Upload
elias-j
View
225
Download
0
Embed Size (px)
Citation preview
8/11/2019 Guidance non-linear.pdf
1/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
NONNON--LINEAR CONTROL OFLINEAR CONTROL OFWHEELED MOBILE ROBOTSWHEELED MOBILE ROBOTS
Maria Isabel Ribeiro
Pedro Lima
[email protected] [email protected]
Instituto Superior Tcnico (IST)
Instituto de Sistemas e Robtica (ISR)
Av.Rovisco Pais, 1
1049-001 LisboaPORTUGAL
May.2002
all rights reserved
8/11/2019 Guidance non-linear.pdf
2/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Course Outline
tarefasPLANEAMENTO DE
TAREFAS
PLANEAMENTO DE
MOVIMENTO
AUTO-LOCALIZAO(long baseline, short baseline,feature matching)
MAPA DOMUNDO
FUSOSENSORIAL
CONTROLO
CONDUO
ACTUADORESSENSORES
VECULO
informao processada apartir dos dados dos sensores
actualizaodo mapa
mapaglobal
Informao enviadaaos actuadores
Deteco deobstculos
Posio e/ouvelocidadesdesejadas
(seguir trajectria
/evitar obstculos)
Motores, rodas,hlices, superfciiesde defleco
mapalocal
Trajectria desejada(pos+vel+ace)
Caractersticas domeio envolvente
Trajectria estimada(pos+vel+ace)
Sonar, laser, viso,encoders, giroscpios
8/11/2019 Guidance non-linear.pdf
3/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Guidance
sensor
measurements
operati
onpoint
posture
estimate
joint
state
feedback
PathPlanner
GuidanceJoint
ControllerVehicle
Localization
obstacles
target
pat
h
or
tra
jectory
jo
int
set
po
ints
(e.g.,
whee
l
ve
loc
it
ies
)
jo
in
t
torques
(e.g.,m
otor
inputs
)
GUIDANCEtake the robot from the current posture to the desired
posture, possibly following a pre-determined path ortrajectory, while avoiding obstacles
GUIDANCEGUIDANCEtake the robot from the current posture to the desired
posture, possibly following a pre-determined path ortrajectory, while avoiding obstacles
8/11/2019 Guidance non-linear.pdf
4/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Guidance Methodologies
State(posture)-feedback methods:
posture stabilization (initial and final postures given; no
path or trajectory pre-determined; obstacles notconsidered; may lead to large unexpected paths)
trajectory tracking (requires pre-planned path) virtual vehicle tracking (requires pre-plannedtrajectory)
Potential-Field like methods
potential fields (holonomic vehicles) generalized potential f ields (non-holonomic vehicles) modified potential fields (non-holonomic vehicles)
Vector Field Histogram (VHF) like methods
nearness diagram navigation (holonomic vehicles) freezone (non-holonomic vehicles)
Some Guidance methodologiesSome Guidance methodologies
NON-LINEAR CONTROL DESIGNFOR MOBILE ROBOTS
8/11/2019 Guidance non-linear.pdf
5/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Control of Mobile Robots
Three distintic problems: Trajectory Tracking or Posture Tracking
Path Following
Point Stabilization
8/11/2019 Guidance non-linear.pdf
6/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Trajectory Tracking
Vehicle of unicycle type
The trajectory tracking problem for a WMR of the unicycletype is usually formulated with the introduction of a
vir tual reference vehicle to be tracked.
=
==
!
!
!
sinvy
cosvx
),y,x(z =
v
position and orientation withrespect to a fixed frame
linear velocity
angular velocity
kinematic model
Is a simplified model, but Captures the nonholonomy property which
characterizes most WMR and is the core ofthe difficulties involved in the control of thesevehicles
reference vehicle
ref
x
y
refy
refx
{ }B
8/11/2019 Guidance non-linear.pdf
7/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Trajectory Tracking
refref
refrefref
refrefref
sinvy
cosvx
=
=
=
!
!
!
