
ia
(2)
In
the
a
bove
expressi
on,
th
e
var
ia
ble
ia
will
rep
rese
nts
input
argum
e
nts
w
hich
is
ba
sic
al
ly
an
int
egr
at
e
d
functi
on
of
error
a
nd
ti
m
e.
T
he
m
at
he
m
a
ti
c
al
exp
re
ssio
n
inclu
des
tw
o
m
or
e
var
ia
bles
v
iz
.
ϕ
1
and
ϕ
2
.
T
he
firs
t
com
po
ne
nt ϕ
1
r
epr
ese
nts a
con
sta
nt att
rib
uted
of
po
sit
ive
val
ue
a
nd can
b
e
re
pr
ese
nted
as,
ϕ
1
=
ϕ
c
(3)
Si
m
il
arly
, th
e second
c
om
po
ne
nt ϕ
2
re
pr
e
sen
ts an e
xpone
ntial
f
unct
io
ns
w
ho
s
e em
pirical
f
orm
is,
ϕ
2
=
e
n
-
q
(4)
The varia
ble
q
is eq
uiv
al
ent t
o (
c
-
1) a
nd the
pro
posed
stu
dy
us
es a
stabil
it
y functi
on as
f
ollow
s
,
2
2
1
S
(5)
Ther
e
f
or
e,
the
der
i
vative ca
ndidate
expres
sio
n for the
stabil
it
y functi
on w
il
l be,
|
|
.
5
.
0
2
dt
d
S
(6)
The
a
bove
e
xp
ressio
n
is
basi
cal
ly
us
ed
f
or
analy
zi
ng
t
he
a
m
ou
nt
of
sta
bi
li
t
y
wh
il
e
desi
gn
i
ng
the
c
on
t
ro
ll
er
syst
e
m
, an
d
t
he
se furt
her
res
ults in
|
|
.
)
(
.
s
i
gn
(7)
The
pro
po
se
d
syst
e
m
is
sai
d
to
offe
r
bette
r
conve
rg
e
nce
pe
rfor
m
ance
to
wards
th
e
surf
ace
of
the
sli
di
ng
i
f
the
conditi
on
μ>0
is
fou
nd.
E
ve
n
if
the
re
a
re
pr
esence
of
a
ny
form
of
the
st
at
e
-
base
d
tra
j
e
ct
or
ie
s
t
han
al
so
t
he
pro
po
se
d
syst
em
cou
ld
su
cce
ssfu
ll
y
re
ac
h
an
eff
ect
i
ve
co
nver
ge
nce
sta
te
wh
e
re
ρ
=0
.
He
nce,
no
w
a
co
nt
ro
ll
er
desig
n
is e
xpre
ssed
em
pirical
ly
as,
α
=
δ
1
-
δ
2
-
δ
3
(8)
It
can
be
seen
that
there
are
th
ree
par
am
et
ers
associat
ed
with
the
con
t
ro
ll
e
r
desig
n.
T
he
f
irst
par
am
et
er
δ
1
can
be fo
rm
ulate
d
as,
.
.
1
1
K
k
(9)
Wh
e
re,
k
K
represents
gain
f
unct
ion.
Sim
i
la
rly
,
the
second
and
thi
rd
par
a
m
et
er
of
ex
pr
e
ssion
(7)
can
be
represe
nted
as
,
.
1
2
k
(10)
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
20
88
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
4310
-
4320
4316
&,
)
(
.
1
3
k
(11)
The
va
riable
χ
an
d
τ
in
a
bove
ex
pr
es
sio
n
r
epr
ese
nts
a
ppr
ox
im
at
ing
f
unct
ion
s
a
nd
pro
du
ct
of
gai
n
and
α
of
fu
zz
y
log
ic
res
pec
ti
vely
con
side
r
ing
the
i
nput
argum
ent
of
ρ
.
The
refor
e
,
it
is
ob
se
rv
e
d
t
hat
the
pro
po
se
d
syst
e
m
pr
ov
ide
s
a
c
losed
-
lo
op
syst
e
m
whose
sta
bi
li
zat
ion
perfor
m
ance
is
ens
ured
wh
il
e
de
vel
op
i
ng
the
novel
sli
din
g
m
od
e
fu
z
z
y
con
tr
oller
syst
e
m
.
This
will
m
ean
that
pr
opos
e
d
syst
em
of
fe
rs
m
uch
red
uce
d
dep
e
ndencies
on
t
he
com
plete
ness
of
na
vi
gational
data
on
s
om
e
sp
eci
fic
instance
of
t
tim
e
on
un
certai
n
seq
uen
ces
of s
uch o
cc
urances
. T
he next secti
on outl
ines
res
ult analy
sis.
4.
