Você está na página 1de 41

Universidade de Aveiro Departamento de Engenharia Mecnica

2014

Tiago Gomes Moura

TTULO DA TESE (MXIMO 70 CARACTERES)


THESIS TITLE (MAX 70 CHARACTERS)

DOCUMENTO
PROVISRIO

Universidade de Aveiro Departamento de Engenharia Mecnica


2014

Tiago Gomes Moura

TTULO DA TESE (MXIMO 70 CARACTERES)


THESIS TITLE (MAX 70 CHARACTERS)

Dissertao apresentada Universidade de Aveiro para cumprimento dos requisitos necessrios obteno do grau de Mestre em (designao do mestrado), realizada sob a orientao cientfica do Doutor (nome do orientador),
Professor (categoria do orientador) do Departamento de Engenharia Mecnica da Universidade de Aveiro, e do Doutor co-orientador, Professor auxiliar
convidado do Departamento de Matemtica da Universidade de Aveiro.

Dedication...

o jri / the jury


presidente / president

Prof. Doutor ...


professor associado da Faculdade de ....

agradecimentos /
acknowledgements

texto Agradecimentos
(opcional)

Contedo
Contedo

Lista de Figuras

iii

Lista de Tabelas

Structure for support of the arms


1.1 Direct kinematics of the arm with 6 DOF . . . . . . . . . . . . . . . . . .
1.2 Workspace . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.3 Structural design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

Kinematics of Cyton Gamma arm


2.1 Direct Kinematic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.2 Inverse Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
2.3 Differential Kinematics . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

11

Movement algorithms and control


3.1 Trajectory Planning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.2 Inverse Kinematics algorithms . . . . . . . . . . . . . . . . . . . . . . . . .
Jacobian Inverse . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

17

A Transformations

2
4
7

11
13
14

17
19
19
23

Lista de Figuras
1.1

Length of body parts as function of body height [?].

. . . . . . . . . . . . . . .

1.2

Cyton Gamma 1500 dimensions. . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.3

Schematic of coordinate system of the Cyton arm with 6 DOF. . . . . . . . . . .

1.4

Workspace representation of right arm (blue) and left arm (red) assembled to
torso (green) in a vertical configuration. . . . . . . . . . . . . . . . . . . . . . . .

1.5

Workspace representation of right arm (blue) and left arm (red) assembled to
torso (green) in a lateral configuration. . . . . . . . . . . . . . . . . . . . . . . .

1.6

3D model of a solution to the structure that allows to adjust the angle and the
distance between the two arms assembled in sideways of the torso. . . . . . . . .

1.8

Workspace representation of right arm (blue) and left arm (red) assembled to
torso (green) in a side mounted configuration with 45. . . . . . . . . . . . . . .

1.7

3D model of a solution to the structure that allows to adjust the distance between
the two arms assembled vertically to the torso. . . . . . . . . . . . . . . . . . . .

2.1

Coordinate frames of 3 DOF Cyton arm. . . . . . . . . . . . . . . . . . . . . . .

12

3.1

Curves of position, velocity and accelaration. . . . . . . . . . . . . . . . . . . . .

18

3.2

Inverse kinematics algorithm using Jacobian inverse with estimated joint positions. 20

3.3

Block diagram of the inverse kinematics algorithm with Jacobian inverse [?]. . .

3.4

Inverse kinematics algorithm using Jacobian inverse with feedback of joint positions. 22

21

A.1 Schematics of 2 arms coordinate system with transformation of vertical configuration. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

23

A.2 Schematics of 2 arms coordinate system with transformation of lateral configuration. 24

iii

Lista de Tabelas
1.1

Denavit-Hartenberg kinematic parameters of the Cyton arm with 6 DOF. . . . .

1.2

Workspace specifications of vertical configuration in mm. . . . . . . . . . . . . .

1.3

Workspace specifications of lateral configuration in mm. . . . . . . . . . . . . . .

1.4

Workspace specifications of lateral configuration with 45 in mm. . . . . . . . . .

2.1

Denavit parameters of 3 DOF Cyton arm.

. . . . . . . . . . . . . . . . . . . . .

12

A.1 Homogenous transformation matrices of vertical configuration. . . . . . . . . . .

23

A.2 Homogenous transformation matrices of lateral configuration. . . . . . . . . . . .

