Robot Dynamics: Q G Q Q Q C Q Q M

1.
Robot Dynamics
Robot arm dynamics deals with the mathematical formulations of the

equations of robot arm motion. The dynamic equations of manipulator motion are a
set of equations describing the dynamic behavior of the manipulator. Such equations
of motion are useful for computer simulation of robot arm motion, the design of
suitable control equations for a robot arm, and the evaluation of the kinematic design
and structure of a robot arm.
Various approaches are available to formulate robot arm dynamics, such as the
Lagrange-Euler, the Newton-Euler, the recursive Lagrange-Euler, and the generalized
d'Alembert principle formulations [Spong and Vidyasagar 2001]. Deriving the
dynamic model of a manipulator using the L-E method is simple and systematic. The
resultant equations of motion, excluding the dynamics of the electronic control device
and the gear friction, are a set of second order, coupled nonlinear differential
equations.
One approach that has the advantage of both speed and accuracy is based on
the N-E vector formulation[Peter I. Corke 1996] was used in this work. The
derivation is simple, although messy, and involves vector cross-product terms. The
resultant dynamic equations, excluding the dynamics of the control device and the
gear friction, are a set of forward and backward recursive equations. These equations
can be applied to the robot links sequentially. There are two problems related to
manipulator dynamics that are important to solve:
inverse dynamics in which the manipulators equations of motion are solved
for given motion to determine the generalized forces. and
direct dynamics in which the equations of motion are integrated to determine
the generalized coordinate response to applied generalized forces.
The equations of motion for an n-axis manipulator are given by:
C q, q q G q
M q q
(1)
2
dq
d q
T
, q 2
where: if we have 4DOF q 1 2 3 4 , q
dt
dt
Obtaining the dynamic equations of motion using the MatLab
program(robotics toolbox) was very powerful process; the toolbox use N-E approach
to compute the equation of motion by feeding the program by necessary data about
the robot system by using the functions(dyn.m, robot.m) to introduce the robot object
in the MatLab program, the robot object then can be used in the MatLab program to
define Simulink blocks. Because of the nature of the formulation and the method of
systematically computing the torques, computations are much simpler, allowing a
short computing time. With this algorithm, about three milliseconds are needed to
compute the feedback joint torques per trajectory set point.
2.
Simulink Model
The robot model studied in this work is the IRIS robot manipulator where all
kinematic and dynamic parameters are illustrated in the appendix A, the dynamic
model for this robot system is a highly nonlinear differential equation due to the
coupled between different degrees of freedom(4DOF). The modeling of the robot
kp
system was done using Robotics Toolbox in the MatLab program, using the
functions(dyn.m, robot.m) to introduce the robot object in the program and then the
robot object can be used with two way dynamics; The forward recursion propagates
kinematics information-such as angular velocities, angular accelerations, linear
accelerations, total forces and moments exerted at the center of mass of each linkfrom the base reference frame (inertial frame) to the end-effector. The backward
recursion propagates the forces and moments exerted on each link from the end
effector of the manipulator to the base reference frame. using the inverse and the
direct dynamic functions which are very important in the usage of dynamic block of
Simulink in the robot control simulation.
The IRIS system was simulated using Simulink program under the robot
control techniques which are stated in the previous section. Computed torque control
approach Simulink model shown in figure 4, feedforward control approach Simulink
model shown in figure 5.
IRIS
kp
K*u
IRIS
kp
noname
IRIS
tau
kv
manipulator.
qdd
qd
Joint position
K*u
Direct Dynamic
Desired Joint Trajectory

noname
IRIS
q
qd
+ torque control Simulink

Figure 4. Computed
qdd
k + + model for IRIS robot
qd
tau
qdd
Inverse Dynamic
Figure 5. Feedforward control Simulink model for IRIS robot.
Joint velocity
Simulation Results
3.
In this section, performances of different controllers for a full motion of the