RESU
LT
A
N
SLYSI
S
A
s
the
propose
d
syst
em
introdu
ce
s
a
sim
ple
and
novel
sli
di
ng
m
od
e
f
uzz
y
con
tr
oller
syst
e
m
,
hen
ce
it
s
eff
ect
ive
ne
ss
is
assesse
d
us
in
g
sp
ee
d
of
t
he
c
ontroll
er
a
nd
co
ntr
ol
signa
l
with
r
espect
to
incr
easi
ng
ob
s
er
vation
ti
m
e.
The
c
om
pu
ta
ti
on
al
e
ff
ec
ti
ven
ess
with
viabili
ty
of
th
e
pro
posed
syst
e
m
is
analy
zed
f
or
distrib
ution
functi
on
a
nd
error
perform
ance.
The
outc
om
e
is
co
m
par
ed
with
e
xisti
ng
s
yst
e
m
witho
ut
fu
zzy
log
ic
im
ple
m
e
ntati
on
(EX
-
W
OF
)
a
s
well
as
existi
ng
syst
e
m
with
fu
zzy
log
ic
im
ple
m
entat
ion
re
pr
ese
nt
ed
as
EX
-
WF
[
51
]
.
The
pro
posed
stud
y
us
es
na
vig
at
io
nal
data
[
52
]
t
o
fi
nd
t
hat
pr
opos
e
d
syst
e
m
of
fe
rs
bette
r
perform
ance.
Fr
om
the
Fig
ure
2
a
nd
Fi
g
ure
3
sho
ws
,
it
is
ob
s
er
ved
t
hat
the
pro
po
se
d
syst
e
m
pr
ov
i
de
s
the
highe
r
con
t
ro
ll
er
s
pee
d
a
nd
optim
al
con
t
ro
l
si
gn
al
with
re
sp
ect
to
increasi
ng
ti
m
e.
The
outc
om
e
show
t
hat
ex
ist
in
g
m
echan
ism
of
sli
din
g
m
od
e
c
on
t
ro
ll
in
g
is
hig
hly
inca
pa
ble
of
pe
rfo
rm
ing
faster
c
on
verg
ence
to
obta
in
bette
r
nav
i
gationa
l
data
and
he
nce
it
s
sp
eed
is
qu
it
e
lower
.
Mo
re
ov
e
r,
just
by
add
i
ng
f
uzzy
lo
gic
in
highly
a
dap
ti
ve
m
ann
er
in
exi
sti
ng
sli
ding
m
od
e
con
tr
oller
desig
n
only
increases
the
sp
ee
d
to
so
m
e
extent
ow
i
ng
to
zer
o
op
ti
m
iz
ation
on
the
data,
whic
h
is
ca
rr
ie
d
ou
t
i
n
pro
pose
d
syst
em
in
th
e
init
ia
l
sta
ge
i
tse
lf.
T
his
is
a
lso
the
pr
im
e
cause
of
higher
nu
m
ber
of
c
on
t
ro
l
s
ign
al
from
bo
th
the
form
s
of
existi
ng
syst
e
m
wh
il
e
pr
opose
d
syst
e
m
o
ff
e
rs h
igh
ly
pre
dicti
ve
co
n
tr
ol sig
na
l i
rr
es
pecti
ve o
f
im
par
ti
al
o
r v
agu
e
contents
of n
a
vig
at
io
nal
d
at
a.
Figure
2
.
S
pee
d
a
naly
sis
Figure
3
.
Co
ntr
ol Signal
a
naly
sis
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Novel
Fr
am
ew
or
k f
or
N
avig
at
ion
us
in
g
E
nh
an
ce
d F
uzzy
A
pp
r
oac
h
wi
th
Sl
iding
Mo
de
…
(
T.
So
meswa
ri
)
4317
Fig
ure
4
an
d
F
ig
ure
5
highli
ghts
the
com
par
at
ive
analy
sis
of
acc
ur
acy
wi
th
resp
ect
to
CDF
of
roo
t
m
ean
sq
ua
re
error
ove
r
incre
asi
ng
val
ue
of
inp
ut
H
an
d
error
pe
rfo
rm
a
nce
with
dif
fe
ren
t
pa
ram
et
er
s.
The
ou
tc
om
es
sh
ows
that
pro
po
sed
syst
em
of
fer
s
sig
nificantl
y
lower
scal
e
of
e
r
r
or
in
c
om
par
ison
to
e
xisti
ng
syst
e
m
.
At
the
sam
e
tim
e,
it
has
bee
n
obser
ve
d
that
pro
posed
syst
e
m
con
su
m
es
appr
ox
im
at
ely
1.552
6
seco
nd
s
of
pro
cessi
ng
wh
e
re
as
the
existi
ng
syst
e
m
of
both
f
or
m
s
con
sum
es
app
r
ox
im
at
el
y
9.