24

A.3 Homogenous trasformation matrix to do a rotation along y0 axis.

25

. . . . . . . .

Chapter 1
Structure for support of the arms

According to the outlined goals, it was necessary to design a structure which allows to connect
two arms in a similar way to the structure of human body torso. In accordance with the
biometrical relations, studied in anthropometry, there is a length estimation of each part of
the body as function of the individual height. In figure 1.1 it is represented a schematic of
these estimation.

Figure 1.1: Length of body parts as function of body height [?].

The distance between the two Cyton arms, as well as the torso height, were based in these
estimation, although, the distance between the arms was designed to be adjustable.

Figure 1.2: Cyton Gamma 1500 dimensions.

Based on specifications of the Cyton Gamma 1500 arm, figure 1.2, its length is 688mm.
Thus, the estimation of the distance between the two arms is about 404mm and the torso
height is about 450mm.

1.1

Direct kinematics of the arm with 6 DOF

In order to define the direct kinematics of the Cyton Gamma 1500, that has 6 DOF, was used
the Denavit-Hartenberg convention. This method defines some rules for the definition of each
link frame. To simplify the analysis it was necessary to ignore 1 DOF, which corresponds
to the joint 7, that only has influence in the orientation of the end-effector. The figure1.3
represents the coordinate frames of each link of the arm according to the coordinate system
defined by the arm manufacture. The coordinate system starts with the reference frame,
in the base of the arm, and ends with the end-effector frame. The Denavit parameters are
shown in the table 1.1.
1

Degrees Of Freedom.

Figure 1.3: Schematic of coordinate system of the Cyton arm with 6 DOF.

Table 1.1: Denavit-Hartenberg kinematic parameters of the Cyton arm with 6 DOF.
i

ji

Li (mm)

di (mm)

ai

p/2+j1

L1

p/2

p/2+j2

L2

-p/2

j3

L3

p/2

j4

L4

-p/2

j5

L5

p/2

j6

L6

With these parameters its possible to determine the arm workspace in order to give an
idea of the volume that the arm can reach. To simplify the kinematics of the arm was reduced

j j

the DOF of the arm form 6 to 3 where the joints 3, 5 and 6 were fixed.

1.2

Workspace

It was developed a simulation in Matlab in order to set the orientation of the arm in the
structure, that give the workspace that best fits to the project situation. In it are shown 2
arms mounted in some different configurations which approximates to the configuration of a
human torso. At this point there were defined 3 kinds of configurations:
Vertically mounted arms;
Sideways mounted arms;
Sideways mounted arms, with angle between the base and the structure;
To simulate these configurations, some transformations were made from global to local frames
of each arm base, that are described in appendix A . With these transformations were made
3 simulations with the different configurations that were previously described. These simulations were based in iterative cycles that run through joint angles form lower to higher limits.
The point cloud of every end-effector position was stored and plotted, representing the arms
workspace in the different configurations.
In the figure 1.5 it is displayed the workspace of the two arms assembled in a vertical
configuration.

Figure 1.4: Workspace representation of right arm (blue) and left arm (red) assembled
to torso (green) in a vertical configuration.

In the table 1.3 are outlined the workspace dimensions of the two arms assembled in a
vertical configuration.

Table 1.2: Workspace specifications of vertical configuration in mm.


X min

-51.1858

X mx

51.1858

Y mn

-71.6950

Y mx

71.6950

Z mn

-19.0411

Z mx

65.6570

In order to approach the robotic arms assembly to the human arms configuration, it was
made a simulation with both arms assembled in sideways of the torso. The point cloud of
the workspace of this configuration is displayed in the figure 1.5.

Figure 1.5: Workspace representation of right arm (blue) and left arm (red) assembled
to torso (green) in a lateral configuration.

The workspace dimensions of the two arms assembled in a lateral configuration are described in the table 1.3.

Table 1.3: Workspace specifications of lateral configuration in mm.


X mn

-51.1858

X mx

51.1858

Y mn

-88.0911

Y mx

88.0911

Z mn

-9.4950

Z mx

93.4950

It was also made a simulation with an intermediate angle, 45, and the point cloud that
represents its workspace is shown in figure 1.6.

Figure 1.6: Workspace representation of right arm (blue) and left arm (red) assembled
to torso (green) in a side mounted configuration with 45.

