Back to search

University of Victoria (Canada)

Global resolution of kinematic redundancy using parameterized self-motion optimization

Abstract

dc:description.abstract

The control of manipulators with redundant degrees of freedom has been of research interest for some time. Many algorithms to solve the one to many mapping of the inverse kinematics of such redundant manipulators have been presented. All are intended to be optimal in some sense, either locally, using only instantaneously available information on the manipulator's state, or globally over the trajectory of the manipulator. The two classes of manipulator control each have their own problems. Local optimization methods suffer from instabilities due to their 'shortsightedness', unable to tell that difficulties are arising until it is too late. Global optimization methods proposed to date are far too computationally expensive to be implemented for real-time control of a manipulator. Thus, an algorithm that fits between the two categories is proposed, the Parameterized Self-Motion Optimization algorithm. Using iterative optimization techniques over a limited parameterization of the solution space, the algorithm does not produce true globally optimal trajectories, but instead what is hoped to be good approximations of said trajectories. Precise accuracy is traded for computational efficiency, generating a solution that is a distinct improvement over that provided by local control methods. The algorithm as implemented is functional, and sufficiently flexible to accommodate changes to the optimization criteria with moderate effort. Its results are shown to be improvements of locally optimal results, improved in a global sense. Thus the algorithm presents a capability to significantly improve the quality of control over local methods, even if the results cannot be shown to be globally optimal.

Degree

thesis:*
Level thesis:degree_level
Master of Applied Science M.A.Sc.
Department thesis:degree_department
Department of Electrical and Computer Engineering
Year dc:date.issued
1993

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Fenger, David Kirk

Rights

dc:rights
Statement dc:rights
  • Available to the World Wide Web

Identifiers

dc:identifier.*
Handle dc:identifier.uri
https://hdl.handle.net/1828/17751

Chain of custody

source
Harvested from
University of Victoria (Canada)
Base URL
dspace.library.uvic.ca/server/oai/request
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
citation

Fenger, David Kirk. Global resolution of kinematic redundancy using parameterized self-motion optimization. Master of Applied Science M.A.Sc. thesis, 1993. https://hdl.handle.net/1828/17751