Proceedings of the 3rd IFAC Workshop on

Proceedings of the 3rd IFAC Workshop on

Automatic Control
Proceedings of thein Offshore
3rd Oil and Gas
IFAC Workshop on Production
Automatic Control in Offshore Oil andAvailable
Gas Production
online at www.sciencedirect.com
May 30 - June
Automatic
Proceedings of1,
Control2018.
in Esbjerg,
Offshore Denmark
Oil and Gas
on Production
May 30 - June 1,the 3rd
2018. IFAC Workshop
Esbjerg, Denmark
May 30 - June
Automatic 1, 2018.
Control Esbjerg,Oil
in Offshore Denmark
and Gas Production
May 30 - June 1, 2018. Esbjerg, Denmark
ScienceDirect
IFAC PapersOnLine 51-8 (2018) 245–250
Hovering
Hovering Control for Automatic Landing
Hovering Control
Control for for Automatic
Automatic Landing
Landing
Operation
Operation
Hovering of
of
ControlAn
An Inspection
Inspection
for Automatic Drone
Drone to
to
LandingA
A
Operation of An
Mobile Inspection
Platform Drone to A
Operation of Mobile
An Platform
Inspection
Mobile Platform Drone to A
Mobile
Shaobao Li, Platform
Petar Durdevic, Zhenyu Yang
Shaobao Li, Petar Durdevic, Zhenyu Yang
Shaobao Li, Petar Durdevic, Zhenyu Yang
Department Shaobao of EnergyLi, Petar
Energy Durdevic,
Technology, Aalborg Zhenyu YangEsbjerg
University,
Department of Technology, Aalborg University, Esbjerg
Campus,
Department Niels of Bohus
Energy Vej 8,
Technology,Esbjerg,
Campus, Niels Bohus Vej 8, Esbjerg, 6700, Denmark (e-mail: 6700,
Aalborg Denmark
University, (e-mail:
Esbjerg
Department shl@et.aau.dk;pdl@et.aau.dk;yang@et.aau.dk).
of Energy Technology,
Campus,shl@et.aau.dk;pdl@et.aau.dk;yang@et.aau.dk).
Niels Bohus Vej 8, Esbjerg, Aalborg
6700,University,
Denmark (e-mail: Esbjerg
Campus,shl@et.aau.dk;pdl@et.aau.dk;yang@et.aau.dk).
Niels Bohus Vej 8, Esbjerg, 6700, Denmark (e-mail:
shl@et.aau.dk;pdl@et.aau.dk;yang@et.aau.dk).
Abstract:
Abstract: This This paper
paper investigates
investigates the the hovering
hovering control
control problem
problem of of aa drone
drone on on aa 6-DOF
6-DOF mobile mobile
Abstract:An
platform. This paper investigates the hovering control problem of a is drone on a 6-DOF mobile
platform. An inner-outer loop strategy based on high-gain observer is presented. In
inner-outer loop strategy based on high-gain observer presented. In the
the outer
outer
loop,
Abstract:
platform.the high-gain
An This paperobserver
inner-outer loop is
investigates designed
the
strategy to estimate
hovering
based on the
control
high-gain states
problem
loop, the high-gain observer is designed to estimate the states of the drone such that its velocity of the
observerof adrone
drone
is such
on
presented. athat
6-DOF
Initsthevelocity
mobile
outer
and
loop, acceleration
platform.the An
high-gain measurements
inner-outer
observer loop is are
strategy
designed not tonecessary,
based on
estimate
and acceleration measurements are not necessary, and then a nested saturation controller and
high-gain
the then
states a
of nested
observer
the is
dronesaturation
presented.
such that controller
Initsthe outer
velocity is
is
designed.
loop,
and the In
acceleration the
high-gain inner
observer
measurements loop, is a hybrid
designed
are not controller
to estimate
necessary, which
the
and
designed. In the inner loop, a hybrid controller which can be effectively used for avoiding can
states
then be
of
a theeffectively
nested drone such
saturationusedthat for
its avoiding
velocity
controller is
designed.
unwinding
and In
acceleration the measurements
phenomenon inner loop, is applieda hybrid
are tonot controller
regulate
necessary,the which
attitude
and canofabe
then the effectively
drone.
nested used afor
Finally,
saturation avoiding
simulation
controller is
unwinding phenomenon is applied to regulate the attitude of the drone. Finally, a simulation
example
designed.is
unwinding
example isIn usedtheto
phenomenon
used demonstrate
toinner loop,
is applied
demonstrate a the
the toeffectiveness
hybrid controller
regulate
effectiveness theof the
the proposed
ofwhich
attitude canof be
proposed thecontrol
drone. method.
effectively
control used aforsimulation
Finally,
method. avoiding
unwinding
example is phenomenon
used to demonstrate is applied thetoeffectiveness
regulate theof attitude
the proposed of thecontrol
drone. method.
Finally, a simulation
© 2018, IFAC
example (International
is Hovering,
used to demonstrate Federation the ofeffectiveness
Automatic Control)
of Hosting by control
the proposed Elsevier method.
Ltd. All rights reserved.
Keywords:
Keywords: Hovering, landing, high-gain observer, hybrid control, underactuated system.
landing, high-gain observer, hybrid control, underactuated system.
Keywords: Hovering, landing, high-gain observer, hybrid control, underactuated system.
Keywords: Hovering, landing, high-gain observer,the
1. INTRODUCTION
INTRODUCTION hybrid control,
position of the
theunderactuated
drone. However,
However, system.
this method
method is is usually
usually
1. the position of drone. this
1. INTRODUCTION the position of the drone. However, this method is usually
not effective for large drones due to large inertia and input
not effective for large drones due to large inertia and input
Drones have
Drones have been 1.
been widelyINTRODUCTION
widely applied
applied for for environment
environment and and not effectiveIn
saturation.
the position
saturation.
for(Tan
of
In
large
the
(Tan etdrones
drone.
et al.
due to large
al. However,
(2016)),
(2016)), an inertia and
invariant
thisinvariant
an method is input
ellipsoid-
usually
ellipsoid-
Drones
production havesurveillance
been widely suchapplied
as offshore for environment
offshore oil & & gas
gas explo-
explo- saturation.
based method
effective In
for(Tan
control
large et al.
law
drones (2016)),
was
and based method control law was proposed for hoveringinput
not due to an
proposed
large invariant
for
inertia ellipsoid-
hovering
and of aa
production surveillance such as oil of
production
ration
Dronesand and surveillance
production,
haveproduction,
been widely such
wind as offshore
farm inspection,
applied inspection, oil
for environment & gas explo-
geography based
drone
saturation.method
on a control
moving
In (Tan ship
et law
deck.
al. was proposed
Only
(2016)),
and drone on a moving ship deck. Only the motion on heave of the
an for
motion
invarianthovering
on heave of
ellipsoid- a
of
ration wind farm geography drone
the
based shipon a
is
method moving
considered
control ship anddeck.
law Only
attitude
was the motion
control
proposed foris on
not
hoveringheave
studied.of of
a
ration
mapping,
productionandand production,
so on
surveillance wind
on (Schäfer
(Schäfer
such as farmet inspection,
al.
offshore (2016);
oil & Lee geography
Lee
gas et al.
explo-al. the ship is considered and attitude control is not studied.
mapping, and so et al. (2016); et the
drone ship
According on is
a considered
to our
moving survey,
ship and attitude
the
deck. existing
Only control
the is
approaches
motion not
on studied.
are
heave still
of
mapping,
(2016)).
ration and and
Take so on
the
production, oil (Schäfer
&wind et al.
