2

Dynamics and Control for Nonholonomic

Mobile Modular Manipulators
Yangmin Li & Yugang Liu
University of Macau
Macao S. A .R., P. R. China

Open Access Database www.i-techonline.com

1. Introduction
The development of a robot requires that it be able to adopt as many configurations as
possible using limited modules, so as to allow the construction of new types of robots
without redesign and remanufacturing. Traditionally, modular manipulators are mounted
on a fixed base whose mobility is constrained. However, with the development of industry
and technology, such modular manipulators as mounted on fixed bases can not meet some
practical requirements any more. An intelligent and autonomous mobile manipulator,
which can fulfil some operations without human interference, has become an active research
topic recently since it has many potential applications such as in modern factories for
transporting materials, in dangerous fields for dismantling bombs or moving nuclear
infected objects, in modern families for doing housework, as well as in the public places for
city maintenance.
In this chapter, a nonholonomic mobile platform is attached to the modular manipulator in
order to increase workspace of the entire robot. Building up the dynamic model for a
nonholonomic mobile modular manipulator is a challenging task due to the interactive
motions between the modular manipulator and the mobile platform, as well as the
nonholonomic constraints of the mobile platform. Also a trajectory following task becomes
even more complex and difficult to achieve.
Such conventional control strategies as computed-torque control require precise apriori
knowledge of the dynamic parameters for the controlled system. However, in practical
applications, it is almost impossible to obtain exact dynamic parameters for a mobile
modular manipulator because of such uncertainties as complex nonlinear frictions,
flexibilities of the joints and links, payload variations, and terrain irregularities. Robust
control techniques provide a natural rejection to external disturbances, which are
provided by a high-frequency commuted control action that constrains the error
trajectories to stay on the sliding surface. Classical sliding mode control law adopts sign
functions and the caused chattering may do harm to the robots. Adaptive control
technique does not rely on precise apriori knowledge of dynamic parameters and it can
suppress such errors as caused by parameter uncertainties by online adjusting dynamic
parameters. Furthermore, adaptive control can counteract the negative influence of highfrequency switching caused by robust control because its action has naturally smooth time
behaviour.
Source: Mobile Robots, Moving Intelligence, ISBN: 3-86611-284-X, Edited by Jonas Buchli, pp. 576, ARS/plV, Germany, December 2006

www.intechopen.com

30

Mobile Robots, moving intelligence

In related work on modular robots, the modular robot concept could be traced back to the
1970s (Will & Grossman, 1975). In early modular robot research, the emphasis was put on
the structure design of self-organizing, self-reconfigurable, self-assembling, and selfrepairable modular robots (Fukuda et al., 1989; Tomita et al., 1999). Kinematic and
dynamic analysis as well as trajectory planning became another active topic in the past
decades (Chen & Yang, 1998; Fei et al., 2001). In recent years, the scholars had turned their
attentions to trajectory following control for modular manipulators (Melek & Goldenberg,
2003; Shen et al., 2002; Stoy et al., 2002). Parameter identification and vibration control for
a 9-DOF reconfigurable modular manipulator were investigated by authors in (Li et al.,
2004a).
Regarding to literatures on mobile manipulators, mobile manipulators were exploited to
install and remove aircraft warning spheres (Campos et al., 2002), to polish aircraft
canopy (Jamisola et al., 2002), to organize furniture in a room (Rus et al., 1995), and to
collectively transport a single palletized load (Stilwell & Bay, 1993). A great deal of
research activities can be found on motion planning of mobile manipulators (Carriker et
al., 1991; Chitta & Ostrowski, 2002; Nagatani et al., 2002). Several kinematic and dynamic
modelling methods were presented for mobile manipulators in the past decade, such as
the Kanes method (Tanner & Kyriakopoulos, 2001), the Newton-Euler method (Chung &
Velinsky, 1999) and the Lagrange method (Li & Liu, 2004b; Liu & Li, 2005a; Yu & Chen,
2002). Tip-over analysis and prevention attracted numerous scholars and several tip-over
stability criteria were defined, such as the potential energy stability level (Ghasempoor &
Sepehri, 1995), the force-angle stability measure (Papadopoulos & Rey, 1996), the zero
moment point criterion (Furuno et al., 2003), and the criterion based on supporting forces
(Li & Liu, 2005b). Extensive literatures can be found on control of mobile manipulators.
Dynamic characteristics between the mobile platform and the onboard manipulator were
investigated (Yamamoto & Yun, 1996). A robust control method was developed to
eliminate the harmful effect of the wheel slip on the tracking performance of a spatial
mobile manipulator (Chung & Velinsky, 1999). A homogeneous kinematic stabilization
strategy and an adaptive control scheme were combined for mobile manipulator control
without any knowledge of the system dynamic model (Colbaugh, 1998). Neural network
and fuzzy logic control for mobile manipulators were also studied by authors (Li & Liu,
2005c, 2005d, 2006a, 2006b).
In the previous research work, modeling for the mobile platform and for the manipulator
was usually carried out separately and control for nonholonomic mobile robots was
mostly limited to kinematic velocity control, while few work on dynamic torque control.
The interactive motions between the manipulator and the mobile platform made the
models established inaccurate, which then affected the control result. Most of the present
controllers were designed in joint space, but few in task space (Ge et al., 1997). However,
in practical applications, the end-effector of a robot is usually specified to fulfil some
operations.
This chapter is organized as follows: an integrated modelling method is proposed
considering nonholonomic constraints and interactive motions in Section 2. In Section 3, a
robust adaptive controller is designed in task space to control the end-effector to follow
desired spatial trajectories. Simulations are conducted on a real mobile modular
manipulator, and a comparison is made with the conventional model-based controller in
Section 4. Section 5 gives some concluding remarks. Some suggested ideas for future
research work are presented in the last section.