3258
s
econd
s
.
This
sho
ws
th
at
pr
op
os
e
d
syst
e
m
of
fer
s
go
od
acc
ur
acy
w
it
h
cost
eff
ect
i
ve
com
pu
ta
ti
onal
perf
or
m
ance
in
con
t
ro
ll
er
oper
at
ion
.
Figure
4
.
Com
par
at
ive
Analy
sis of CD
F
of
RM
SE
Figure
5
.
Com
par
at
ive
Analy
sis of E
rror Per
form
ance
5.
CONCL
US
I
O
N
This
discu
ssio
n
of
pa
pe
r
basi
cal
ly
pr
ese
nts
a
bri
dge
t
o
cl
ose
the
trade
off
of
im
ple
m
entin
g
inte
gr
at
e
d
GP
S
an
d
I
NS
base
d
nav
i
gation
syst
em
wit
h
sli
di
ng
m
od
e
co
ntr
oller.
T
he
c
on
t
rib
utio
n
as
well
as
novelty
factors
of
the
pr
op
os
e
d
stu
dy
are:
i)
un
l
ike
existi
n
g
a
ppr
oach
es
of
i
m
ple
m
enting
fu
zzy
sli
ding
m
od
e
con
t
ro
ll
er,
t
he
pro
po
se
d
syst
e
m
do
esn
’t
us
e
fu
zzy
r
ules
li
ke
existi
ng
syst
e
m
.
The
co
ns
tr
uction
of
f
uzz
y
ru
le
s
are
create
d
on
the
ba
sis
of
accu
racy
to
be
obta
ine
d
in
the
en
d
st
age
co
ns
id
erin
g
va
rio
us
no
n
-
li
nea
r
par
am
et
ers,
ii
)
diff
e
ren
t
f
r
om
existi
ng
na
vi
ga
ti
on
al
syst
e
m
,
the
input
data
is
no
t
process
ed
as
it
is
bu
t
it
i
s
su
bject
e
d
to
optim
iz
at
ion
us
ing
non
-
li
near
op
ti
m
iz
ation
pri
nciple
an
d
th
en
it
is
su
bjec
te
d
to
f
uzzy
sli
ding
m
od
e
con
tr
oller,
ii
i)
by
doin
g
the
ab
ov
e
t
wo
op
e
rati
on,
the
it
erati
ve
char
e
ct
erist
ic
s
of
op
tim
iz
at
ion
proc
ess
i
s
reduce
d
to
a
la
rg
e
exte
nt
to
obta
ined
fa
ste
r
conve
rg
e
nce
pe
rfor
m
ance.
T
he
st
ud
y
outc
om
e
is
fou
nd
to
off
e
r
sign
ific
a
nt
im
pr
ovem
ent
in
te
r
m
s
of
er
ror
m
i
nim
iz
at
ion
w
he
n
c
om
par
ed
w
i
th
existi
ng
sli
ding
m
od
e
c
on
trolle
r
desig
n wit
h
a
nd
without
fu
zz
y appr
oach.
Evaluation Warning : The document was created with Spire.PDF for Python.
IS
S
N
:
20
88
-
8708
In
t J
Elec
&
C
om
p
En
g,
V
ol.
8
, N
o.
6
,
Dece
m
ber
201
8
:
4310
-
4320
4318
REFERE
NCE
S
[1]
Nila
nj
an
De
y
,
A
m
arty
a
Mukher
j
ee
,
Embedde
d
S
y
stems
and
Rob
oti
cs
wi
th
Open
Source
Too
ls,
C
RC Pre
ss
,
2016
.
[2]
Marion
Hersh, M
ic
hael
A Johns
on,
As
sistive
Technol
og
y
for
Visually
Im
pai
r
ed
a
nd
Bli
nd
People,
Springer
Scie
nc
e
&
Business Med
ia
,
201
0
.
[3]
V.E
.
Balas, L.
C
.
Jain, X.
Zha
o
,
I
nform
at
ion
T
ec
h
nolog
y
and
Int
ellige
nt
Tr
ansport
at
ion
S
y
st
ems
,
I
OS
Press
,
2017
.
[4]
Bogusław
Dołęga
,
Robert
Głębo
cki
,
Dam
ia
n
Kordos
,
Marc
in
Żug
aj
,
Advanc
es
in
Aerospac
e
Guid
anc
e
,
Navigation
and
Contro
l, 201
7
.
[5]
W
ei
Quan
,
Jianli
Li
,
Xiao
li
n
Gong
,
Jianc
heng
F
ang
,
INS
/CNS
/G
NS
S
Inte
gra
t
ed
Naviga
t
ion
Te
ch
nolog
y
,
Springer
,
2015
.
[6]
Stephe
n
W
.