IRIS robot is evaluated. The computation of the dynamics for the full robot model is
much more complicated than the single joint case due to gravity and coupling effects
between joints. Some robot controllers, therefore, compensate for only a part of these
effects. In this article the integral feedback term has been not included; since the
interesting in the dynamic trajectory errors rather than in the steady state errors, which
can be reduced by the integral term. The gains k p and k v adjusted to achieve
acceptable performance. The reference trajectory of the movement studied is
generated by a fifth order polynomial, which gives smooth velocities and
T
accelerations. For this smooth trajectory, the joint moves from qi 0 0 0 0 to
q f 2 * / 3 pi / 3 pi / 2 pi / 4 in 6s. The evaluations of the following two
control methods mentioned before used for high speed movements of all four joints of
the manipulator include the following:
1. Computed torque control.
2. Feedforward control.
The values of the gains k p and k v were adjusted to reduce the trajectory
and velocity errors and the values chosen as k p 50 , and k v =10 for all joints. The
trajectory errors for the first of the above 2 controllers are shown on Figure 6., and the
velocity errors are shown in Fig. 7, the trajectory errors and the velocity errors for the
first controller can be considered a large; With this controller,
1st the
jointpeak trajectory
errors are 0.401 , 0.172 , 0.258 and 0.688 for joints2nd
1, joint
2, 3, and 4
3rd
joint
respectively. The trajectory and velocity errors for the feedforward controller are
4th joint
shown on Figures 8 and 9., and these errors can be considered more acceptable from
the first controller; With this controller, the peak trajectory errors are 0.005 ,
0.0028 , 0.0046 and 0.002 for joints 1, 2, 3, and 4 respectively.
Figures 10 and 11 shows the input torques for the joint moves from qi to q f
in 6 sec of two types of controller, the difference between values of input torques for
the two types was very small and can be considered unperceptible.
T
0.01
joints error (rad)
0.005
-0.005
-0.01
3
time(sec)
Figure 6. Trajectory errors of computed torque control.
0.015
1st J. error (rad)
0.005
-0.005
2nd J. error (rad)
3rd J. error (rad)
-0.01
4th J. error (rad)
velocity error (rad/sec)
0.01
x 10
1st joint
2nd joint
3rd joint
4th joint
-4
0
-2
0
5
x 10
-5
0
-5
0
2
x 10
-4
3
time (sec)
Figure 7. Velocity errors of computed torque control.

-2
0
5
x 10
3
time(sec)
-5
0
-5
0
Figure 8. Trajectory errors of feedforward control.
1st J. velocity error

2nd J. velocity error
3rd J. velocity error
4th J. velocity error
x 10
-4
0
-5
0
2
x 10
3
time(sec)
-4
0
-2
0
2
x 10
-4
0
-2
0
5
x 10
-4
0
-5
0
Figure 9. Velocity errors of feedforward control.
25
1st joint
2nd joint
3rd joint
4th joint
20
10
-5
0
time(sec)
Figure 10. simulated Input torque of the computed torque controller.
25
1st joint
2nd joint
3rd joint
4th joint
20
15
Input torque(N/m)
Input torque(N/m)
15
10
-5
0
time(sec)
Figure 11. simulated Input torque of the feedforward controller.
4.
Conclusions
In this paper the simulation of the dynamic system and two types of control for
robot manipulator was done by using a Simulink based on robotics toolbox. the
models derived in this paper was very effected and can easily be used the controller
and also can be easily regulate the gains of controllers presented some simulation
results of using derived dynamic model of the manipulator for dynamic CTC and
feedforward control. The results indicate that feedforward control can improve the
trajectory following accuracy significantly. Such a performance using feedforward
compensation also indicates that the estimated rigid body model of the manipulator is
quite accurate and adequate for control purposes for all joints.
5.
References
Jorge Angeles, 2003 "Fundamentals of Robotic Mechanical Systems: Theory,
Methods, and Algorithms", Second Edition, Springer-Verlag, New York.
Kelly R., Salgado R., 1994 "PD control with computed feedforward of robot
manipulators: A design procedure", IEEE Transactions on Robotics and
Automation, Vol. 10, No. 4, August, pp. 566571.
Kelly, R., Santibanez V. and Loria, A. 2005 "CONTROL of ROBOT
MANIPULATOR in JOINT SPACE", Advanced Textbooks in Control and Signal
Processing, Springer-Verlag London.
M. Spong and M. Vidyasagar, 2001 "ROBOT MODELING and CONTROL", John
Wiley and Sons INC., London.
Mohammad Reza Emami, 1997 "Systematic Methodology of Fuzzy-Logic Modeling
and Control and Application to Robotics", Master thesis, Dept. of Mechanical
and Industrial Engineering, University of Toronto.
Peter I. Corke, 1996 "Robotics TOOLBOX for MATLAB (Release 7.1)", IEEE
Robotics and Automation Magazine, Vol.3, No. 1, pp. 24-32.
Rafeal M. Inigo, and James S. Morton, 1991 "Simulation of the Dynamics of an
Industrial Robot", IEEE transaction on Education, Vol. 34, No. 1, pp.89-99.
Steven T. Karris, 2006 "Introduction To Simulink With Engineering Applications",
Orchard Publications, United States of America.
Teng, F. C. 2000 "Real Time Control using MatLab and Simulink", IEEE press.
Nomenclature:
q : vector of generalized joint coordinates describing the pose of the manipulator.