gas production
production
farm (2016);
inspection, anLee
as an et al. According to our survey, the existing approaches are still
example.
geography
(2016)). Take the oil & gas as example.
(2016)).
Drones
mapping, Take
haveand the
been
so oil & gasto production
applied
on (Schäfer increase
et
Drones have been applied to increase the effecitivity al. the
(2016);as effecitivity
anLee example.
et al.of According
of
not
the ship
not is to
effective
effective
ourhovering
for
considered
for
survey,
hovering andthe and
and
existing
landing
attitude
landing
approaches
controlcontrol
is not
control of
aredrone
of studied. still
aa drone
Drones
areial
(2016)). have
inspection been
Take the applied
foroil
facility
& gas to increase
integrity
production the effecitivity
check,aspollution
pollution
an example. of not
on a effective
According 6-DOF to for
moving
our hoveringdeck.
survey,
mon- on a 6-DOF moving deck. More recently, some methodsand
the More landing
recently,
existing control
some
approaches of aaredrone
methods still
areial inspection for facility integrity check, mon- on
not aeffective
based 6-DOF moving
on inner-outer
inner-outer
for hoveringdeck.
loopand More
control recently,
landing strategy some
control of methods
providea drone an
areial
itoring,
Drones inspection
as
havewell as
been for facility integrity
facility/pipeline
applied to increase check,
inspection the pollution
etc. Drone
effecitivity mon- is
of based on loop control strategy provide an
itoring, as well as facility/pipeline inspection etc. Drone is based inspiration
on a on
6-DOF inner-outer
for hovering
moving loop
deck.controlcontrol
More on a strategy
6-DOF
recently, provide
moving
some methods an
deck.
itoring,
often as well remotely
operated
areial operated
inspection asforfacility/pipeline
by professional
facility professional
integrity inspection
check,operators
pollutionfrom
etc. from
Drone the
mon- is inspiration for hovering control on a 6-DOF moving deck.
often remotely by operators the
often operated
platform
itoring, asor remotely
boats.
well as However,by professional
facility/pipeline to realize
inspection
platform or boats. However, to realize a smooth and safe aoperators
smooth etc. from
and
Drone the
safeis inspiration
Cao
basedand
Cao and for hovering
Lynch
on Lynch
inner-outer (2016)
(2016)
control
looppresented
presented
on a 6-DOF
control an
strategy
an
moving
inner-outer
provide
inner-outer
deck.
loop
an
loop
platform
landing
often ofora commissioning
operated boats. However,
remotely by to realize
drone
professionalon the aoperators
smoothdeck
moving andissafe
from far
the Cao
control and
inspiration Lynch
strategy
for for(2016)
hovering presented
underactuated
control on a an inner-outer
drone
6-DOF with input
moving loop
and
deck.
landing of a commissioning drone on the moving deck is far control strategy for underactuated drone with input and
landing
more
more
ofora commissioning
complicated,
platform boats. and
complicated, and there
However,
there droneseems
to
seems
on to
realizetheus
to us amoving deck
that there
smooth
that andis
there no control
far
issafe
no
state
Cao and
state
strategy
Lynch for
constraints,
constraints,
underactuated
achieving
(2016)
achieving globally drone
presented
globally
with input
asymptotically
an inner-outer
asymptotically
and
stable
loop
stable
more
much
landingcomplicated,
study and there
or development
development
of a commissioning seems
being
drone to
theus
found
onfound that there
about
moving automatic
deck is far state
results constraints,
for position achieving
and globally
attitude
no results for position and attitude tracking. Naldi etand
control strategy for underactuated asymptotically
tracking.
drone withNaldi
input stable
et al.
much study or being about automatic al.
much
landing study or
of drone development
drone on onand being found about automatic results
(2017)
state for
proposedposition
constraints, a and
robust
achieving attitude
inner-outer
globally tracking.
loop
asymptoticallyNaldi
control et
scheme al.
stable
more complicated,
landing of aa moving
moving
there seems platform
platform to us either.
that there is no (2017) proposed a robust inner-outer loop control scheme
either.
landing
much study of drone on a movingbeing
or development platformfoundeither. about automatic (2017) for
for
proposed
aa drone
results drone with auncertain
for position
with
robust
and inner-outer
uncertain attitude
intertial
loop control
intertialtracking.
matrix.
matrix. Naldischeme
However,
However, the
et the
al.
Similar
Similar to the
to drone manned
the manned aircrafts’
aircrafts’ landing
landing on a mobile for
(2017)a
velocity droneand
proposedwith uncertain
acceleration
a robust
on a mobile velocity and acceleration are required to be measurableare intertial
required
inner-outer matrix.
to
loop be However,
measurable
control schemethe
in
landing
Similar of
toprocedure, on
the manned a moving platform
aircrafts’ landing either. onof aa mobile in
platform the automatic landing drone velocity
these
for a
platform procedure, the automatic landing of a drone these two works. In most of low-cost drones, acceleration two
droneand acceleration
works.
with In most
uncertain are
of required
low-cost
intertial to be
drones,
matrix. measurable
acceleration
However, in
the
platform
on toprocedure, the aircrafts’
automatic landinghoveringonof aa mobile
drone
on aa moving
Similarmoving deck
thedeck consists
manned
consists of two
of two landing
phases:
phases: hovering and these
and cannot
velocity
cannot
two
be
be
works.
measured.
and In most
acceleration
measured. Therefore,of low-cost
are
Therefore, hovering
required
hovering
drones,
control
to control acceleration
for drones
be measurable
for drones in
on a moving
landing.
platform During deck
procedure, consists
hoveringthe of two
phase,
automatic dronesphases:
landingneed hovering
to
of track
a and
the cannot
drone without
these beusing
two measured.
works. velocity
In Therefore,
most and
of hovering
acceleration
low-cost control
drones, for drones
measurements
acceleration
landing. During hovering phase, drones need to track the without using velocity and acceleration measurements
landing.
3D During
trajectory of the moving deck, keep
on trajectory
3D a moving deck
of the consists
hovering
moving of two
phase,
deck, drones
keep aa fixed
phases: fixed height
needhovering
to track
height toand
to the without
the would
cannotbe
would beusing
be more
measured.
more
velocity
desirable. and acceleration
Therefore,
desirable. hovering control measurements
for drones
3D trajectory
moving
landing. deck
During of
and the moving
adjust
hovering theirdeck,
phase, keep
attitudes
drones a fixed
properly
need height
to for
track to the
final would
without be more
using desirable.
velocity and acceleration measurements
moving deck and adjust their attitudes properly for final In this paper, we investigate the hovering control problem
moving
landing. deck
In this
3D trajectory thisand adjust
paper,
ofpaper,
the we their
moving willdeck, attitudes