Hin
ch
,
Outdoor
Nav
iga
ti
on
with
GP
S,
W
il
der
n
ess Press,
2010
.
[7]
Adam
W
ei
ntri
t
,
Acti
vities
in
Na
viga
ti
on
:
Mar
ine Navi
ga
ti
on
and S
afe
t
y
of
Sea
Tr
ansporta
t
ion
,
CRC Press,
2015
[8]
Hass
an
A.
Kari
m
i,
Advanc
ed L
oca
t
ion
-
Based
T
ec
hnolog
ie
s a
nd
Servic
es,
CRC P
ress,
2016
.
[9]
Paul
D.
Groves
,
Princi
pl
es
of
GN
SS
,
Ine
rti
al,
a
nd
Multi
sen
sor
Inte
gra
te
d
Nav
i
gat
ion
S
y
stems
,
Second
Ed
it
ion
,
Artec
h
Hous
e
,
2
013
.
[10]
George
j
.
Refu
t
o
,
Evol
u
ti
on
of
the
US
sea
-
b
a
sed
nucl
e
ar
m
issile
de
te
rr
ent
:
warfi
ghti
ng
ca
p
abi
litie
s
,
Xlibr
is
Corpora
ti
on
,
20
11
.
[11]
W
il
son W
ong
,
Jam
es
Gordo
n
Fe
rguss
on
,
Mili
t
ar
y
Sp
ac
e
Pow
er:
A Guide
to
th
e
I
ss
ues,
ABC
-
CLIO,
2010
.
[12]
Hari
dwar
Singh
,
Him
anshu
Shekhar
,
Sol
id
Rock
e
t
Propellants:
Sc
ie
nc
e
and
T
ec
hn
olog
y
Challeng
e
s,
Ro
y
al
Societ
y
of
Chemistr
y
,
20
16
.
[13]
P.D.
Groves,
L.
W
ang,
D.
W
al
t
er,
H.
Mar
ti
n,
K.
Voutsis
and
Z.
Jiang
,
"The
f
our
ke
y
challe
n
ges
of
adva
nc
e
d
m
ult
isensor
n
avigati
on
and
posi
tioning"
,
2014
I
E
EE
/ION
Posit
io
n,
Locat
ion
and
Navi
gati
on
Sym
posium
-
PLANS
2014
,
Monter
e
y
,
CA,
2014
,
pp
.
7
73
-
792.
[14]
Jinkun L
iu
,
Slid
i
ng
Mode
Contro
l
Us
ing
MA
TL
A
B,
Aca
d
emic
Pr
e
ss
,
2017
.
[15]
L.
Dac
h
eng,
L.
Zhi
guo
and
Z
.
Shuai,
"A
naly
sis
on
the
infl
u
ence
of
the
INS
pre
c
ision
in
an
ultra
-
ti
ghtly
coupl
ed
INS
/GP
S
int
egr
at
ion"
,
2016
IE
EE
Information
Technol
ogy
,
N
et
working
,
Elec
tronic
and
Autom
ati
on
Control
Confe
r
ence
,
Cho
ngqing,
2016
,
pp
.
1106
-
1111
.
[16]
S.Y.
Cho,
"IM
-
fil
te
r
for
INS
/GPS
-
int
egr
a
te
d
nav
iga
ti
on
s
y
stem
c
onta
ini
ng
low
-
c
ost
g
y
ros"
,
in
IE
EE
Tr
ansacti
ons
on
Ae
ros
pace an
d
Elec
troni
c
S
yst
ems
,
vol
.
50
,
no
.
4,
pp.
2619
-
262
9,
Octob
er
2014
.
[17]
B.
P.
Duong
and
V.H.
Ngu
y
e
n
,
"D
eve
lopment
of
a
GP
S/IN
S
int
egr
at
ed
nav
iga
t
ion
s
y
stem
for
m
ode
l
ai
rcr
aft
"
,
201
4
14th
Int
ernati
on
al
Conf
ere
nce o
n
Control, A
uto
mation
and
S
yst
ems (
ICCAS
2014)
,
Seoul,
2014
,
pp.
201
-
206
.
[18]
J.
Fang
and
X.
Gong,
"P
red
ic
t
i
ve
Ite
r
ated
Kal
m
an
Filt
er
for
I
NS
/GP
S
Inte
gra
ti
on
and
I
ts
Applicati
on
to
SA
R
Motion
Com
pensa
ti
on"
,
in
I
EE
E
Tr
ansacti
ons
on
Instrum
ent
ati
o
n
and
Me
asur
e
ment
,
vo
l.
59,
n
o.
4,
pp.
909
-
91
5
,
April
2010.
[19]
I.
U.
L
ee,
H.
Li,
N.M.
Hoang
and
J.M.