www.intechopen.com

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

31

2. An Integrated Modelling Method

2.1 Kinematics analysis
A mobile modular manipulator is normally composed of an m-wheeled holonomic or
nonholonomic mobile platform and an n-DOF onboard modular manipulator, as shown in
Fig. 1(a). In this chapter, we analyze the 3-wheeled nonholonomic mobile platform, which
has two driving wheels and one castor wheel. The two driving wheels are coaxial and
mounted in front of the platform and the castor wheel in the rear is orientable with respect
to the cart. The robot is assumed to move on a horizontal plane; then the motion of the
mobile platform can be illustrated as shown in Fig. 1(b). An arbitrary inertial base frame
OB X BYB Z B is fixed on the motion plane, while a frame Om X mYm Z m is attached to the mobile
platform. In frame Om X mYm Z m , Om ( xm , ym ) is selected as the midpoint of the line segment

connecting the two driving-wheel centres; OmYm is along the coaxial of the two driving

wheels; Om X m is perpendicular to OmYm and passes through Om . The heading angle m

determines the posture of the mobile platform.

end-effector

modular
manipulator

driving
wheel

castor
wheel

(a) Prototype

YB

Ym

driving
wheel

dm

castor
wheel

driving
wheel

modular
manipulator

Om ( xm , ym )

rf
R

Xm

lG

XB

OB
(b) coordinate system definition
Fig. 1. Prototype and coordinate system definition for a mobile modular manipulator.

www.intechopen.com

32

Mobile Robots, moving intelligence

Since the modular manipulator is mounted on the mobile platform, it can be assumed that
the relative position of frame Om X mYm Z m and the frame O0 X 0Y0 Z 0 is steady; here O0 X 0Y0 Z 0 is
the frame attached to the 1st module of the modular manipulator. Therefore the mobile
platform can be treated as a special module added to the bottom of the modular
manipulator, which can both move on the motion plane and rotate about the vertical axis.
From Fig. 1(b), transformation matrices between the frames OB X BYB Z B , Om X mYm Z m and
O0 X 0Y0 Z 0 are given by

cosm
s in
m
B

mT =

s in m
cosm
0
0

0
0
1
0

where rf is the radius for the driving wheels;

xm
ym
rf
1

1
0
m

0T =
0

0
1
0
0

0
0
1
0

lG
0
hG
1

(1)

lG and hG denote offsets of the 1st module with

respect to Om X mYm Z m along Om X m , OmYm respectively.

By defining the link four parameters according to the Denavit-Hartenberg (D.H.) notation
(de Wit et al., 1996), the transformation matrix between two adjacent links for the modular
manipulator can be derived as follows

cos qi
sin q cos
i
i
i 1

iT =
sin qi sin i

sin qi
cos qi cos i
cos qi sin i
0

0
sin i
cos i
0

li

di sin i

di cos i

(2)

where li , i , di , qi are D.H. parameters.

Therefore the transformation matrix of the end effector with respect to the inertial base
frame can be derived by

T = B0T 01T 21T L nn1T

B
n

(3)

Hence, the position vector pre and the posture vectors nr, or, ar of the end-effector with respect
to the inertial base frame can be derived, which can be observed from the fourth column and
the first three columns of the matrix BnT correspondingly.
However, it is inconvenient to describe the posture of the end-effector using these posture
vectors with nine parameters. The posture can be determined by only three independent
parameters using Z-Y-Z Euler angles , and . Relationship between the posture vectors

= a tan 2 ( a y , ax ) ,

and the Euler angles can be described as follows

= a tan 2 ( oz , nz ) ,

= a tan 2 ( ax cos + a y sin , az )

www.intechopen.com

(4)

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

33

2.2 Nonholonomic constraints and Jacobian

The motion of a mobile modular manipulator during a very short time interval
[t i t i +1 ] is shown in Fig. 2. lr , dr and d m represent the distances from the fixed bar to

OmYm , from the fixed bar to the horizontal axis of the castor wheel, and that between
the two driving wheels respectively; ri and i are steering radius and yaw angle during

the time interval; L , R , r and r denote rotating angles of the left driving wheel, the
right driving wheel, the castor wheel around its horizontal axis, and that around the
fixed bar respectively; S L , S R and Sm represent advances of the left driving wheel,
the right driving wheel, and the mobile platform during the time interval
correspondingly.
YB

Xm

L
Omi +1 ( xmi +1 , ymi +1 )

S L d r r

ri

S m

Ym

driving
wheel

X m S R

rf
lr

rr

mi

dm

castor driving
wheel wheel

mi +1

Ym

lG

Omi ( xmi , ymi )

XB

OB

Fig. 2. Nonholonomic constraints derivation.

Let the time interval t = t i +1 t i 0 , on the assumption of low speeds, velocities during
this time interval can be treated as constants. Then,

xmi

ymi

S L

S R

S
m

i
r
i

www.intechopen.com

= Omi Omi +1 cos mi

= Omi Omi +1 sin mi

= i ri m
2

= i ri + m
2

Omi Omi +1

=

i
m

Sm

(a)
(b)
(c)
(5)

(d)
(e)
(f)
(g)

34

Mobile Robots, moving intelligence

where xmi = xmi +1 xmi , ymi = ymi +1 ymi and mi = mi +1 mi .

Substituting Eq. 5(f, g) into Eq. 5(c, d) yields

dm

i
S L = S m 2 m

S = S + d m i
m
m
R
2

Adding Eq. 6(a) and Eq. 6(b) yields

Sm =
Subtracting Eq. 6(a) for Eq. 6(b) yields

m =

(a )

(6)

(b )

S L + S R
2

(7)

SL SR
dm

(8)

Substituting Eq. 7 and Eq. 5(e) into Eq. 5(a,b) yields

c o s mi
i
( S L + S R )
x m =
2

i
y i = s in m ( S + S )
m
L
R

(9)

From Eq. 8 and Eq. 9,

i
i
xmi cosmi SL
SR cosm & &
lim
SL + SR
x&m = lim

=
+ lim

=
t 0 t
t
t

0
2 t
2
t

ymi sinmi SL
S sinm & &
i
=
lim
+ lim R =
SL + SR
y&m = lim

t 0
2 t0 t t0 t
2
t

i
S
S
1
1
&mi = lim m = lim L lim R = S&L S&R
t

0
t
t

0
0
dm t
dm
t
t

(10)