focus on the
keep the properly
a fixedhovering
height for tofinal
controlthe In wouldthis bepaper,
morewe investigate the hovering control problem
desirable.
landing. In we will focus on hovering control In
of aathis
dronepaper,
above we investigate the hovering control problem
landing.
of a
movingdrone.In this
deck andpaper,
Normally, adjust we their
due will
to focus
the on theproperly
functional
attitudes
of a drone. Normally, due to the functional controllability of
hovering
controllability
for final
control drone above aa 6-DOF
6-DOF moving
moving deck under
deck under disturbances.
disturbances.
of a drone.
constraint
landing. In ofNormally,
of
thisthe drone,
paper, due
weit tocannot
itwill the
focus functional
track
on thethe the controllability
trajectory
hovering control of
The a drone
control
this above
paper, we a
objective 6-DOF
is
investigatetomoving
makethe the
of The control objective is to make the drone track theproblem
In drone
hovering
deck under track
control the trajec-
disturbances.
constraint the drone, cannot track trajectory of trajec-
constraint
the moving
of a moving of the drone,
deck while
drone. Normally, whiledue it cannot
keeping
to thetheir track
their the
attitudes
functional trajectory of The
tory
of a control
of
dronethe objective
moving
above a is
deck
6-DOF to make
while
moving
matching. tory of the moving deck while achieving the best matching
controllability the
achievingdrone
deck the
under track
best the trajec-
matching
disturbances.
the deck keeping attitudes matching.
the
In
In
moving
recent
constraint
recent years,
ofdeck
years,
while
some
thesomedrone, keeping
approaches
it cannot
approaches
their
have
track
have
attitudes
been matching.
developed
the developed
been trajectory of tory
to
to
attitude
attitude
of the
The control moving
between
objective
between
deck
the
the dronewhile
isdrone
to make achieving
and
and thedrone
the
the deck
deck
the bestthe
while
track
while
matching
safisfying
trajec-
safisfying
In
therecent
solve the years,
moving decksome
similar whileapproaches
hovering control
keeping have
problem.
their been For
attitudes developed
example, attitude
the
tory of thebetween
functional the drone
controllability
moving deck while and the
constraint.
achieving
to the functional controllability constraint. Towards this end, deck the while
Towards
best safisfying
this
matching end,
solve the similar hovering control problem. Formatching.
example, the functional
mapping
attitude between controllability
between the the
drone deckconstraint.
and and
the the
deck Towards
desired
while this
trajectory end,
safisfying
solve
Choi the Ahn
and similar hovering
(2015) studied control
the point problem.
point been For
tracking ofexample,
In recent
Choi and years,
Ahn some
(2015) approaches
studied the have tracking of aa drone
developed drone to the mapping between the deck and the desired trajectory
Choi
in 3D
solve and
the Ahn
space
similar(2015)
and studied
presented
hovering athe
control point tracking
backstepping-like
problem. For a drone the
ofexample,
feedback of the mapping
the between
drone satisfying
functional satisfying the
controllability thedeck and thecontrollability
functional
constraint. desired
Towardstrajectory con-
this con-
end,
in 3D space and presented a backstepping-like feedback of of
the the
straint
drone
drone
is
mapping satisfying
constructed,
between and
the
the
the
functional
functional
then
deck an
and
controllability
controllability
inner-outer
the desired loop con-
control
trajectory
in 3Dand
Choi space
linearizationAhn and
method
(2015)presented
for
studied athe
control backstepping-like
design.
point tracking The feedback
proposed
of a drone
linearization method for control design. The proposed straint is constructed, and then an inner-outer loop control
linearization
method
in 3D space
method is only
is onlymethod
and designed for for
presented
designed
control
static
fora static
design.
position
backstepping-like
position
The proposed
tracking.
feedback
tracking. In straint
In
scheme
of the drone
scheme
isbased
constructed,
on high-gain
on high-gain
based satisfying
andthethen
observer
observer
an inner-outer
functional is employed
is employed loop
for control
controllability
for control
con-
method
(Serra is
et al.
linearization only
al. (2016)),designed
(2016)),
method aa for for static
visual-servo-based position
control design. control tracking.
control law was
The proposed In scheme
development,
straint based
is on
where
constructed,high-gain
a observer
nested
and then
was development, where a nested saturation controller based is
saturation
an employed
inner-outer for
controller
loop control
based
(Serra et visual-servo-based law
(Serra
proposed
method etisfor
al.only
(2016)),
landing a visual-servo-based
control
designed for of a drone
static
proposed for landing control of a drone on a moving target, on
position a control
moving law
tracking. was
target, In development,
on
scheme
on the high-gain
the high-gain
based on where a nested
observer
high-gain
observer is
saturation
is proposed
proposed
observer forcontroller
is employed
for the outer
the outer based
for controlloop
loop
proposed
where
(Serra the
et for landinginner-loop
high-gain
al. (2016)), control
a of acontrol
drone is
visual-servo-based onapplied
acontrol
moving such
lawtarget,
that
was on the
control,
development, high-gain
and a hybrid
where observer
acontrol
nestedis proposed
is applied
saturation for
for the
the outer
inner-loop
controller loop
based
where the high-gain inner-loop control is applied such that control, control, and a hybrid control is applied for the inner-loop
where
the thefor
attitude
proposed high-gain
dynamcis
landing inner-loop
are of
control fastacontrol
enough
drone is
onapplied
tomoving
ato such
neglect control
that control
their
target, and
to
on the high-gain a hybrid
avoid control
unwinding
observer is applied for
phenomenon.
isphenomenon.
proposed for the
the outer
inner-loop loop
the attitude dynamcis are fast enough neglect their to avoid unwinding
the attitude
affects
where the dynamcis
on high-gain
the outer-loop
outer-loop are fast
control,
inner-loop enough
whichis is
control is to neglect
used to
applied to
such their
control
that control
control, to
andavoid
a unwinding
hybrid control phenomenon.
is applied for the inner-loop
affects on the control, which used control
affects on the dynamcis
the attitude outer-loopare control, which istoused
fast enough their control to avoid unwinding phenomenon.
to control
neglect
affects
Copyright
2405-8963 on© the
2018outer-loop
2018, IFAC control,Federation
IFAC (International which is ofused to control
Automatic Control)245 Hosting by Elsevier Ltd. All rights reserved.
Copyright © 2018 IFAC 245
Peer review©under
Copyright 2018 responsibility
IFAC of International Federation of Automatic
245Control.
10.1016/j.ifacol.2018.06.384
Copyright © 2018 IFAC 245
IFAC OOGP 2018
246
Esbjerg, Denmark. May 30 - June 1, 2018 Shaobao Li et al. / IFAC PapersOnLine 51-8 (2018) 245–250