L
ee,
"N
avigati
on
s
y
stem
de
vel
opm
ent
of
th
e
Underwater
V
ehi
c
le
s
using
t
h
e
GP
S
/
INS
sensor
fusion
"
,
2014
1
4th
Inte
rnat
ional
Confe
renc
e
on
Control,
Au
tomati
on
and
Syst
ems
(
ICCAS
2014)
,
Seoul,
2014
,
pp
.
610
-
612.
[20]
Q.
Li
and
F.
Sun,
"S
trong
tr
ac
king
cubature
Kalman
fi
lt
er
al
gori
thm
for
GP
S/INS
Inte
gra
te
d
Navigati
on
S
y
stem"
,
2013
I
EE
E
Inte
rnat
ion
al
Conf
ere
nce o
n
Me
cha
tronic
s
and
Aut
omat
ion
,
Ta
k
amatsu,
201
3,
pp
.
1113
-
111
7.
[21]
Z.
W
u,
Y.
Zha
n
g,
J.
Sun
and
Z.
Yin,
"A
filter
algorithm
for
GP
S/INS
int
egr
at
ed
navi
ga
ti
on
S
y
s
tem
base
d
on
IM
M
-
AF"
,
2016
IEEE
Inte
rnat
ional
G
eosc
ie
n
ce and
R
emote
S
ensing
S
ymposium (
IGA
RSS)
,
Bei
j
ing, 2
016,
pp
.
838
-
84
1.
[22]
W
.
Yan,
L.
W
a
ng,
Y.
Jin
and
G.
Shi,
"H
igh
a
cc
ura
c
y
Nav
iga
t
ion
S
y
st
em
using
GP
S
and
INS
s
y
stem
integra
tion
strat
eg
y
"
,
2016
IEE
E
In
te
rnatio
nal
Confe
ren
ce
on
Cybe
r
Tec
hnology
in
Aut
omation,
Contr
ol,
and
In
te
l
li
g
en
t
Syste
ms
(
CY
BE
R
)
,
Chengdu,
201
6,
pp
.
365
-
369
.
[23]
H.
Naka
nishi
,
S.
Kana
t
a
and
T.
Sawara
gi,
"G
PS
-
INS
-
BARO
hy
brid
navi
g
at
ion
s
y
stem
ta
k
ing
in
to
account
grou
nd
eff
ect
for
au
tono
m
ous
un
m
anne
d
hel
i
copt
er"
,
201
2
IEE
E
In
te
rnat
i
onal
Sy
mpos
ium
on
Safe
t
y,
S
ec
u
rity
,
and
Re
scu
e
Robot
ic
s
(
SSRR
)
,
Col
le
ge
Sta
ti
on
,
TX
,
2012
,
pp
.
1
-
6.
[24]
H.
Qin,
L.
Con
g
and
X.
Sun,
"
Acc
ura
c
y
improvem
ent
of
GP
S/M
EMS
-
INS
int
egr
at
ed
navi
g
at
io
n
s
y
stem
durin
g
GP
S
signal
outage
for
la
nd
veh
i
cl
e
n
avi
ga
ti
on"
,
in
Journal
of
S
y
stems
Engi
nee
ri
ng
and
El
ectroni
cs
,
vol.
23
,
no.
2
,
pp.
256
-
264
,
Ap
ril
2012
.
[25]
Chun
-
Jung
Sun,
Chang
-
W
en
H
ung,
W
ei
-
Cheng
Huang
and
C.
E.
Li
n
,
"Im
ple
m
ent
at
ion
of
GP
S/INS
navi
gat
i
on
s
y
stem
using
low
-
cost
MEMS
s
ensors"
,
2010
5th
IEE
E
Confe
re
nce
on
Industrial
El
ectronics
and
Appl
ic
at
ions
,
Ta
i
chung, 2010,
pp.
1588
-
1592
.
[26]
S.Z
.
Jam
al
,
"T
i
ghtly
coupled
GP
S/INS
ai
rborne
navi
gation
s
y
st
em"
,
in
IEEE
A
erospace
and
Elec
troni
c
Syste
m
s
Magazine
,
vo
l. 2
7,
no
.
4
,
pp
.
39
-
42,
April
2012.
[27]
M.A.K.
Jara
da
t
and
M.F.
Abdel
-
Hafe
z,
"N
o
n
-
Li
ne
ar
Autore
gre
ss
ive
Delay
-
Depe
ndent
INS
/
GP
S
Naviga
ti
on
S
y
stem Us
ing
N
eur
al Net
works
"
,
in
I
EEE
Sensors
Journal
,
vol
.
17
,
no
.
4
,
pp
.
1105
-
1115,
Feb
.
15,
1
5
2017.
[28]
W
.
Xin,
L.
Ji,
X.