The respective dimensions of the two arms workspace, assembled in lateral configuration
with 45, are described in the table 1.4.

Table 1.4: Workspace specifications of lateral configuration with 45 in mm.


X mn

-41.3066

X mx

60.9748

Y mn

-90.0248

Y mx

90.0248

Z mn

-9.1858

Z mx

93.1858

The workspace of the robot should be, preferentially, in front of robot torso and as far
forward as possible. By the graphical comparison, that gives a visual idea of the workspace,
and by the workspace dimensions analysis of each configuration, it can be concluded that
the vertical configuration is the one that best fits the workspace specifications. Furthermore,
that configuration have a large volume of workspace that both arms can reach simultaneously.
That allows the robot to perform grasp movements with both end-effectors at the same time.
Given this features, the adopted configuration was the one with the two arms assembled
in a vertical configuration. Hereupon it was made a design of a structure that supports both
arms with the dimensions previously mentioned and with variable distance between the arms.

1.3

Structural design

In this part it was necessary to meet the physical aspects of the arms as well as the functional
aspects desired to the structure. Thus the structure must be light, flexible, strong and
easily portable. In order to design a structure with these features there were mainly used
aluminum components, going of against the material used in the humanoid robots available
in the market. Another specification was that the distance between both arms should be
adjustable, thus, there were designed guidance systems.
At this stage were proposed and designed some structures that met the different configurations initially desired. In the figure 1.7 is shown a solution that allows to fix the arms in
sideways of the torso with possibility to adjust the angle between the base of the arms and
the torso, as well as the distance between both arms.

Figure 1.7: 3D model of a solution to the structure that allows to adjust the angle and
the distance between the two arms assembled in sideways of the torso.

In this solution, the change of the distance between the two arms is done through the
guidance system of the horizontal aluminum profile and the setting of the angle between the
arms base and torso is done by the half moon shape profile. This was not the adopted
solution because it was concluded, in previous section, that the best arms configuration is
the vertical assembly. Furthermore, this solution presents fewer robustness due to the the
hinges of the angle adjustment and it is a heavier solution.
For the vertical configuration, the proposed solution is presented in figure 1.8.

Figure 1.8: 3D model of a solution to the structure that allows to adjust the distance
between the two arms assembled vertically to the torso.

This solution is lighter than the solution of figure 1.7 and the arms are assembled with
its bases fixed to the smaller section aluminum profiles, that allows to adjust the distance
between the arms. This a lighter solution because it doesnt need an intermediate plate
between the arms base and the torso.

Chapter 2
Kinematics of Cyton Gamma arm

In order to do the imitation movements with both arms, first it was necessary to make a study
of the Cyton Gamma 1500 arms in terms of its kinematics. The kinematics of a manipulator
is a relation between position, velocity and acceleration of its links.
In this chapter is discussed the kinematics of Cyton Gamma 1500, the algorithm used to
do a linear movement and also 2 algorithms that apply control with feedback.

2.1

Direct Kinematic

The direct kinematic of a robotic arm is the determination of end-effector position and orientation as a function of the joint angles. To the Cyton Gamma arms were constrained some
joint variables, and considered only 3 joints, to simplify the kinematic of the arm.
In order to set the direct kinematics of the arm was adopted the Denavit-Hartenberg
convention.
In the figure 2.1 it is represented the coordinate systems of each joint to the 3 DOF
simplification of Cyton arm, and in the table 2.1 are the respective Denavit-Hartenberg
parameters.

11

Figure 2.1: Coordinate frames of 3 DOF Cyton arm.

Table 2.1: Denavit parameters of 3 DOF Cyton arm.


i

ji

Li

di

ai

90+j1

L1

90

90+j2

L2

j3

L3

From the Denavit parameters its possible to determinate the position of end-effector (x,y

j j

and z coordinates) in terms of joint angles ( 1, 2 and 3).


This can be done by the successive multiplication of the transformation matrices, each
one corresponding to the respective link. The transformation matrix applied to each link is
given by the following expression2.1.

12

cos(i ) sin(i )cos(i )

sin( )
i

Ai =

cos(i )cos(i )

sin(i )sin(i )

Li cos(i )

cos(i )sin(i ) Li sin(i )

sin(i )

cos(i )

di

(2.1)

By the multiplication of all the transformation matrices, result the expressions for Cartesian coordinates of the end-effector, that are shown in the equations 2.2, 2.3 and 2.4.