Y
zY
Y
yY
ƌ TĚ ,Ĩ
 y
K
K
IĚ
K z z

z
TĚ
y z

;ĂͿ ;ďͿ
Fig. 1. Coordinate frames and hovering problem descrip-
tion
Notations: Throughout this paper, R, Rm and Rn×m
denote the spaces of real numbers, real m-vectors and real
n × m matrices, respectively. The terms  · p and  · ∞
denote the p-norm and infinity norm, respectively.
2. PROBLEM FORMULATION
2.1 Dynamic Model of Drone
To describe the motion of a drone and a 6-DOF moving
deck, three reference frames are defined as shown in Fig.
1(a): the inertial reference frame {I} fixed to the earth
surface, the platform body-fixed frame {B} attached to the
deck surface and the drone body-fixed frame {Q} attached
to the drone’s gravity center.
The drone is described by the following dynamic model:
ṗ = v
M v̇ = −uf Re3 + M ge3 + df
(1)
Ṙ = RS(ω)
J ω̇ = S(Jω)ω + uτ + dτ
where p = [x, y, z]T ∈ R3 and v = [vx , vy , vz ]T ∈ R3
denote the position of the gravity center of the drone and
its velocity in the inertial reference frame {I}, respectively,
uf ∈ R is the scale translational control force, uτ ∈ R3
is the attitude control torque, M > 0 ∈ R and J =
J T > 0 ∈ R3×3 are the mass and the inertia matrix of
the drone, respectively, df ∈ R3 and dτ ∈ R3 are the
bounded unknown disturbances, g = 9.8 is the acceleration
of gravity, e3 = [0, 0, 1]T , S(·) ∈ R3×3 is a skew symmetric
matrix regarding a vector x = [x1 , x2 , x3 ]T ∈ R3 given as
 
0 −x3 x2
S(x) = x3 0 −x1 ,
−x2 x1 0
and R ∈ SO(3) is the rotation matrix from {Q} to {I},
where SO(3) is the special orthogonal group of 3rd order:
SO(3) = {R ∈ R3×3 : RT R = RRT = I3 , |R| = 1}. (3)
The element of SO(3) can be parametrized by a unit
 T p∗d = pd + H(t)e3 +
quaternion as q = η T through Rodrigues formula:
R(q) = I3 + 2ηS() + 2S()2 ,
where η ∈ R and  ∈ R3 are the scalar and vector
components of the unit quaternion q. See more details in
(Shuster (1993)). Then the kinematic equation (the third
equation of (1)) can be replaced by the following equation:
 
1 1 −T
q̇ = q ⊗ ν(ω) = ω, (4)
2 2 ηI3 + S()
 T
where ν(ω) = 0 ω T and ⊗ is an operator between two
zY  
T T
quaternions qi = ηi i , i = 1, 2, defined as follows:
yY
  