S&L = rf &L and S&R = rf &R , we can obtain

Equation 10 always holds during the course of plane motion, so the superscript i can be
omitted. Considering that

rf cos m &
r cos m &

L + f
R
x& m =
2
2

r f s in m &
r s in m &

L + f
R
y& m =
2
2

rf
rf
&
&
&
m = d L + d R
m
m

www.intechopen.com

(a )
(11)

(b )
(c )

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

In the same way, we can obtain

&
co s ( r + m
r =
rr

& = s i n ( r + m
r
dr

) x&

) x&

Adding Eq. 11(a) multiplied by

sin ( r + m
rr

co s ( r + m
dr

y& m
y& m

lr s in r &
m
rr

d r lr c o s r &
m
dr

rf
2

&L +

rf
2

&R

(13)

dm
to Eq. 13 yields
2

cos m x& m + sin m y& m +

Subtracting Eq. 11(c) multiplied by

Subtracting Eq. 11(b) multiplied by

dm &
m r f &R = 0
2

dm &
m r f &L = 0
2

(15)

cos m from Eq. 11(a) multiplied by sin m yields

s in m x& m c o s m y& m = 0

(16)

From Eq. 12 and Eq. 14-16, we have

sin m
m
cosm
2

d
m
cos
sin m
m

2
sin
0
cosm
m

cos ( r + m ) sin ( r + m )
lr sin r

sin ( r + m ) cos ( r + m ) lr cos r dr

in short A ( ) & = 0 ; here = [ xm

(14)

dm
from Eq. 13 yields
2

c o s m x& m + s i n m y& m

ym

m r

From Eq. 11 and Eq. 12, we can easily obtain

www.intechopen.com

(12)

cos m to Eq. 11(b) multiplied by sin m yields

c o s m x& m + s in m y& m =
Adding Eq. 11(c) multiplied by

35

rf

0
rr

0
0

0
0

d r

T
r L R ] .

x&m
0
y& m
&

rf m
& = 0
r
0 &

r
0 &
L
0 &
R

(17)

36

Mobile Robots, moving intelligence

rf cosm

rf sinm

x&m
2
y&
r
m
f
&m
dm
&
r = rf ( dm cos r + 2lr sin r )
&r
2dmrr
&
L r ( d sin + 2d 2l cos )
r
r
r
r
& f m
R
2dmdr

in short & = S ( ) & & T .

R
L
Substituting Eq. 18 into Eq. 17 yields

A (

Let x = [ px

py

rf cosm

rf

&
dm
L
rf ( dm cos r 2lr sin r ) &
R
2dmrr

rf ( dm sin r 2dr + 2lr cos r )

2dmdr

2
rf sin m

) S ( ) =

pz ]T and q = [ L R

q1

and joint space respectively; and define = [ T

then the Jacobian matrix can be derived by

J =

(18)

(19)

q2 L qn ]T be coordinates in task space

q1 L qn ]T as extended coordinates, and

x
x S ( )
=

q
0 n 2

0 7 n

I n n

(20)

2.3 Dynamic modelling

With the assumption of moving on a horizontal plane, the mobile platform has a constant
potential energy U m . To calculate the kinetic energy, the mobile platform can be divided into
four parts, the cart (including all the driving units in the box), the left front wheel, the right
front wheel and the rear wheel, that is

T m = T c + T lf + T r f + T r

where

Tm is

(21)

the entire kinetic energy for the mobile platform; Tc , Tlf , Trf and Tr denote

respectively kinematic energy of the cart, the left and right front wheel and the rear wheel.
Kinetic energy for the cart can be calculated by

Tc =

where

mc

1
1
1
mc x&c2 + mc y& c2 + I c &m2
2
2
2

(22)

is the mass of the cart; xc = xm + lc cos m , yc = ym + lc sin m denote coordinates

for the mass centre of the cart, with

lc

being distance between the mass centre and Om ;

represents the moment of inertia for the cart around the axis of Om Z m .

www.intechopen.com

Ic

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

37

With the assumption of rolling without slipping, the contact points of the wheels with the
motion plane can be treated as instant rotating centres. Then, kinematic energy for the front
driving wheels can be given by

3
Tlf = m f rf2 &L2 ,
4

where

3
Trf = m f rf2 &R2
4

(23)

m f is the mass for the front driving wheels.

The castor wheel can rotate about both its axis and the fixed bar that does not pass through
its centre, as shown in Fig. 3. So, the kinetic energy of the rear castor wheel includes two
parts as given by Eq. 24.

Tr =

1
1
I r1 &r2 + I r 2 &r2
2
2

(24)

where the moments of inertia I r1 and I r 2 can be determined by

I r1 =

1
3
mr rr2 + mr rr2 = mr rr2
2
2
r2

1
I r 2 = mr rr2 + mr d r2 = mr r + d r2
4
4

castor
wheel

dr

driving
wheel

(25)

Ym

fixed bar

Xm

lr

dm

dr

driving
wheel

Fig. 3. Structure for the castor wheel.

Kinematic energy for the onboard modular manipulator can be calculated by
n
1
1

TM = vciT mi vci + iT I i i
2

i =0 2

(26)

vci and i represent the linear and angular

where TM denotes kinetic energy for the modular manipulator; miand Ii are mass and inertial
moment matrix for the ith module respectively;
velocities for the mass centre of the ith module.

www.intechopen.com

38

Mobile Robots, moving intelligence

U M = {mi g T pci }

Potential energy for the onboard modular manipulator can be calculated by

n

where g = [0

0 9.81]T and pci = [ xiB

i =0

yiB

(27)

ziB ]T .