2.2

X = sin(1 ) (L2 sin(2 ) + L3 sin(2 + 3 ))

(2.2)

Y = cos(1 ) (L2 sin(2 ) + L3 sin(2 + 3 ))

(2.3)

Z = L1 + L2 cos(2 ) + L3 cos(2 + 3 )

(2.4)

Inverse Kinematics

The Inverse Kinematics (IK) consists of the determination of the joint variables corresponding
to a given end-effector position and orientation. The IK may have multiple solution or it can
even have no solution, depending on the complexity and number of DOF of the arm. That
is a reason to the 3 DOF simplification that was made.

Usually, the IK is determined by manipulating the DK expression. The expression of 1,


presented in the equation 2.5, can be obtained by dividing X by Y as follows:
X
Y

sin(1 )(L2 sin(2 )+L3 sin(2 +3 ))


cos(1 )(L2 sin(2 )+L3 sin(2 +3 ))

X
1 = arctan
Y
The 2 expression can be obtained by the manipulation of X and Z expressions.


(2.5)

X
sin(1 )

Z L1 =

L2 sin(2 ) + L3 sin(2 + 3 )
L2 cos(2 ) + L3 cos(2 + 3 )

From the manipulation results:


2L2

Y
sin(1 )

sin(2 ) + 2L2 (Z L1 ) cos(2 ) = L23 + L22 +

X2
sin(2 )2

+ (Z L1 )2

That result is in the form of K1 sin() + K2 cos() = K3 that have a know solution.

But when 1 tends to zero,

x
sin(1 ) tends

to infinite, so its necessary to use the Y expression

instead of the X. Thus, K2 remains the same because it doesnt depends on 1, K1 and K3
take the following expressions:

13

K1 = 2L2

Y
sin(1 )

2L2

Y
cos(1 )

K2 = 2L2 (Z L1 )
K3 = L23 + L22 +

X2
sin(2 )2

+ (Z L1 )2 L23 + L22 +

Y2
cos(1 )2

+ (Z L1 )2

The solution for 2 is given by the equation 2.6.

K1
2 = atan
K2


atan(

K12 + K22 K32


K3

(2.6)

To obtain 3 can be used the same method, but squaring both X and Z expression and
Y and Z expressions. Thus, K1, K2 and K3 are:
K1 = 0
K2 = 2L2 L3
K3 =

X
sin(1 )

2

+ (Z L1 )2 L22 + L23

Y
cos(1 )

2

+ (Z L1 )2 L22 + L23

The expression of 3 is given by the equation 2.7.

3 = arctan

2.3

K22 K32
K3

(2.7)

Differential Kinematics

The differential kinematics aims to describe the arm movement between two configurations.
It gives the relationship between the joint velocities and the corresponding end-effector linear
and angular velocity. This relation is obtained by the arm Jacobian (J). This relation can be
defined by the equation 2.8.

d~r = J d~q

14

(2.8)

where d~r is a vector with Cartesian position of end-effector, d~qis a vector with joint angles
and the Jacobian matrix is given by equation 2.9.

J =

r1
q1
r2
q1

r1
q2
r2
q2

..
.

...
..
.

rm
q1

rm
q2

...

..
.

...

r1
qn
r2
qn

..
.

rm
qn

(2.9)

where m represents the number of Cartesian variables and n the number of joints.
For the 3 DOF simplification there are 3 joints and the Cartesian variables are X, Y and
Z, despising the orientation. With these considerations, the Jacobian of the arm is presented
in equation 2.10.

cos(1 ) (L2 sin(2 ) + L3 sin(2 + 3 ))

sin(1 ) (L2 cos(2 ) + L3 cos(2 + 3 ))

J =
sin(1 ) (L2 sin(2 ) + L3 sin(2 + 3 ))

L3 sin(1 )cos(2 +

cos(1 ) (L2 cos(2 ) + L3 cos(2 + 3 )) L3 cos(1 )cos(2 +


L2 sin(2 ) L3 sin(2 + 3 )

L3 sin(2 + 3 )
(2.10)

The IK can be derived by the inverse matrix of Jacobian and the expression that represents
the joint velocities as function of end-effector Cartesian coordinates is shown in equation 2.11
.

d~q = J 1 d~r

(2.11)

The calculation of an inverse of a matrix can be computationally slow, and that could be
crucial when all parts of the arm control are running at the same time. So this matrix was
calculated analytically and is expressed by the equation 2.12.