Zha
ng
and
J.
Zha
o,
"F
aul
t
d
et
e
ct
ion
and
di
a
gnosis
i
n
the
IN
S/GP
S
navi
gat
io
n
sy
st
em
"
,
2014
World
Aut
omat
i
on
Congress
(
W
AC)
,
W
ai
koloa, H
I,
2014,
pp.
27
-
32.
[29]
J.
Ye,
P.
Mal
y
s
z
and
A.
Emad
i,
"A
Fixed
-
Sw
it
chi
ng
-
Frequ
ency
Int
egr
al
Slid
i
ng
Mode
Curre
nt
Control
l
er
for
S
witc
hed
Re
luc
t
anc
e
Moto
r
Driv
es
"
,
in
I
EE
E
Jou
rnal
of
Eme
rgin
g
and
Selec
t
ed
T
opic
s
in
Powe
r
El
e
ct
ronics
,
vol.
3,
no
.
2
,
pp
.
381
-
394,
June
2015
[30]
C.
C.
Chen
,
"A
n
opti
c
al
image
s
ta
biliz
at
ion
usin
g
novel
ultrason
ic
li
n
ea
r
m
otor
and
fuz
z
y
sl
idi
n
g
-
m
ode
cont
roller
for
porta
bl
e
dig
it
al
ca
m
cor
de
rs
"
,
in
I
EEE
Tr
ans
act
ions
on
Con
sum
er
El
ec
tr
oni
cs
,
vol.
63
,
no.
4,
pp.
343
-
349
,
Novem
ber
2017.
Evaluation Warning : The document was created with Spire.PDF for Python.
In
t J
Elec
&
C
om
p
En
g
IS
S
N:
20
88
-
8708
Novel
Fr
am
ew
or
k f
or
N
avig
at
ion
us
in
g
E
nh
an
ce
d F
uzzy
A
pp
r
oac
h
wi
th
Sl
iding
Mo
de
…
(
T.
So
meswa
ri
)
4319
[31]
Rao,
D.
Nara
si
m
ha,
T
.
Surned
ra,
and
S.
Ta
ra
Kal
y
ani,
"Im
prove
d
Perform
an
ce
of
DP
FC
Us
ing
Slidi
ng
Mo
de
Control
le
r
Meth
od
"
,
Int
ern
ationa
l
Journal
of Elec
tri
c
al
and
Com
p
ute
r Engineering
,
6
.
5
(2016):
207
3.
[32]
Bou
arr
oudj,
No
ure
ddine,
Djame
l
Boukhet
ala,
an
d
Fare
s
Boudje
m
a
,
"S
li
ding
-
Mode
Control
l
er
Ba
sed
on
Frac
ti
ona
l
Order
Calculus
for
a
Cl
ass
of
Nonlinear
S
y
st
ems
"
,
Inte
rna
ti
on
al
Journal
of
Elec
tri
c
al
and
Com
pute
r
Eng
ineeri
ng
,
6.
5
(2016):
2239
.
[33]
Othm
an,
Siti
Marha
in
i
s,
et
.
al.
,
"O
pti
m
iz
at
ion
of
Modifie
d
Sli
ding
Mode
Control
ler
for
an
E
le
c
tro
-
h
y
dra
ul
i
c
Actua
tor
s
y
s
te
m
with
Mism
at
che
d
Disturbance
"
,
Inte
rn
at
ion
al
Journal
of
Elec
tr
ic
a
l
and
Com
pute
r
Engi
n
ee
r
in
g
(IJECE)
,
8
.
4
(20
18).
[34]
T.
D.
Do,
"D
ist
urba
nce
Obs
erv
er
-
Based
Fuzz
y
SM
C
of
W
EC
Ss
W
it
hout
W
i
nd
Speed
Mea
s
ure
m
ent
"
,
in
IE
EE
Ac
c
ess
,
vol
.
5
,
p
p.
147
-
155
,
201
7.
[35]
A.A.A.
El
gamm
al
and
M.F.
El
-
n
agga
r,
"M
OP
SO
-
base
d
opti
m
al
c
ontrol
of
shunt
ac
ti
v
e
power
fil
ter using
a
var
ia
bl
e
struct
ure
fuz
z
y
l
ogic
slid
ing
m
ode
controll
er
for
h
y
br
id
(FC
-
PV
-
W
ind
-
Bat
te
r
y
)
e
ner
g
y
utilisat
ion
sche
m
e"
,
in
IE
T
Re
newab
le
Pow
er
Gene
ration
,
v
ol.
11
,
no
.
8
,
pp
.
1148
-
1156,
6
28
2017.
[36]
J.
Fei,
Y.
Chu
and
S.