η1 −Ti η2
y
q1 ⊗ q2 = .
1 η1 I3 + S(1 ) 2
\Ě y
2.2 Hovering Control Problem

This paper focuses on hovering control of a drone above
a 6-DOF moving deck. We assume that the desired tra-
jectory and attitude of the deck are determined by the
T
nonlinear functions pd (t) = [xd , yd , zd ] ∈ R3 and γd (t) =
T
[φd , θd , ψd ] ∈ R3 with respect to time t, respectively,
where pd (t) and γd (t) are 4th continuous differentiable.
Therefore, we can define the velocity and angular velocity
of the deck as vd (t) = ṗd and wd (t) = γ˙d , respectively.
Hovering control aims at tracking the position of the
moving deck and keeping a constant heaving distance
to the deck. This process is very important for a drone
landing on a moving deck safely. The most desirable
control is that the controlled drone can track the deck
while keeping the same attitude as that of the deck.
However, it is noted that the second equation of (1) is
underactuated, which means that as the drone flies along
the desired trajectory pd , its desired attitude R∗ must
satisfy the following functional controllability constraint
by neglecting disturbance df :
M ge3 − M p̈d
R ∗ e3 = . (5)
uf
Obviously, the moblie platform has its own intrinsic at-
titude Rd (γd ) that usually cannot satisfy the constaint
condition (5). Since position plays a major role in landing
control, this paper will study how to make the drone track
the position of the deck while achieving the best attitude
matching.
To make the descent of the drone smooth, a descenting
function H(t) is defined. It is conveniently designed to be
a smooth sigmoid function. An example will be given in
Section 4. Denote H0 = H(t0 ) be the initial height and
Hf = H(∞) be the final height of the drone in the heave
of the deck as shown in Fig. 1(b).
It is noted that, to keep the heaving distance Hf , the
desired position z ∗ of the drone in the inertial frame needs
(2) to follow the following trajectory:
Hf
z∗ = + zd . (6)
cos(θd )
Therefore, the hovering control problem can be solved if
the drone can track the desired trajectory p∗d described as
Hf e3
. (7)
cos(θd )
while keeping an attitude satisfying the constraint (5).
3. HOVERING CONTROLLER DESIGN
In this section, an inner-outer loop control strategy will be
introduced to design the hovering controller. The structure
of the inner-outer loop control is shown in Fig. 2.

246
 IFAC OOGP 2018
Esbjerg, Denmark. May 30 - June 1, 2018 Shaobao Li et al. / IFAC PapersOnLine 51-8 (2018) 245–250 247

pd uf Recall the translational motion control design with veloc-

ĞƐŝƌĞĚ dƌĂŶƐůĂƚŝŽŶĂů
dƌĂũĞĐƚŽƌLJ ŽŶƚƌŽů Id T d KƵƚĞƌ>ŽŽƉ ity feedback in (Naldi et al. (2017)) first. Let
z1 = p − p∗d , z2 = v − ṗ∗d .
\ d ƚƚŝƚƵĚĞ uW
ŽŶƚƌŽů With the first two equations of (1), we have

qd  Zd /ŶŶĞƌ>ŽŽƉ ż1 = z2
(14)
,ŝŐŚŐĂŝŶ
M ż2 = −uf Re3 + M ge3 − M p̈∗d + df
ƌŽŶĞ
pÖ  vÖ ŽďƐĞƌǀĞƌ
p
A vectored-thrust control strategy can be applied to design
the scalar controller uf . Define a control force vector
Fig. 2. Structure of Inner-Outer Loop Control
Fc (z1 , z2 ) as follows:
3.1 High-Gain Observer Fc (z1 , z2 ) = M ge3 − M p̈∗d + κ(z1 , z2 ). (15)
The high-gain observer is introduced in the outer loop where κ(z1 , z2 ) is a state feedback law satisfying κ(0, 0) =
control because velocity and acceleration measurement 0.
may be unavailable for some low-cost drones. It is designed Substituting (15) into the second equation of (14), we have
based on the following lemma:
M ż2 = −uf Re3 + Fc (z1 , z2 ) − κ(z1 , z2 ) + df . (16)
Lemma 1. (Behtash (1990)). Consider a system with its
output y(t) ∈ Rm and its first n − 1 derivative be- We can define a desired attitude Rc (z1 , z2 ) ∈ SO(3)
ing bounded, i.e., there are postive constants Yk , k = satisfying
1, · · · , n − 1 such that y (k)  < Yk . Given a high-gain uf Rc e3 = Fc (z1 , z2 ).
observer
 Choose the translational motion control law as
δ ξ˙i = ξi+1 , i = 1, · · · , n − 1
(8) uf = Fc (z1 , z2 ). (17)
δ ξ˙n = −λ1 ξn − λ2 ξn−1 − · · · − λn−1 ξ2 − ξ1 + y(t)
Then, we have
with δ being any small positive constant, ξi , i = 1, · · · , n
being the observer states, and λi , i = 1, · · · , n − 1, being Fc (z1 , z2 )
R c e3 = . (18)
chosen such that the polynomial sn +λ1 sn−1 +· · ·+λn−1 s+ Fc (z1 , z2 )
1 is Hurwitz, then we have Then, system (14) can be rewritten as
ξk+1

1) − y (k) = −δΨ(k+1) , k = 0, 1, · · · , n − 1, where ż1 = z2
δk (19)
Ψ = ξn + λ1 ξn−1 + · · · + λn−1 ξ1 . M ż2 = −uf (R − Rc )e3 − κ(z1 , z2 ) + df .
2) There exist positive constants t∗ and bk , only depend-
ing on Yk , k = 1, · · · , n, δ and λi , i = 1, · · · , n − 1, such A nested saturation function σ(·) can be used to design
that Ψ(k)  ≤ bk for all t > t∗ . κ(z1 , z2 ) to meet the saturation requirement of the scalar
control force uf . According to (Naldi et al. (2017)), the
Obviously, Lemma 1 implies that y (k) can be estimated by following state feedback law is available to guarantee the
ξk+1
δk bounded by δbk+1 , k = 0, 1, · · · , n − 1. It is noted
stability of (19) under the scalar control (17).
   
that the high-gain observer does not need to know the k2 k1
system dynamics but only its output. This property has κ(z1 , z2 ) = γ2 σ z2 + λ 1 σ z1 , (20)
γ2 γ1
made the high-gain observer widely applied to estimate
complex nonlinear systems. where k1 , k2 , γ1 , and γ2 are some positive constants,
and σ(·) is the saturation function satisfying the following
In this paper, a high-gain based on Lemma 1 is designed properties for any s ∈ R:
to estimate the position and velocity of the drone under
disturbances. The designed high-gain observer is given as 1) |σ̇(s)| ≤ 2;
follows: 2) |σ̈(s)| ≤ d for some d > 0;