Then, the Lagrange function can be given by

L = Tm U m + TM U M

(28)

The constraint dynamics for the nonholonomic mobile modular manipulator can be
determined by

d L L
= 1 cosm + 2 cosm + 3 sinm + 4 cos ( r + m ) + 5 sin ( r + m )

dt x&m xm

d L L = 1 sinm + 2 sinm 3 cosm + 4 sin ( r + m ) 5 cos ( r + m )

dt y&m ym

dm
dm
d L L
dt & = 1 2 + 2 2 4lr sin ( r ) + 5 ( lr cos r dr )
m
m
d L L
= 4rr
&
dt r r

d L L = d
5 r
dt &
r
r
d L L
xT
r
Fext

+
+
&
L
1 f
L
dt L L

T
(29)
d L L = 2rf + R + x Fext
dt &R R
R

xT
d L L
Fext , ( i = 1,L, n)

=
+

i
dt q&
qi
i qi

where

i ( i = 1,L ,5 ) are Lagrange multipliers associated with the nonholonomic

constraints, and Fext 6 is a vector for external forces; L , R , i represent driving torques

of the left wheel, the right wheel and the ith joint for the onboard modular manipulator
respectively.
Equation 29 can be written into the matrix form

d L L

= B ( ) ( + J T Fext ) + C ( )
dt &
T

www.intechopen.com

(30)

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

where B ( ) = 0( n + 2)5

I ( n + 2)( n + 2)

39
T
= [ L R 1 L n ] ,

, C ( ) = A ( ) 05n T ,

T
= [ 1 L 5 ] .
Equation 30 can be rewritten into the following standard form

M ( ) && + V ( , & ) & + G ( ) = B ( ) ( + J T Fext ) + C ( )

where

is the inertial matrix;

represents the centrifugal and coriolis matrix; and

G denotes the gravitional force vector.

M is symmetric and positive semi-definite, i.e.

Remark 1: Matrix

MT = M 0.

Remark 2: Matrix M& 2V is skew-symmetric, i.e., for any r n + 7 ,

From Eq. 20,

S ( )
0 n 2

( ) q&& + V ( , & ) q&

V = S T V S + M S&

(32)

+ G (

Substituting Eq. 32 and its derivative into Eq. 31 yields

where M = S T M S ,

r T ( M& 2V ) r = 0

0 7 n
q&
I n n

& =

(31)

)=

, G = S T G , = + J T Fext ;

T
S T B = I ( n + 2)( n + 2) and S C = 0 n + 2 are eliminated.

(33)
the

terms

If n <

4 , there are some dead zones that the end-effector can not reach; on the other hand, if
n > 4 the entire mobile modular manipulator will be a redundant one; this two cases are

beyond the disscussion of this chapter. Assume the Jacobian matrix is full rank, soving the
differential kinematics x& = J q& and its derivative, yields

q& = J 1 x& ,

q&& = J 1 &&
x J 1 J& J 1 x&
J T yields

( ) &&x + V% ( , & ) x&

Substituting Eq. 34 into Eq. 33 and left multiplying

M%

+ G% (

(34)

) = %

M% = J T M J 1 , V% = J T (V M J 1 J& ) J 1 , G% = J T G and % = J T
% is symmetric and positive definite, i.e., M% T = M% > 0 .
Remark 3: Matrix M
where

Proof: From Eq. 33 and Eq. 35,

M% = J

According to Remark 1,
M%

= J

M% = J

www.intechopen.com

S J

M S J

M S J
= J

(S

(36)

M S J

(35)

(S

1
1

= M%

)>

(37)

40

Mobile Robots, moving intelligence

T
%&
%
n+7
Remark 4: Matrix M%& 2V% is skew-symmetric, i.e., for any r , r M 2V r = 0 .

Proof: From Eq. 33 and Eq. 35,

J T M% J = S

M S

M% J& = S

M& S + 2 S

(38)

Differentiating Eq. 38 yields

M&% J + 2 J

M S&

(39)

Then
M%& = J

M& S J

+ 2J

V S J

+ J

M S& J

2 M% J& J

M S& J

M% J& J

(40)

From Eq. 33 and Eq. 35

V% = J

Subtracting Eq. 41 multiplied by 2 from Eq. 40 yields

&
M% 2 V% = J

According to Eq. 41
r

&
M% 2 V%

) r

(S

( M& 2 V

) ( M&
T

) S

2V

) (S

(41)

(42)

)=

Remark 5: Matrices M% , V% and G% are all bounded as long as J keeps nonsingular.

(43)

independent of the strucuture parameters , M% && + V% & + G% = Y ( , q& , &&, & ) . Here
Remark 6: The dynamic equation in Eq. 35 is linearity in parameters, i.e., for any

Y ( , q&, &&, & ) is also independent of .

3. Robust Adaptive Controller Design

3.1 A model-based controller
Assume xd , x&d and &x&d are the desired end-effector positions, velocities and accelerations. If
precise dynamic parameters can be obtained in advance, the model-based controller can be
exploited, as shown in Fig. 4 and Eq. 44.

= J T {M% &&
xd + K D ( x&d x& ) + K P ( xd x ) + V% x& + G% Fext }

where K P > 0 and K DT = K D > 0 are proportional and differential gain matrices.

www.intechopen.com

(44)

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

Fext

xd

+ +

KD
xd

xd

KP

JT

41

x x

Fig. 4. A model-based controller.

Substituting Eq. 44 into Eq. 35 yileds

&&
e + K D e& + K P e = 0

where

e = x xd is the error vector.

(45)

Theorem 1: If K P > 0 and K DT = K D > 0 , the system described in Eq. 45 is asymptotically

stable, i.e., e 0 as t .
Proof: Considering the nonnegative Lyapunov candidate as follows

1
1
VS = e&T e& + eT K P e 0
2
2

(46)

The time derivative of Lyapunov candidate is

V&S = e&T e&& + e&T K P e

(47)

Substituting Eq. 45 into Eq. 47 yields

V&S = e&T K D e& 0

From Eq. 46 and Eq. 48, VS is a Lyapunov function. Notice that when and only when

(48)

e& = 0 ,

V&S = 0 may hold, so e& 0 as t + . To analyze the steady-state error of the system,
substituting e& = 0 and e&& = 0 into Eq. 45 yields K P e = 0 , which is followed by e = 0 . That is

to say, e 0 as t + . That is all for the proof.

www.intechopen.com

42

Mobile Robots, moving intelligence

3.2 A robust adaptive controller

To avoid measuring accelerations, the error system can be defined as follows

e ( t ) = x ( t ) xd ( t )

x&s ( t ) = x&d ( t ) e ( t )

(49)

s ( t ) = x& ( t ) x&s ( t )

where s ( t ) is the tracking error measure, and 66 is a positive definite matrix.

