Development of an Adaptive Algorithm for Solving the Inverse Kinematics Problem for Serial Robot Manipulators
In order to overcome the drawbacks of some control schemes, which depends on modeling the system being controlled, and to overcome the problem of inverse kinematics which are mainly singularities and uncertainties in arm configuration. Artificial Neural Networks (ANN) technique has been utilized...
Saved in:
Main Author: | |
---|---|
Format: | Thesis |
Language: | English |
Published: |
2005
|
Subjects: | |
Online Access: | http://psasir.upm.edu.my/id/eprint/6078/1/FK_2005_58.pdf |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Summary: | In order to overcome the drawbacks of some control schemes, which depends
on modeling the system being controlled, and to overcome the problem of
inverse kinematics which are mainly singularities and uncertainties in arm
configuration.
Artificial Neural Networks (ANN) technique has been utilized where learning is
done iteratively based only on observation of input-output relationship.
The proposed technique does not require any prior knowledge of the kinematics
model of the system being controlled; the main idea of this approach is the use
of an Artificial Neural Network to learn the robot system characteristics rather
than having to specify an explicit robot system model.Since one of the most important problems in using Artificial Neural Networks, is
the choice of the appropriate networks' configuration, two different networks'
configurations were designed and tested, they were trained to learn desired set
of joint angles positions from a given set of end effector positions.
Experimental results have shown better response for the first configuration
network in terms of precision and iteration.
The developed approach possesses several distinct advantages; these
advantages can be listed as follows :(First) system model does not have to be
known at the time of the controller design, (Second) any change in the physical
setup of the system such as the addition of a new tool would only involve
training and will not require any major system software modifications, and
(Third) this scheme would work well in a typical industrial set-up where the
controller of a robot could be taught the handful of paths depending on the task
assigned to that robot.
The efficiency of the proposed algorithm is demonstrated through simulations of
a general 6 D.O.F. serial robot manipulator |
---|