Kinematic model of the Reference Vehicle
)t(vref )t(ref Bounded Bounded derivatives Do not tend to zero as t tends to infinity
),y,x(z refrefrefref =
Control Objective
Drive the errors refrefref -,yy,xx to zero
Express the errors in the {B} frame
=
=
ref
ref
ref
3
2
1
yy
xx
100
0cossin
0sincos
e
e
e
e
DifferentiatingIntroducing the change of inputs
=+=
ref2
3ref1
u
ecosvvu
8/11/2019 Guidance non-linear.pdf
8/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Trajectory Tracking
+
+
=2
1
ref3u
u
10
00
01
v
0
esin
0
e
100
00
00
e!
non-linear dynamic systemcontrol variables
Question? Is it possible to design a feedback law u=f(e) such that the error
converges to zero? Is this law linear or non-linear ?
Two different solutions:
Linear feedback control
Nonlinear feedback control
8/11/2019 Guidance non-linear.pdf
9/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Trajectory Tracking
LINEAR FEEDBACK CONTROL
+
+
=2
1
ref3u
u
10
00
01
v
0
esin
0
e
100
00
00
e!
0u
0e
==linearize about the
equilibrium point
+
=2
1
ref
ref
u
u
10
00
01
e
100
00)t(
0)t(0
e!
linear time-varying dynamic system
refref
refref
v)t(v
)t(
==
Assuming
linear time invariant dynamic system
Is it possible to design a linear feedback law u=f(e) such thatthe error converges to zero?
Is the dynamic system controllable ?
8/11/2019 Guidance non-linear.pdf
10/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Trajectory Tracking
=
000010
00v00v0001
refref
refref
2
ref
c
0v refref ==If the SLIT is non-controllable
the reference robot at rest
the error cannot be taken tozero in finite time
otherwise
Keu=
=
3
2
1
232221
131211
2
1
e
e
e
KKK
KKK
u
u
ijK Chosen by pole placement
closed loop poles 0)aas2s)(a2s( 22 =+++
undetermined
system
ku1 =
332ref22
111
eke)vsgn(ku
eku
==
a2k
|v|
ak
a2k
3
ref
2r
2
2
1
=
=
=
If vr tends to zero, k2 increases withouth boun
8/11/2019 Guidance non-linear.pdf
11/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Trajectory Tracking
closed loop poles 0)aas2s)(a2s( 22 =+++
332r22
111
eke)vsgn(ku
eku
==
2ref
2ref3
ref2
2ref
2ref1
bvw2k
|v|bk
bvw2k
+=
=
+=
To avoid the previous problem:
The closed-loop poles depend on the values of vrand wr
2ref
2ref bvwa +=
8/11/2019 Guidance non-linear.pdf
12/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Trajectory Tracking
NONLINEAR FEEDBACK CONTROL
+
+
=2
1
ref3u
u
10
00
01
v
0
esin
0
e
100
00
00
e!
3refref32
3
3ref42
1refref11
e)w,v(kee
esinvku
e)w,v(ku
=
=
k4 positive constant
k1 continuos function strictly positive in RxR-(0,0)k3 continuos function strictly positive in RxR-(0,0)
See the analogy with the linear contol
PropertyProperty: This control globally asymptoticallystabilizies the origin e=0
demonstration using Lyapunov stability theory
bk
bvw2)w,v(k)w,v(k
4
2ref
2refrefref3refref1
=
+==
8/11/2019 Guidance non-linear.pdf
13/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Path Following
Objective:Objective:
Steer the vehicle at a constant forward speedalong a predefined geometric path that is givenin a time-free parametrization.
x
y{ }B
"
Path
d
Approach:Approach: The controller should compute
The distance of the vehicle to the path
The orientation error between the vehicles main axis andthe tangent to the path
The controller should act on the angular velocity
to drive both to zero
8/11/2019 Guidance non-linear.pdf
14/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Path Following
x
y{ }B
"
Path
d
M
M is the orthogonal projection of the robots position P on the pathM exists and is uniquely defined if the path satisfies some conditions
),,P( NT Serret-Frenet frame moving along the path. The vector is the tangent vector to the path in the
closest point to the vehicle The vector is the normal
N
N
T
T
is the distance between P and M
s is the signed curvilinear distance along the path, from some initial path tothe point M
d(s) is the angle betwen the vehicles x-axis and the tange to the path at thepoint M.
c(s) is the paths curvature at the point M, assumed uniformly bounded anddifferentiable
is the orientation error
"
d~ =
8/11/2019 Guidance non-linear.pdf
15/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Path Following
A new set of state coordinates for the mobile robot
The error dynamics can be derived writing thevehicle kinematic model in the Serret-Frenet frame:
)~,,s( "
They coincide with x,y, in the particular caswhere the path coincides with the x-axis
"
!