: vector of joint velocities.
q
: vector of joint accelerations.
q
q d : vector of desired generalized joint coordinates describing the pose of the

manipulator.
q d : vector of desired joint velocities.
d : vector of desired joint accelerations.
q
M : symmetric joint-space inertia matrix, or manipulator inertia tensor.
C : Coriolis and centripetal effects; Centripetal torques are proportional to q i2 ;
while the Coriolis torques are proportional to q i , q j .

G :gravity loading.
: vector of generalized torques or forces associated with the generalized

coordinates q .
k p , k v : are symmetric positive definite matrices, called gains of position and velocity
respectively.
~
q
vector of joint errors.
~
vector of joint velocity errors.
q
IRIS: Intelligent Robot Industrial System.
Appendix A: Robot Model Studied

Figure 12 illustrates the geometric form of the robot manipulator studied in this paper,
and the kinematics and dynamic parameters are listed in the tables 1[Mohammad Reza 1997],
below.
Table 1. kinematics parameters of IRIS robot
Link i
Center of
Inertia matrix (kg.m2)
i
ai
d i Mass
i
mass (m)
(deg) (m) (m) (kg)
1
1 90 0.28
0.0420
0
0.0075
0.0035
0
0.0035
0.0400
0.063
0.026
0.0208
0
0.0070
.0035
.0035
0.0160
0.0076
0.0059
0
0.0388
0
0.0059
0
0.0340
0.0014
0
0.0063
0
0
0
0.0064
2 90
0.32
2.191
3 90
0.25
2.81
0.19 1.385
0
0.13
0.009
4.023
0.108
0.020
0.146
Figure 12. The desired configuration of the IRIS robot with the link coordinate frames.

Robot Dynamics: Q G Q Q Q C Q Q M

Uploaded by

Copyright:

Available Formats

Robot Dynamics: Q G Q Q Q C Q Q M

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Robot Dynamics: Q G Q Q Q C Q Q M

Uploaded by

Copyright:

Available Formats

1.

Robot arm dynamics deals with the mathematical formulations of the

Desired Joint Trajectory

+ torque control Simulink

Figure 5. Feedforward control Simulink model for IRIS robot.

In this section, performances of different controllers for a full motion of the

joints error (rad)

Figure 6. Trajectory errors of computed torque control.

1st J. error (rad)

2nd J. error (rad)

3rd J. error (rad)

4th J. error (rad)

velocity error (rad/sec)

Figure 7. Velocity errors of computed torque control.

Figure 8. Trajectory errors of feedforward control.

1st J. velocity error

Figure 9. Velocity errors of feedforward control.

Figure 10. simulated Input torque of the computed torque controller.

Figure 11. simulated Input torque of the feedforward controller.

q : vector of generalized joint coordinates describing the pose of the manipulator.

q d : vector of desired generalized joint coordinates describing the pose of the

while the Coriolis torques are proportional to q i , q j .

: vector of generalized torques or forces associated with the generalized

Appendix A: Robot Model Studied

You might also like

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.