J 1 =

cos(1 )
L2 sin(2 )+L3 sin(2 +3 )
sin(1 )sin(2 +3 )
L2 (sin(2 )cos(2 +3 )cos(2 )sin(2 +3 ))
sin(1 )(L2 sin(2 )+L3 sin(2 +3 ))
L2 L3 (sin(2 )cos(2 +3 )cos(2 )sin(2 +3 ))

sin(1 )
L2 sin(2 )+L3 sin(2 +3 )
cos(1 )sin(2 +3 )
L2 (sin(2 )cos(2 +3 )cos(2 )sin(2 +3 ))
cos(1 )(L2 sin(2 )+L3 sin(2 +3 ))
L2 L3 (sin(2 )cos(2 +3 )cos(2 )sin(2 +3 ))

cos(1 )sin(2 +
L2 (sin(2 )cos(2 +3 )cos(

L2 L3 (si

(2.12)

15

Chapter 3
Movement algorithms and control

The main interest of a manipulator or, in this case, two robotic arms, is the capability
to preform a movement from an initial to a final posture. It is necessary to implement
algorithms that takes into account the end-effector position, velocity and/or acceleration.
These algorithms are known by trajectory planning and aim to define the behavior of the
end-effector, not only in terms of the path that the end-effector has to follow but also define
suitably trajectories.

3.1

Trajectory Planning

After the study of the relationship between operational space and joint space either in geometric (direct and inverse kinematics) and kinetic (differential kinematics), its necessary
to proceed to the trajectory planning. This includes the set of studies and methods that
define the velocities behavior of each joint, in order to meet the movement goal. When the
end-effector movement is as simple as moving from one point to another no matter what
path it does, its called point-to-point planning. On the other hand, when its necessary for
the end-effector to execute a well-defined path in the operation space, obeying to precise
temporal criteria, its called path motion planning.
To perform movements that are human imitations it was resorted to the path motion
planning and the most common planning is with linear trajectory. In this trajectory planning,
the end-effector follows a straight line from the initial position to the final one. This planning
defines the evolution of each Cartesian coordinate of the end-effector through time so that
the position, velocity and even acceleration, can be verified. The movement should start
at time instant ti and finish at tf starting from initial position Xi, Yi and Zi to the final
position Xf, Yf and Zf. In the simplest case there is a start and finish position and the start
and finish velocities are null. The simple way to do that is to use a polynomial function of
t. Thus, the velocity expression is a second order polynomial equation because it has two
roots and, consequently, the position expression is a third order equation. Considering p a
.
..
generic Cartesian position, p the velocity of this position and p its acceleration, the general

17

expressions of position, velocity and acceleration are presented in equations 3.1, 3.2 and 3.3.[?]

p(t) = a0 + a1 t + a2 t2 + a3 t3

(3.1)

p(t) = a1 + 2a2 t + 3a3 t2

(3.2)

..

p(t) = 2a2 + 6a3 t

(3.3)

The expressions 3.1, 3.2 and 3.3 can be represented graphically as shown in the figure .

Figure 3.1: Curves of position, velocity and accelaration.

Considering the terms imposed in the expression 3.4, and combining the equations 3.1,3.2
and 3.3, it can be shown that the equations 3.5, 3.6 and 3.7 represent the linear position
movement with null initial and final velocities.

p(ti ) = 0; p(tf ) = pf ; p(0) = pi ; p(tf ) = pf

p(t) =

2
3
(pf pi ) t3 + 2 (pf pi ) t2 + pi
3
tf
tf

(3.5)

6
6
(pf pi ) t2 + 2 (pf pi ) t
3
tf
tf

(3.6)

12
6
(pf pi ) t + 2 (pf pi )
3
tf
tf

(3.7)

p(t) =

..

(3.4)

p(t) =

18

3.2

Inverse Kinematics algorithms

As seen before, the inverse kinematics of a manipulator can be difficult to determine and it
only gives the relation between end-effector position and the joint angles. The arm Jacobian
can be useful, among other things, to determine the inverse kinematics algorithms. In the
following sections are presented two algorithms to implement the IK using the Jacobian
matrix.