From Eq. 49,

x& ( t ) = s ( t ) + x&s ( t )

&&
xs ( t ) = &&
xd ( t ) e& ( t )

(50)

&&
x ( t ) = s& ( t ) + &&
xs ( t )

Substituting Eq. 50 into Eq. 35 yields

M% s& ( t ) + V% s ( t ) + M% &&
xs ( t ) + V% x&s ( t ) + G% = %

From Remark 6,

(51)

M% &&
xs ( t ) + V% x&s ( t ) + G% = Y ( , q& , &&
xs , x&s )

(52)

Substituting Eq. 52 into Eq. 51 yields

M% s& ( t ) + V% s ( t ) + Y ( , q& , &&

xs , x&s ) = %

(53)

be the estimate of the structure parameter , the robust adaptive controller

Let
presented in this chapter is given by Eq. 54, and a control system diagram is shown in
Fig. 5.

K s ( t ) K s ( t ) dt ( s ) F
= J T Y ( , q& , &&
xs , x& s )
P
I
r
ext
t

(54)

where the 1st term forms an adaptive controller; the 2nd and 3rd terms form PI controller;
r ( s ) is the robust sliding mode term, with its element determined by

where kri > 0 is a constant.

www.intechopen.com

ri ( si ) = kri e

1
si

sgn ( si )

(55)

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

q&
x
+

xd

x&d
+
-

e&

Y ( , q, xs , xs )

x&s
&x&s

Kr
x&s

x&

KP

+
-

JT
Fext

x&

Fig. 5. A robust adaptive controller.

ri

ri

ri

Fig. 6. A new sliding mode function.

www.intechopen.com

KI
&x&s

q&

Adaptive
Law

&x&d

JT

43

dis

44

Mobile Robots, moving intelligence

The sliding mode control law is not only continuous but also infinitely differentiable, which
can eliminate the chattering caused by the classical sign function effectively. The output
varies according to

si

. The funciton is plotted in Fig. 6.

Substituting Eq. 54 into Eq. 53 and considering Eq. 35 at the same time yields

t
+ = 0
M% s& ( t ) + V% s ( t ) + K P s ( t ) + K I s ( t ) dt + Y ( , q& , &&
xs , x&s )
r
0

(56)

Theorem 2: The closed-loop system in Eq. 56 is asymptotically stable under the adaptation
law by Eq. 57. The error signals are convergent with time, i.e., e ( t ) , e& ( t ) 0 as t .
Furthermore, the signals in the system are all bounded.

&
= Y

( , q& , &&x s , x& s ) s

(57)

where T = > 0 is a constant matrix.

Proof: Considering the nonnegative Lyapunov candidate as follows

T
t
t
1
1
1 % T 1 %
VS = s ( t ) dt K I s ( t ) dt + sT ( t ) M% s ( t ) +

0
0

2
2
2

(58)

% =
.
where
Time derivative of the Lyapunov candidate is

t
1
&
%& T 1
%
V&S = sT M% s& + K I s ( t ) dt + sT M% s +

2
0

(59)

From Eq. 56, we can obtain

t
% s T
sT M% s& + K I s ( t ) dt = sT V% s sT K P s + sT Y ( , q& , &&
xs , x& s )
r
0

(60)

Substituting Eq. 60 into Eq. 59 and considering Remark 4 at the same time yields

% sT +
%& T 1
%
V&S = sT K P s + sT Y ( , q& , &&
xs , x&s )
r

(61)

&
&% =
, substituting Eq. 57 into Eq. 61 yields
Considering that

V&S = sT K P s sT r

(62)

Substituting Eq. 55 into Eq. 62 yields

1
6

s
V&S = sT K P s kri e i si sgn ( si ) 0
i =1

(63)

is the pnorm space, from Eq. 58 and Eq. 63, s ( t ) l 2 . From

According to LaSalles theorem, we can conclude that the system is asymptotically stable
and s 0 as t + . Let

www.intechopen.com

lp

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

45

% l can
Eq. 50, e ( t ) l 2 l , e& ( t ) l 2 and e ( t ) 0 as t + . Because VS 0 and V&S 0 ,

l ; therefore from the error equation, we can

be obtained; from Remark 5, l , so

conclude that s& ( t ) l . Since s ( t ) l 2 and s& ( t ) l , s 0 as t + , which is followed by

e& ( t ) 0 . This is the end of the proof.

4. Simulation Results
The mobile modular manipulator exploited in this simulation is made up of a 4-DOF
onboard modular manipulator and 3-wheeled nonholonomic mobile platform, as shown in
Fig. 1(a). The modular manipulator consists of 4 rotation modules named PowerCube
produced by the AmtecGmbH Corporation of Germany. The Pioneer3-DXe produced by
ActiveMedia Corporation of USA is used as the mobile platform.
According to the Denavit-Hartenberg notion, the simplified model of the mobile modular
manipulator is drawn in Fig. 7.
X 2 , Y3 , Z 4 , Z E

Y4 , YE

D. H. Parameters
d 1 = 0.250 m

d 2 = 0.000 m
d 3 = 0.000 m

d 4 = 0.225 m

l1 = 0.000 m

l 2 = 0.000 m

1 = 0

2 =
3 =
4 =

X4, XE

X3

Z3

X 1 , Y2

Y1 , Z 2

Z 0 , Z1

l3 = 0.140 m

lG

l 4 = 0.000 m

Ym

Zm

Y0

Xm
R

Om

Fig. 7. A simplified model for the mobile modular manipulator.