Hou,
"A
Bac
kstepp
ing
Neura
l
Globa
l
Sli
ding
Mode
Control
Us
ing
Fuzz
y
Approxim
at
or
for
Thre
e
-
Phase
Act
ive
Pow
er
Fi
lt
e
r
"
,
in
IEEE Ac
ce
s
s
,
vol. 5, pp. 160
21
-
16032,
2017
.
[37]
H.
Li
,
J.
W
ang,
L.
W
u,
H.
K.
Lam
and
Y.
Gao,
"
Optimal
Guara
nt
ee
d
Cost
Slidi
ng
-
Mode
Control
of
Inte
rva
l
T
y
p
e
-
2
Fuzz
y
T
ime
-
Del
a
y
S
y
st
ems
"
,
in
I
EE
E
Tr
ansacti
o
ns on
Fuzzy
S
yst
ems
,
vol
.
26
,
no
.
1,
pp.
246
-
257,
Feb.
2018
.
[38]
Q.
Shen,
P.
Shi
and
Y.
Shi,
"D
istri
bute
d
Adaptive
Fuzz
y
Contr
ol
for
Nonline
ar
Multi
age
nt
S
y
s
te
m
s
Via
Slidi
ng
Mode
Obs
erv
ers
"
,
in
IEEE
Tr
ansacti
ons on
Cyb
e
rneti
cs
,
vol
.
46
,
no.
12
,
pp
.
3086
-
3097,
De
c. 201
6.
[39]
T.
W
ang
and
J.
Fei,
"A
daptive
Neura
l
Contro
l
of
Acti
ve
Pow
er
Filt
er
Us
in
g
Fuzz
y
Slidi
ng
Mode
Control
l
e
r
"
,
in
IE
EE A
c
ce
ss
,
vol.
4
,
pp
.
6816
-
6822,
2016
.
[40]
S.
W
en,
M.Z
.
Q.
Chen,
Z.
Ze
ng,
X.
Yu
and
T.
Huang,
"F
uzzy
Control
for
Unce
rtain
Vehicle
Acti
ve
Sus
pensi
on
S
y
stems
via
D
y
n
amic
Slidi
ng
-
Mode
Approac
h"
,
in
IEEE
Tr
ansacti
ons
on
Syste
ms
,
Man,
and
Cybe
rnetics:
Syste
ms
,
vol.
47
,
no
.
1
,
pp
.
24
-
32
,
Jan
.
201
7.
[41]
S.
W
en,
M.Z
.
Q
.
Chen,
Z.
Z
eng,
T.
Huang
and
C.
Li
,
"A
dapt
iv
e
Neura
l
-
Fuz
z
y
Sl
id
ing
-
Mode
Fault
-
Tol
er
ant
Control
for
Unce
rtain
Nonlinear
S
y
stem
s"
,
in
IEEE
Tr
ansacti
ons
on
Syst
ems,
Man,
and
Cybe
rnet
ic
s:
Sys
te
ms
,
vol.
47
,
no
.
8,
pp
.
2268
-
227
8,
Aug.
2017.
[42]
J.
Yu,
J.
Li
u,
Z.
W
u
and
H.
Fang
,
"D
ept
h
Control
of
a
Bioi
nspired
Roboti
c
Dolphi
n
Based
on
Sl
idi
ng
-
Mode
Fuzz
y
Control
Me
thod
"
,
in
IEEE
Tr
ansacti
ons on
Industrial
E
lectronic
s
,
vol. 65, no. 3, p
p.
2429
-
2438
,
M
arc
h
2018
.
[43]
X.
Zha
o,
H.
Yan
g,
W
.
Xia
and
X.
W
ang,
"A
dapt
i
ve
Fuzz
y
Hi
er
ar
chi
c
al
Slidi
ng
-
M
ode
Control
for
a
Cla
ss
of
MIMO
Nonline
ar
T
ime
-
Delay
S
y
st
ems
W
it
h
Input
Satura
ti
on"
,
in
IE
EE
Tr
ansacti
ons
on
Fuzzy
Syste
ms
,
vol.
25,
no.
5
,
pp.
1062
-
1077,
Oc
t.
2017.
[44]
A.
Mohant
y
,
S.
Patra
and
P.
K.
Ra
y
,
"Robus
t
fuz
z
y
-
slid
ing
m
o
de
base
d
UP
FC
cont
roll
er
for
t
ran
sient
stabili
t
y
ana
l
y
sis
in
aut
o
nom
ous
wind
-
di
ese
l
-
PV
h
y
br
id
sy
stem"
,
in
IET
Gene
ration,
Tr
a
nsm
ission
&
Di
s
tribut
ion
,
vol.
1
0,
no.
5
,
pp
.
1248
-
1257,
4
7
2016
.
[45]
S.
Raj
endiran,
P.
La
kshm
i
and
B.