Jacobian Inverse
As seen before, in section 2.3, the Jacobian matrix gives a relation between the joint velocity
space and the operational velocity space. The equation 3.8 shows this relation.

d~q = J 1 (q) d~r

(3.8)

The Jacobian matrix depends on the current configuration of the arm, so as the initial
posture q0 is known, joint positions can be estimated resorting to a simple technique based

Dt and the joint


positions and velocities at time t are known, the the joint positions at time t+Dt are given

on the Euler integration method. If the integration interval is given by


by the expression presented in the equation 3.9.

q(t + 4t) = q(t) + q(t) 4t

(3.9)

With this technique, a first approach of IK using differential kinematics can be applied.
The algorithm used to apply the IK is described in figure 3.2.

19

Figure 3.2: Inverse kinematics algorithm using Jacobian inverse with estimated joint
positions.

This is a simple algorithm that can be applied to the robotic arms. Although, as the
joint positions were estimated, is expected some error between the desired positions of the
end-effector and the real positions during the movement. By this, it was necessary to apply
a control algorithm. The first approach, that cannot be considered control because it doesnt
measure the error between the desired and the real positions, was to get the joint angles
feedback from the motors encoders and introduce them in the inverse Jacobian matrix. This
approach only allow to have the real joint positions instead of the estimated ones, witch
improves the followed trajectory, but it doesnt ensures that the end-effector is following the
defined trajectory.

20

To guaranty that the defined trajectory is followed with minimal error it was necessary to
apply a control algorithm, first using the Jacobian inverse. Taking in account the operational
space error e between the desired r~d and the real ~r positions of the end-effector, given by the
expression 3.10 .
e = r~d ~r

(3.10)

According to [?], considering a square and non singular Jacobian matrix, the relation
between the joint velocity space and the operational velocity space can be described as shown
in the equation 3.11, considering the error measure e.
.

q = J 1 (q) (r + K e)

(3.11)

The error converges to zero depending on the K and the sampling time. The block
diagram of the inverse kinematics algorithm is presented in the figure .

Figure 3.3: Block diagram of the inverse kinematics algorithm with Jacobian inverse
[?].

The DirKin refers to the direct kinematics of the arm and is used to calculate the error
between the real and the desired positions. This algorithm enables the end-effector to follow
the desired trajectory with more precision than the previous algorithms and is described in
the figure 3.4.

21

Figure 3.4: Inverse kinematics algorithm using Jacobian inverse with feedback of joint
positions.

This algorithm can be computationally slow because it uses the Jacobian inverse, witch
requires some computational time. This problem was bypassed using the analytical Jacobian
inverse, so is not necessary calculate it in real time.

22

Appendix A
Transformations

The coordinate system of the arms in a vertical configuration is shown in figureA.1 .

Figure A.1: Schematics of 2 arms coordinate system with transformation of vertical


configuration.

The R frame represents the global coordinate system and 0R and 0L frames represents
the base of right and left arms frame, respectively. The transformation matrices of right and
left arm, in vertical configuration are shown in table A.1.

Table A.1: Homogenous transformation matrices of vertical configuration.

RR =

1 0
0
0

0 1 0 202

0 0 1 450

0 0
0
1

RL =

(a) Right arm.

1
0
0
0

(b) Left arm.

23

0 0
0

1 0 202

0 1 450

0 0
1

In figure A.2 is presented the coordinate system for side mounted arms with Z axis of
each arm perpendicular to zR.

Figure A.2: Schematics of 2 arms coordinate system with transformation of lateral


configuration.

The homogeneous transformation matrices for lateral configuration are shown in table
A.2.

Table A.2: Homogenous transformation matrices of lateral configuration.

RR =

1 0
0
0

0 1 0 202

0 0 1 450

0 0
0
1

RL =

(a) Right arm.

1
0
0
0

0 0
0

1 0 202

0 1 450

0 0
1

(b) Left arm.

To define a configuration that the arm has an angle along y0 axis, it was necessary to do
a local rotation defined by the matrix presented in table A.3, where is the pretended angle.

24

Table A.3: Homogenous trasformation matrix to do a rotation along y0 axis.

R =

cos()
0
sin()
0

25

0 sin() 0

1
0
0

0 cos() 0

0
0
1

Você também pode gostar