Because of page limitations, the kinematics and dynamics model will not be detailed.
To ensure the controllers valid, the desired end-effector trajectory should be selected as far
away from singularities. The Jacobian matrix is derived from Eq. 20 as follows

www.intechopen.com

46

Mobile Robots, moving intelligence

J11
J
21
0

rf
J = dm

J12
J 22

J13
J 23

0
rf

rf

J12 =

rf

cos m
cos m +

l3 cos q2 + d 4 cos q2 sin q3

d 4 sin q2 cos q3

cos q2
cos 2 q3 + cos 2 q2 sin 2 q3

cos q2 sin q3

sin q2 cos q3

cos 2 q3 + cos 2 q2 sin 2 q3

cos 2 q3 + cos 2 q2 sin 2 q3

cos q3
cos q2 + sin 2 q2 cos 2 q3

sin q2 cos q2 sin q3

cos 2 q2 + sin 2 q2 cos 2 q3

where

J11 =

J15
J 25

sin q2 sin q3 cos q3

cos 2 q3 + cos 2 q2 sin 2 q3

dm

J14
J 24

rf
dm
rf

( J13 lG sin m ) , J 21 =

( J13 lG sin m ) , J 22 =

rf
2

sin m
sin m +

rf

rf
dm

0
0

( J 23 + lG cos m ) ,

rf

( J 23 + lG cos m ) ,

J13 = l3 sin ( q1 + m ) cos q2 + d 4 sin ( q1 + m ) cos q2 sin q3 + d 4 cos ( q1 + m ) cos q3 ,

2

dm

(64)

dm

(65)

J 23 = + l3 cos ( q1 + m ) cos q2 d 4 cos ( q1 + m ) cos q2 sin q3 + d 4 sin ( q1 + m ) cos q3 ,

J14 = l3 cos ( q1 + m ) sin q2 + d 4 cos ( q1 + m ) sin q2 sin q3 ,
J 24 = l3 sin ( q1 + m ) sin q2 + d 4 sin ( q1 + m ) sin q2 sin q3 ,

J15 = d 4 cos ( q1 + m ) cos q2 cos q3 d 4 sin ( q1 + m ) sin q3 ,

J 25 = d 4 sin ( q1 + m ) cos q2 cos q3 + d 4 cos ( q1 + m ) sin q3 .

The determinant of Jacobian matrix is

det ( J ) =

rf2 lG l3 sin q2 cos q2 cos q3

d m cos q3 + cos q2 sin q3

2

(66)

So, to avoid singularities, the boundary of joint angles in the joint space can be selected as
follows

= L , R , q1 , q 2 , q3 , q 4 0 < q 2 < , < q3 <

2
2
2

(67)

Simulations are conducted for two control schemes, i.e., the model-based controller (MBC)
and the robust adaptive controller (RAC). To examine the disturbance suppression
characteristics, a series of disturbance torques are introduced. Both the MBC and the RAC
are required to control the end-effector to follow a sine like spatial trajectory as shown in Fig.
8, which has been planned to ensure the system far away from singularities. All the joints
and velocities are initialized to be zeros, except that q2 = 3 . Simulation time interval is
selected as 10 seconds. The gain matrices for the RAC are selected as follows

www.intechopen.com

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

47

K P = diag {100} , K I = diag {10.0} , K r = diag {10.0} ,

= diag {2.0} ,
= diag {0.1}.

(68)

For the MBC, nominal dynamic parameters are adopted, which are assumed to be deviated
from the real values by 10%; the gain matrices are selected as K P = diag {100} , K D = diag {10.0} .
Simulation results are presented by Fig. 8-14. The desired and controlled locus for the MBC
and the RAC are shown in Fig. 8(a) and Fig. 8(b); Figure 9(a) and Fig. 9(b) present the
tracking position errors; tracking Euler angular errors are shown in Fig. 10(a) and Fig. 10(b);
Figure 11(a) and Fig. 11(b) give the tracking linear velocity errors; tracking Euler angular
velocity errors are given by Fig. 12(a) and Fig. 12(b); time-varying control torques for the
MBC and the RAC are shown in Fig. 13 and Fig. 14.
0.6

0.6

RACcontrolled

MBCcontrolled

0.4

pz, m

pz, m

0.4
0.2

0.2
desired

desired

0
10

0
10
5
p ,m

0
px, m

10

10

5
py, m

0
5

10

0
px, m

(a) Locus for the MBC

(b) Locus for the RAC
Fig. 8. Desired and controlled locus for the MBC and RAC.
0.4

0.2

Tracking position errors for RAC, m

Tracking position errors for MBC, m

0.4

ez
ey

0
ex

0.2

0.4
0

10

Time, s

ez

0.2

0
ey
ex

0.2

0.4

0.6
0

Time, s

10

1.5
1

0.5
0
0.5
e

1.5
2
2.5

Time, s

10

Tracking Euler angular errors for RAC, rad

Tracking Euler angular errors for MBC, rad

(a) Position errors for the MBC

(b) Position errors for the RAC
Fig. 9. Tracking position errors for the MBC and RAC.
1
e

0.5
0
e

0.5
1

1.5
2
2.5

6
Time, s

10

(a) Euler angular errors for MBC

(b) Euler angular errors for RAC
Fig. 10. Tracking Euler angular errors for the MBC and RAC.

www.intechopen.com

10

48

Mobile Robots, moving intelligence

ex

0.5

ez

0
0.5
1
1.5
0

10

1.5

Tracking linear velocity errors for RAC, m/s

Tracking linear velocity errors for MBC, m/s

1.5

ex

0.5

ey
0

ez

0.5
1
1.5
0

Time, s

6
Time, s

10

15

10

10

Time, s

10

Tracking angular velocity errors: RAC, rad/s

Tracking angular velocity errors: MBC, rad/s

(a) Liner velocity errors for MBC

(b) Liner velocity errors for RAC
Fig. 11. Tracking linear velocity errors for the MBC and RAC.
3
2

1 e

1
2

e
0

10

Time, s

10

5
, N m

10

, N m

(a) Euler angular velocity errors for MBC (b) Euler angular velocity errors for RAC
Fig. 12. Tracking Euler angular velocity errors for the MBC and RAC.

5
10

4
6
Time, s

10

10

10

Time, s
20

, N m

20

1, N m

20

20
0

4
6
Time, s

10

20

4
6
Time, s

10

10

0.5

, N m

, N m

10

10
20

4
6
Time, s

10

0.5

Fig. 13. Time-varying control torques for the MBC.

www.intechopen.com

6
Time, s

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

49

Remark 7: From the simulation results, we can conclude that the proposed algorithms are
effective to control the end-effector to follow some definite spatial trajectories. The robust
adaptive controller behaves better than the model based controller, especially when
parameter uncertainties and external disturbances exist.