δ π̇1 = π2 3) sσ(s) > 0 for all s = 0, and σ(0) = 0;
(9) 4) σ(s) = sgn(s) for |s| ≥ 1;
δ π̇2 = −λ1 π2 − π1 + p
5) |s| < |σ(s)| < 1 for |s| < 1.
where π1 , π2 ∈ R3 are the observer states. Choose λ1 such
that s2 + λ1 s + 1 is Hurwitz, and then the estimation of For a vector s, σ(·) is simply extended by each element
the position and velocity of the drone can be described as satisfying the above properties. Note that each dimension
p̂ = π1 (10) of κ(z1 , z2 ) is bounded by γ2 .
π2 Next, we will further discuss the translational motion
v̂ = (11)
δ control design independent of velocity and acceleration
and there exist constants b1 and b2 such that measurements. With the high-gain observer proposed in
p − p̂ ≤ δb1 (12) subsection 3.1, a new control law replacing the position
v − v̂ ≤ δb2 . (13) and velocity by their estimates is designed as
   
k2 k1
3.2 Translational Motion Control κ (ẑ1 , ẑ2 ) = γ2 σ ẑ2 + γ1 σ ẑ1 , (21)
γ2 γ1
Translational motion control aims at making the drone where ẑ1 = p̂ − p∗d and ẑ2 = v̂ − ṗ∗d . Then we can define a
track the desired trajectory p∗d such that the qudrotor can new control force vector Fc (ẑ1 , ẑ2 ) as
hover on the deck while keeping a desired heaving distance. Fc (ẑ1 , ẑ2 ) = M ge3 − M p̈∗d + κ (ẑ1 , ẑ2 ). (22)

247
IFAC OOGP 2018
248
Esbjerg, Denmark. May 30 - June 1, 2018 Shaobao Li et al. / IFAC PapersOnLine 51-8 (2018) 245–250
Let the translational motion control law in the form as
uf = Fc (ẑ1 , ẑ2 ), (23)
and the desired attitude Rc∗ satisfying functional control-
lability constraint is given by
Fc (ẑ1 , ẑ2 )
Rc∗ e3 = . (24)
Fc (ẑ1 , ẑ2 ) where G is the matrix with the first, second and third rows
With the new translational motion control law (23), sys-
tem (19) is changed into

ż1 = z2
(25)
M ż2 = −uf (R − Rc∗ )e3 − κ (ẑ1 , ẑ2 ) + df .
Then, we have the following results.
Theorem 2. Consider the drone system (1) tracking the
desired trajectory p∗d described in (7). Design a high-gain
observer in the form of (9) with the ploynomial s2 +λ1 s+1
being Hurwitz by choosing proper constant λ1 . Assume
that there exists a constant χ∗ > 0 such that uf (R −
Rc∗ )e3 ∞ ≤ χ∗ . Under the translational motion control
law (23) with κ (ẑ1 , ẑ2 ) and Fc (ẑ1 , ẑ2 ) given in (21) and
(22), respectively, and k1 , k2 , γ1 and γ2 satisfying
γ2 γ1 γ2 144k1
< , 4k1 γ1 < , < 1,
k2 4 4M k2
system (25) is ISS with respect to the input −uf (R −
Rc∗ )e3 +df , without restricions on the initial states and the
input. In particular, the states of system (25) are bounded
by the following asymptotic bound
6M 3M
z1 p ≤ v2 p , z2 p ≤ v2 p ,
k1 k2 k2
where v2 = (uf (R−Rc∗ )e3 +(κ(z1 , z2 )−κ (ẑ1 , ẑ2 ))+df )/M .
Proof. Rewrite (25) in the form as

 ż1 = z2
M ż2 = −κ(z1 , z2 ) − uf (R − Rc∗ )e3

+(κ(z1 , z2 ) − κ (ẑ1 , ẑ2 )) + df
Let χ(R, Rc∗ ) = −uf (R−Rc∗ )e3 . System (28) can be viewed
as a special case of system (C.1) in Appx C of (Isidori et al.
(2012)) with n = 2, q1 (t) = 1, q2 (t) = M 1
, v1 = 0 and
v2 = (χ(R, Rc∗ ) + (κ(z1 , z2 ) − κ (ẑ1 , ẑ2 )) + df ) /M .
It is noted that the inequalities (12) and (13) imply that
there exists a constant κ∗ > 0 such that κ(z1 , z2 ) −
κ (ẑ1 , ẑ2 )∞ ≤ κ∗ . Therefore, there exists a constant
v2∗ > 0 such that v2 ∞ ≤ v2∗ .
Directly using the Lemma C.2.1 in (Isidori et al. (2012)),
this theorem can be proved.
3.3 Attitude Control
In order to satisfy χ(R, Rc∗ )∞ ≤ χ∗ , an attitude control
will be designed to make R track Rc∗ . Note that the
rotation matrix Rc∗ is a mapping [φ∗c , θc∗ , ψc∗ ]T → Rc∗ ,
and have the relationship described in XYZ convention
as shown in (30). Therefore, through (24) we can fix ψc∗
and θc∗ . ψc∗ can be chosen randomly, but in order to achieve
the best matching attitude with the deck, the thrid Euler
angle of the desired attitude is fixed by ψc∗ = ψd . In this
way, the desired attitude of the drone is determined, and
the quaternion qc∗ corresponding to Rc∗ can be obtained
based on the quaternion algebra (Shuster (1993)).
Based on the third equation of (1), the desired angular
velocity ωc∗ corresponding to Rc∗ is calculated by
ωc∗ = GRc∗T Ṙc∗ e3 + ψ̇d e3 ,
given by [0, −1, 0], [1, 0, 0] and [0, 0, 0], respectively, and ψ˙d
is the angular velocity of the deck in yaw.
Define the quanternion error q̃ and the angular velocity
error ω̃ as
q̃ = qc∗−1 ⊗ q (31)
ω̃c = ω − ω̄c , (32)
where ω̄c = R(q̃)T ωc∗ . Then, the attitude error system can
be
 written as  
 1 0
q̃˙ = q̃ ⊗
2 w̃c
 ˙
J ω̃c = Λ(ω̃c , ω̄c )ω̃c + S(J ω̄c )ω̄c − JR(q̃)T ω̇c∗ + uτ + dτ
(33)
with Λ(ω̃c , ω̄c ) defined as
Λ(ω̃c , ω̄c ) = S(J ω̃c ) + S(J ω̄c ) − S(ω̄c )J − JS(ω̄c ).
Remark 3. q̃ = [±1, 0, 0, 0]T are both the equilibriums of
(26)
the first equation of (33), because q̃ = [±1, 0, 0, 0]T denote
the same attitude in the 3D space.
Referring to the hybrid control method proposed in (May-
hew et al. (2011)), the attitude control law is designed as
uτ = JR(q̃)T ω̇c∗ − S(J ω̄c )ω̄c − k3 h˜
 − k4 ω̃c , (34)
(27) where ˜ is the vector component of q̃, k3 and k4 are postive
constants, and h ∈ {−1, 1} is a variable governed by