Raj
kum
ar,
"F
ra
ct
ion
al
orde
r
fu
z
z
y
sl
idi
ng
m
o
de cont
roller
for
th
e
quar
te
r car
with
drive
r
m
odel a
n
d
dual actuators"
,
in
IET
El
e
ct
ri
c
al
Syst
ems i
n
Tr
anspor
tat
ion
,
vo
l.
7
,
no
.
2
,
pp
.
14
5
-
153,
6
2017.
[46]
Kai
-
W
ei
Chi
ang
,
Th
anh
Trung
Duong
and
Jhen
-
Kai
L
ia
o
,
“
The
Perform
anc
e
Anal
y
s
is
of
a
R
eal
-
Ti
m
e
In
te
gr
at
e
d
INS
/GP
S Vehi
cle
Navig
at
ion
S
ystem wit
h
Abnor
m
al
GP
S Mea
surem
ent
E
li
m
inati
on”,
Sensor
,
201
3
[47]
Xun
y
uan
Yin
,
Yingbo
Sun,
and
Changhong
W
ang,
“
Pos
it
ioni
n
g
Err
ors
Predic
t
ing
Method
of
Strapdown
Ine
rt
ia
l
Naviga
t
ion
S
y
s
t
ems
Based
on
P
SO
-
SVM”,
Hind
awi,
2013
[48]
N.K.
Fil
y
ashkin
and
V.S.
Yatskivsk
y
,
"P
red
i
ction
of
ine
rtial
navi
ga
ti
on
s
y
st
e
m
e
rror
d
y
nami
cs
in
INS
/GP
S
s
y
stem
"
,
2013
IEE
E
2nd
Int
ern
at
ion
al
Confer
e
nce
Actu
al
Problems
of
Unm
a
nned
Air
Vehicle
s
Deve
lopmen
ts
Proce
edi
ngs (AP
UA
VD
),
Kiev,
2
013,
pp
.
206
-
20
9.
[49]
Y.
Han,
M.G.
Safonov
and
J.
Chen,
"A
n
in
-
m
oti
on
alignm
ent
m
et
hod
for
inertia
l
n
avi
g
at
ion
s
y
stem
base
d
o
n
unfa
lsified
ad
ap
ti
ve
cont
rol
theor
y
"
,
Proce
ed
in
gs
of
2014
IE
EE
Chin
ese
Gu
ida
nc
e,
Nav
igat
ion
and
Con
tro
l
Confer
ence, Yan
ta
i
,
2014
,
pp
.
19
00
-
1905.
[50]
R.
Th
anga
r
aj
,
M
.
Pant
,
T.
R
.
Che
l
li
ah
and
A
.
Abra
ham,
"O
ppositi
o
n
base
d
Ch
aot
i
c
Diffe
ren
t
ia
l
Evo
l
uti
on
al
gori
thm
for
solving
globa
l
opti
m
izati
on
proble
m
s"
,
2012
Fourth
W
orld
Congress
on
Natur
e
and
Bio
logi
c
al
l
y
Inspire
d
Com
puti
ng
(Na
BIC),
Mexi
co
C
i
t
y
,
2012
,
pp
.
1
-
7.
[51]
Khoo
ban,
M.H.,
Niknam,
T.
,
‘A
new
and
robust
cont
rol
str
at
eg
y
for
a
class
of
nonli
ne
ar
power
sy
stems
:
ada
p
tiv
e
gene
ra
l
t
y
p
e
-
II
f
uzzy
’, Proc. Inst. Mec
h.
Eng
.
I
,
J
.
S
y
st
.
Contro
l E
ng.
,
2015,
229,
(
6),
pp
.
517
–
528
[52]
“
U.S.
Envi
ronm
ent
a
l
Prote
ct
ion
Agenc
y
”, https:/
/
ww
w.e
pa.
gov
/,
Ret
ribv
ed
on
02
April,
2018
BIOGR
AP
H
I
ES
OF
A
UTH
ORS
T.
Som
eswari
w
orking
as
assistant
profe
ss
or,
EEE
dept
a
t
Th
e
O
xford
Coll
eg
e
of
Engi
ne
eri
ng
has
publi
shed
3
int
ern
a
ti
on
al
jou
rna
l
papers
.
I
a
m
pursuing
Ph.
D
in
VTU
under
the
guida
n
ce
of
Dr.
Anilkumar
and
Dr.
Naga
r
a
j.
R.
I
hav
e
co
m
ple
te
d
m
y
B
.
E
degr
ee
in
20
07
from
JN
TU,
H
y
de
arb
ad
in El
ec
tr
ic
a
l and Elec
tron
i
cs
and
M.
E from
jnt
u,
Anant
apur
in
2010.
Evaluation Warning : The document was created with Spire.PDF for Python.