5. Conclusions
In this chapter, 3-wheeled nonholonomic mobile modular manipulators are investigated.
First, an integrated modelling method is proposed in consideration of the nonholonomic
constraints and the interactive motions. Second, a model based controller and a robust
adaptive controller are presented in task space directly. Third, simulations are carried out on
a mobile modular manipulator composed of a 3-wheeled mobile platform and a 4-DOF
onboard modular manipulator, and the simulation results demonstrate the effectiveness of
the proposed algorithms. The proposed algorithms can be easily extended to some other
mobile manipulators as well.
10

, N m

L, N m

10

10

Time, s
15

2, N m

1, N m

10

10

10

10

10

Time, s

4
Time, s

4, N m

10

3, N m

10

10

6
Time, s

0.2

0.4

10

10

Time, s

Fig. 14. Time-varying control torques for the RAC.

www.intechopen.com

0.6

Time, s

50

Mobile Robots, moving intelligence

6. Future Research Work

In this chapter, although the number of joints for the onboard modular manipulator is just
assumed to be 4, the proposed method can be applied to m-DOF manipulator atop n-DOF mobile
platform system in principle. The modelling and control for a redundant mobile modular
manipulator is more challenging since self-motions have to be taken into consideration, (Li & Liu,
2006b; Maciejewski & Klein, 1989). The proposed algorithms are just verified by simulations, there
is still a lot of work to do to verify the algorithms by real experiments. The task space information
in this simulation is calculated via forward kinematics and differential kinematics, how to acquire
task-space information directly and precisely is also a complex work, which needs deep research
work on sensory technology and method.

7. Acknowledgement
The authors appreciate the fund support from the research committee of University of
Macau under grant no.: RG082/04-05S/LYM/FST.

8. References
Campos, M.; Pereira, G.; Vale, S.; Bracarense, A.; Pinheiro, G. & Oliveira, M. (2002). A
mobile manipulator for installation and removal of aircraft warning spheres on
aerial power transmission lines. Proceedings of IEEE International Conference on
Robotics and Automation, pp. 3559-3564, Washington, DC, USA.
Carriker, W.; Khosla, P. & Krogh, B. (1991). Path planning for mobile manipulators for
multiple task execution. IEEE Trans. on Robotics and Automation, Vol. 7, pp. 403408.
Chen, I. & Yang, G. (1998). Inverse kinematics for modular reconfigurable robots. Proceedings
of IEEE International Conference on Robotics and Automation, pp. 1647-1652, Leuven,
Belgium.
Chitta, S. & Ostrowski, J. (2002). Motion planning for heterogeneous modular mobile
systems. Proceedings of IEEE International Conference on Robotics and Automation, pp.
4077-4082, Washington, DC, USA.
Chung, J. & Velinsky, S. (1999). Robust control of a mobile manipulator-dynamic modelling
approach. Proceedings of the American Control Conference, pp. 2435-2439, San Diego,
California, USA.
Colbaugh, R. (1998). Adaptive stabilization of mobile manipulators. Proceedings of the
American Control Conference, pp. 1-5, Philadelphia, Pennsylvania, USA.
de Wit, C.; Siciliano, B. & Bastin, G. (1996). Theory of Robot Control, Springer-Verlag.
Fei, Y.; Zhao, X. & Song, L. (2001). A method for modular robots generating dynamics
automatically. Robotica, Vol. 19, pp. 59-66.
Fukuda, T.; Nakagawa, S.; Kawauchi, Y. & Buss, M. (1989). Structure decision method for
self organizing robots based on cell structures-cebot. Proceedings of IEEE
International Conference on Robotics and Automation, pp. 695-700, Scottsdale, AZ,
USA.
Furuno, S.; Yamamoto, M. & Mohri, A. (2003). Trajectory planning of mobile manipulator
with stability considerations. Proceedings of IEEE International Conference on Robotics
and Automation, pp. 3403-3408, Taipei, Taiwan, China.