ḣ = 0, hη̃ > −ζ
(35)
h+ ∈ sgn(η̃), hη̃ ≤ −ζ
where η̃ is the scalar of q̃, ζ ∈ (0, 1) is the hysteresis
(28) threshold, and sgn(η̃) is defined as

sgn(η̃), |η̃| > 0
sgn(η̃) =
{−1, 1}, η̃ = 0.
The attitude control law (34) is effective to handle the
unwinding phenomenon (Bhat and Bernstein (2000)) and
noise induced chattering (Mayhew et al. (2011)).
Then, we have the following theorem:
Theorem 4. Given the attitude error system (33), under
the control law (34), system state x̃ = [q̃ T , ω̃cT , h]T is
globally uniformly ultimately bounded. In particular, if
dτ ≡ 0, system (33) is asymptotically stable.

d Fc (ẑ1 ,ẑ2 )
Remark 5. Equation (24) implies Ṙc∗ e3 = dt |Fc (ẑ1 ,ẑ2 )| .
Then, ω̇c∗ is computed by
ω̇c∗ = GṘc∗T Ṙc∗ e3 + GRc∗T R̈c∗ e3 + ψ̈d e3
d Fc (ẑ1 , ẑ2 ) d2 F  (ẑ1 , ẑ2 )
= GS(ωc∗ )T Rc∗T + G 2 c + ψ¨d e3 .

dt |Fc (ẑ1 , ẑ2 )| dt |Fc (ẑ1 , ẑ2 )|
3.4 Stability of the Closed-Loop System
Combing Theorem 2 and Theorem 4, we can further obtain
the stability of the whole closed-loop control system.
Theorem 6. Consider the drone system (1) and the desired
trajectory p∗d described in (7). Under the scalar trans-
lational motion controller (21) with γ1 , γ2 , k1 and k2
248
 IFAC OOGP 2018
Esbjerg, Denmark. May 30 - June 1, 2018 Shaobao Li et al. / IFAC PapersOnLine 51-8 (2018) 245–250 249

 
cos(θc∗ ) cos(ψc∗ ) − cos(θc∗ ) sin(ψc∗ ) sin(θc∗ )
Rc∗ = cos(φc ) sin(ψc∗ ) + cos(ψc∗ ) sin(φ∗c ) sin(θc∗ )
∗
cos(φc ) cos(ψc∗ ) − sin(φ∗c ) sin(θc∗ ) sin(ψc∗ )
∗
− cos(θc∗ ) sin(φ∗c ) (30)
sin(φ∗c ) sin(ψc∗ ) − cos(φ∗c ) cos(ψc∗ ) sin(θc∗ ) cos(ψc∗ ) sin(φ∗c ) + cos(φ∗c ) sin(θc∗ ) sin(ψc∗ ) cos(φ∗c ) cos(θc∗ )

satisfying (26), and the attitude controller (34) with k3 > 0

and k4 > 0, the tracking errors z1 , z2 , q̃ and ω̃c are globally 10

x/m
0 p*
d
uniformly ultimately bounded. -10 p
0 5 10 15 20 25 30 35 40
This theorem is obviously based on the ISS property in 10

y/m
Theorem 2 and the globally uniformly ultimately bounded 0

property in Theorem 4, and thus the proof is omitted here. -10

0 5 10 15 20 25 30 35 40
Remark 7. It can be observed from Theorem 2 and Theo- 10
rem 4 that the tracking errors are dependent on df , dτ and

z/m
0
the estimation errors of the high-gain observer. Since small -10
0 5 10 15 20 25 30 35 40
δ can enhance the accuracy of the high-gain observer, the time/s
tracking errors of the drone can also mitigated by choosing
small δ. Fig. 5. The position trajectory of the drone.

4. SIMULATION STUDIES 1
0 q*
c

-1 q
0 2 4 6 8 10
2 1
x

1
0 0
-1
-2 0 2 4 6 8 10
0 10 20 30 40 2
1
0
2 -1
y

0 2 4 6 8 10
0 1
-2
3

0
0 10 20 30 40 -1
0 2 4 6 8 10
time/s
1
z

0
-1 Fig. 6. The attitude trajectory of the drone.
0 10 20 30 40
time/s
u (N/m)

60
Fig. 3. The position estimation error of the high-gain 40
20
f

observer. 0 5 10 15 20
u (N/m) u (N/m) u (N/m)

50
0
1
x
x

-50
v

0 0 5 10 15 20
-1
20
0 10 20 30 40
0
y

1 -20
y
v

0 0 5 10 15 20
-1 10
0 10 20 30 40 0
z

-10
1 0 5 10 15 20
z
v

0
time/s
-1
0 10 20 30 40
time/s
Fig. 7. The translational motion control force and attitude
control torques of the drone.
Fig. 4. The velocity estimation error of the high-gain  π   − cos(2t) 
6−5 cos( 30 t)
observer. pd (t) = π
−0.5−4 sin( 30 t) , γd (t) = 0.05 sin(2t) ,
sin(0.8t) − cos(2t)
In this section, a numerical example is presented to verify
the effectiveness of the proposed control solution. The respectively. The descending function H(t) is defined as
dynamic model of the drone is given by (1) with system Hf − H0
H(t) = H0 + , (36)
parameters: M = 3.25kg, J = diag(0.032, 0.032, 0.164), −6(2t−10)
1 + e 10
and df and dτ being the Gaussian white noises with
where, the initial height H0 of the drone is determined by
maximum aplitude of 1.5N and 0.05Nm, respectively.
its initial position in Z axis. The drone is desired to hover
The position trajectory and Euler angles of the 6-DOF at the height Hf = 1m above the deck. Then, the desired
deck follow the following nonlinear functions: position trajectory p∗d in (7) is obtained.