"!
"
!
)s(c1
)s(c~
cosvw
~
~sinv
)s(c1
~cosv
s
=
=
=
PROBLEM FORMULATION
Given a path in the x-y plane and the mobile robot translational velocity,v(t), (assumed to be bounded) together with its time-derivative dv(t)/dt,
the path following problem consists of finding a (smooth) feedbackcontrol law
such that
))t(v,~
,,s(k = "
0)t(~
lim
0)t(lim
t
t
=
=
"
8/11/2019 Guidance non-linear.pdf
16/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Path Following
LINEAR FEEDBACK CONTROL
"
!
"!
"
!
)s(c1
)s(c~
cosvw
~
~sinv
)s(c1
~cosv
s
=
=
=
u~
~sinv
)s(c1
~cosv
s
=
=
=
!
"!
"
!
")s(c1
)s(c~
cosvwu
=
Two different solutions:
Linear feedback control
Nonlinear feedback control
Linearize the dynamics around the equilibrium point )0~,0( =="
)t(u)t(~
)t(~
)t(v)t(
=
=!
"!
Linearization of thelast two equations
8/11/2019 Guidance non-linear.pdf
17/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Path Following
The linear system isCONTROLLABLE
ASYMPTOTICALY
STABILIZABLE BY LINEARSTATE FEEDBACK
Linear stabil izing feedback
= ~|v|kvku 32 " 0k,0k 32 >>
0vk|v|k 223 =++ ""!"!! Closed-loop differential equation
0kk 2'
3'' =++ """
0aas2s 22 =++
=
=
t
0
d|v|
' "
"transformation
Distance gone by point M along the p
Desired closed-loopcharaceristic equation
If
v(t)=v=cte 0
8/11/2019 Guidance non-linear.pdf
18/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Path Following
NONLINEAR FEEDBACK CONTROL
u~
~sinv
)s(c1
~cosv
s
=
=
=
!
"!
"
!
Nonlinear control law
= ~)v(k~
~sin
vku 2 "
with
0k2>
).(k continuous function strictly positive
In order to have the two (linear and nonlinear) controllersbehave similarly near chose0~,0 =="
|v|k)v(k 3=
with
22
3
ak
a2k
=
=
8/11/2019 Guidance non-linear.pdf
19/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Path Following
Property
Under the assumption
0)t(vlimt
the non-linear control
=~
)v(k~
~sin
vku 2 "asymptotically stabilizes )0
~,0( =="
provided that the vehicles initial configuration is such that
2
2
2
2
)s(csup(lim
1)0(
~
k
1)0(
8/11/2019 Guidance non-linear.pdf
20/21
2002 - Pedro Lima, M. Isabel RibeiroRobtica Mvel
Point Stabilization
Given
An arbitrary posture
Find
A control law
which stabilizes asymptotically z-zd about zero,whatever the initial robots posture z(0)
)t,zz(kv
d=
),y,x(zd =
COROLLARY
There is no smooth control law k(z) that can solve thepoint stabilization problem for the considered class ofsystems.
3 ALTERNATIVES
Smooth (differentiable) time-varying nonlinear feedback k(z,t) Piecewise continuous control laws k(z) Time-varying piecewise continuous control laws k(z,t)
8/11/2019 Guidance non-linear.pdf
21/21
References
C. Canudas de Wit, H. Khennouf, C. Samson, O.Sordalen, Nonlinear Control Design for Mobile Robots inRecent Developments in Mobile Robots, World Scientific,1993. Reading assignment
Carlos Canudas de Wit, Bruno Siciliano and Georges
Bastin (Eds), "Theory of Robot Control", 1996.