www.intechopen.com

Dynamics and Control for Nonholonomic Mobile Modular Manipulators

51

Ge, S.; Hang, C. & Woon, L. (1997). Adaptive neural network control of robot manipulators
in task space. IEEE Trans. on Industrial Electronics, Vol. 44, No. 6, pp. 746-752.
Ghasempoor, A. & Sepehri, N. (1995). A measure of machine stability for moving base
manipulators. Proceedings of IEEE International Conference on Robotics and Automation,
pp. 2249-2254, Nagoya, Aichi, Japan.
Jamisola, R.; Ang, M.; Oetomo, D.; Khatib, O.; Lim, T. & Lim, S. (2002). The operational
space formulation implementation to aircraft canopy polishing using a mobile
manipulator. Proceedings of IEEE International Conference on Robotics and Automation,
pp. 400-405, Washington, DC, USA.
Li, Y.; Liu, Y.; Liu, X. & Peng, Z. (2004a). Parameter identification and vibration control in
modular manipulators. IEEE/ASME Trans. on Mechatronics, Vol. 9, No. 4, pp. 700705.
Li, Y. & Liu, Y. (2004b). Control of a mobile modular manipulator moving on a slope.
Proceedings of IEEE International Conference on Mechatronics, pp. 135-140, Istanbul,
Turkey.
Liu, Y. & Li, Y. (2005a). Dynamics and model-based control for mobile modular
manipulators. Robotica, Vol. 23, No. 6, pp. 795-797.
Li, Y. & Liu, Y. (2005b). Kinematics and tip-over stability analysis for the mobile modular
manipulator. Proceedings of the Institution of Mechanical Engineers Part C: Journal of
Mechanical Engineering Science, Vol. 219, No. 3, pp. 331-343.
Li, Y.; Liu, Y. & Yan, S. (2005c). Adaptive Neural-Network Control for Redundant
Nonholonomic Mobile Modular Manipulators, Advances in Neural Networks, Eds. by
J. Wang, X. Liao and Z. Yi, Springer, LNCS 3498, Part III, pp.271-276.
Li, Y. & Liu, Y. (2005d). Obstacle Avoidance for Redundant Nonholonomic Mobile Modular
Manipulators via Neural Fuzzy Approaches. Advances in Natural Computation, Eds
by L. Wang, K. Chen and Y.S. Ong, Springer, LNCS 3612, pp.1109-1118.
Li, Y. & Liu, Y. (2006a). Online fuzzy logic control for tip-over avoidance of autonomous
redundant mobile manipulators. International Journal of Vehicle Autonomous Systems,
Vol.4, No.1, pp.24-43, 2006.
Li, Y. & Liu, Y. (2006b). Real-time tip-over prevention and path following control for
redundant nonholonomic mobile modular manipulators via fuzzy and neuralfuzzy approaches. ASME Trans. of Dynamic Systems, Measurement, and Control, (to
appear).
Maciejewski, A. A. & Klein, C. (1989). The Singular Value Decomposition: Computation and
Application to Robotics. International Journal of Robotics Research, Vol. 8, No. 6, pp.
63-79.
Melek, W. & Goldenberg, A. (2003). Neurofuzzy control of modular and reconfigurable
robots. IEEE/ASME Trans. on Mechatronics, Vol. 8, No. 3, pp. 381-389.
Nagatani, K.; Hirayama, T.; Gofuku, A. & Tanaka, Y. (2002). Motion planning for mobile
manipulator with keeping manipulability. Proceedings of IEEE/RSJ International
Conference on Intelligent Robotics and Systems, pp. 1663-1668, EPFL, Swizerland.
Papadopoulos, E. & Rey, D. (1996). A new measure of tipover stability margin for mobile
manipulators. Proceedings of IEEE International Conference on Robotics and
Automation, pp. 3111-3116, Minneapolis, MN, USA.
Rus, D.; Donald, B. & Jennings, J. (1995). Moving furniture with teams of autonomous
robots. Proceedings of IEEE/RSJ International Conference on Intelligent Robotics and
Systems, pp. 235-242, Pittsburgh, USA.

www.intechopen.com

52

Mobile Robots, moving intelligence

Shen, W.; Salemi, B. & Will, P. (2002). Hormone-inspired adaptive communication and
distributed control for conro self-reconfigurable robots. IEEE Trans. on Robotics and
Automation, Vol. 18, No. 5, pp. 700-712.
Stilwell, D. & Bay, J. (1993). Toward the development of a material transport system using
swarms of ant-like robots. Proceedings of IEEE International Conference on Robotics and
Automation, pp. 766-771, Atlanta, Georgia, USA.
Stoy, K.; Shen, W. & Will, P. (2002). Using role-based control to produce locomotion in
chain-type self-reconfigurable robots. IEEE/ASME Trans. on Mechatronics, Vol. 7,
No. 4, pp. 410-417.
Tanner, H. & Kyriakopoulos, K. (2001). Mobile manipulator modelling with Kanes
approach. Robotica, Vol. 19, pp. 675-690.
Tomita, T.; Murata, S.; Kurokawa, H.; Yoshida, E. & Kokaji, S. (1999). Self-assembly selfrepair method for a distributed mechanical system. IEEE Trans. on Robotics and
Automation, Vol. 15, No. 6, pp. 1035-1045.
Will, P. & Grossman, D. (1975). An experimental system for computer controlled mechanical
assembly. IEEE Trans. on Computer, Vol. 24, pp. 879-888.
Yamamoto, Y. & Yun, X. (1996). Effect of the dynamic interaction on coordinated control of mobile
manipulators. IEEE Trans. on Robotics and Automation, Vol. 12, No. 5, pp. 816-824.
Yu, Q. & Chen, I. (2002). A general approach to the dynamics of nonholonomic mobile
manipulator systems. ASME Trans. of Dynamic Systems, Measurement, and Control,
Vol. 124, No. 4, pp. 512-521.

www.intechopen.com

Mobile Robotics, Moving Intelligence

Edited by Jonas Buchli

ISBN 3-86611-284-X

Hard cover, 586 pages

Publisher Pro Literatur Verlag, Germany / ARS, Austria

Published online 01, December, 2006

Published in print edition December, 2006

This book covers many aspects of the exciting research in mobile robotics. It deals with different aspects of the
control problem, especially also under uncertainty and faults. Mechanical design issues are discussed along
with new sensor and actuator concepts. Games like soccer are a good example which comprise many of the
aforementioned challenges in a single comprehensive and in the same time entertaining framework. Thus, the
book comprises contributions dealing with aspects of the Robotcup competition. The reader will get a feel how
the problems cover virtually all engineering disciplines ranging from theoretical research to very application
specific work. In addition interesting problems for physics and mathematics arises out of such research. We
hope this book will be an inspiring source of knowledge and ideas, stimulating further research in this exciting
field. The promises and possible benefits of such efforts are manifold, they range from new transportation
systems, intelligent cars to flexible assistants in factories and construction sites, over service robot which assist
and support us in daily live, all the way to the possibility for efficient help for impaired and advances in
prosthetics.

How to reference

In order to correctly reference this scholarly work, feel free to copy and paste the following:
Yangmin Li and Yugang Liu (2006). Dynamics and Control for Nonholonomic Mobile Modular Manipulators,
Mobile Robotics, Moving Intelligence, Jonas Buchli (Ed.), ISBN: 3-86611-284-X, InTech, Available from:
http://www.intechopen.com/books/mobile_robotics_moving_intelligence/dynamics_and_control_for_nonholono
mic_mobile_modular_manipulators

InTech Europe

University Campus STeP Ri

Slavka Krautzeka 83/A
51000 Rijeka, Croatia
Phone: +385 (51) 770 447
Fax: +385 (51) 686 166
www.intechopen.com

InTech China

Unit 405, Office Block, Hotel Equatorial Shanghai

No.65, Yan An Road (West), Shanghai, 200040, China
Phone: +86-21-62489820
Fax: +86-21-62489821