249
IFAC OOGP 2018
250
Esbjerg, Denmark. May 30 - June 1, 2018 Shaobao Li et al. / IFAC PapersOnLine 51-8 (2018) 245–250

of the drone achieve the best matching with the attitude

of the deck while satisfying the functional controllability
10 p
constraint of the drone. The hybrid control is also available
d to aviod the unwinding phenomenon. The stability of the
p whole closed-loop system is theoretically proved. In the
5 future, it would be interesting to further investigate visual
z/m

based landing control of a drone on a moving deck and

0 make an implementation on a practical drone.

10 ACKNOWLEDGEMENTS
5 15
10 This work is partially supported by the National Natural
0 5
y/m
-5 0 Science Foundation of P. R. China under Grant 61403334
x/m
and the Natural Science Foundation of Hebei Province
Fig. 8. The tracking trajectory of the drone in 3D. under Grant F2017203109.
REFERENCES
10
Behtash, S. (1990). Robust output tracking for non-linear
x/m

0
-10 systems. International Journal of Control, 51(6), 1381–
0 10 20 30 40 1407.
10
Bhat, S.P. and Bernstein, D.S. (2000). A topological ob-
struction to continuous global stabilization of rotational
y/m

5
0 motion and the unwinding phenomenon. Systems &
0 10 20 30 40 Control Letters, 39(1), 63–70.
10 Cao, N. and Lynch, A.F. (2016). Inner–outer loop control
z/m

5 for quadrotor uavs with input and state constraints.

0 IEEE Transactions on Control Systems Technology,
0 10 20 30
time/s
Fig. 9. The tracking errors of the drone.
The parameters of the high-gain observer (9), the control
force (21) and the control torques (34) are given as:
δ = 0.1, λ1 = 2, k1 = 0.1, k2 = 26, γ1 = 0.5, γ2 = 26,
k3 = 40 and k4 = 2.
Simulation results are shown in Figures 3-9. Figures 3
and 4 show the position estimation error and velocity
estimation error of the high-gain observer, respectively.
It is observed that the estimation errors are very small
and thus the designed high-gain observer is effective. In
Figure 5, the blue dash line denotes the desired trajectory
p∗d and the red line is the trajectory of the drone. It is
observed that the drone tracks the desired trajectory with
very small tracking errors. Figure 6 illustates that the
attitude of the drone tracks the desired attitude qc∗ . Figure
7 shows that the control force uf and the control torques
uτ are bounded. In Figure 8, the blue line denotes the
trajectory of the deck and the red line is the trajectory of
the drone. From Figure 8 as well as Figure 9, it is observed
that the drone tracks the moving deck while keeping an
approximate height Hf = 1m above the deck. In summary,
the simulation results demonstrate the effectiveness of the
proposed control solution.
5. CONCLUSION
This paper studies the hovering problem of a drone above
a 6-DOF moving deck. A high-gain observer is designed
to estimate the position and velocity of the drone. Then,
a nested saturation control is designed based on the high-
gain observer to make the drone track the trajectory of the
deck and keep a fixed heaving distance above the moving
deck. A hybrid control is applied to make the attitude
40 24(5), 1797–1804.
Choi, Y.C. and Ahn, H.S. (2015). Nonlinear control of
quadrotor for point tracking: Actual implementation
and experimental tests. IEEE/ASME transactions on
mechatronics, 20(3), 1179–1192.
Isidori, A., Marconi, L., and Serrani, A. (2012). Ro-
bust autonomous guidance: an internal model approach.
Springer Science & Business Media.
Lee, H., Hwang, Y.S., Jang, G., Kim, T., and Lee, J.M.
(2016). 3d map building using the lrf and sinusoidal
trajectory of a quadrotor. In Robotics and Biomimet-
ics (ROBIO), 2016 IEEE International Conference on,
690–695. IEEE.
Mayhew, C.G., Sanfelice, R.G., and Teel, A.R. (2011).
Quaternion-based hybrid control for robust global at-
titude tracking. IEEE Transactions on Automatic Con-
trol, 56(11), 2555–2566.
Naldi, R., Furci, M., Sanfelice, R.G., and Marconi, L.
(2017). Robust global trajectory tracking for underac-
tuated vtol aerial vehicles using inner-outer loop control
paradigms. IEEE Transactions on Automatic Control,
62(1), 97–112.
Schäfer, B.E., Picchi, D., Engelhardt, T., and Abel, D.
(2016). Multicopter unmanned aerial vehicle for auto-
mated inspection of wind turbines. In Control and Au-
tomation (MED), 2016 24th Mediterranean Conference
on, 244–249. IEEE.
Serra, P., Cunha, R., Hamel, T., Cabecinhas, D., and
Silvestre, C. (2016). Landing of a quadrotor on a moving
target using dynamic image-based visual servo control.
IEEE Transactions on Robotics, 32(6), 1524–1535.
Shuster, M.D. (1993). A survey of attitude representations.
Navigation, 8(9), 439–517.
Tan, C.K., Wang, J., Paw, Y.C., and Liao, F. (2016). Au-
tonomous ship deck landing of a quadrotor using invari-
ant ellipsoid method. IEEE Transactions on Aerospace
and Electronic Systems, 52(2), 891–903.

250