Applying hierarchical and adaptive control to coordinating simple robots Public Deposited

http://ir.library.oregonstate.edu/concern/graduate_thesis_or_dissertations/rx913s06m

Descriptions

Attribute NameValues
Creator
Abstract or Summary
  • Coordinating multiple robots to achieve a complex task requires solving two distinct control problems: the high-level control problem of ensuring that each robot aims to perform a useful task (e.g., coordination) and the low-level control problem of ensuring that each robot actually performs the correct actions to achieve its task (e.g., navigation and locomotion). Though addressing both problems simultaneously with one algorithm is appealing, this is often difficult to impossible in domains requiring a combination of complex actions (goal selection, navigation, obstacle avoidance). This thesis establishes a hierarchical control structure, presents an adaptive navigation method, compares it to reactive navigation, and applies established adaptive coordination techniques under severe restrictions. The development and experimentation process produced results showing the following: 1) Hierarchical control structure proves effective and useful for use on resource limited robotic platforms allowing the subsequent navigation and coordination analyses to be addressed individually. 2) Adaptive navigation is an effective approach for dense environments with limited and noisy sensing, providing improvement over reactive navigation by up to 75%. 3) The application of an abstract difference objective function to training for coordination remains effective under limited information and physical robot motion restrictions, outperforming traditional system or local objectives by up to 50%. Specifically, this work establishes that neuro-evolutionary methods are applicable and beneficial both for the discovery of successful navigation techniques, as well as for the generation of coordination behavior in realistic multi-robot teams where individuals are strongly limited in sensing, communication, and computational ability. Possible extensions include increased levels of communication among individuals as well as configuring individual sensing abilities for heterogeneous teams.
Resource Type
Date Available
Date Copyright
Date Issued
Degree Level
Degree Name
Degree Field
Degree Grantor
Commencement Year
Advisor
Committee Member
Academic Affiliation
Non-Academic Affiliation
Keyword
Subject
Rights Statement
Language
File Format
File Extent
  • 734340 bytes
Replaces
Additional Information
  • description.provenance : Made available in DSpace on 2008-04-07T21:54:37Z (GMT). No. of bitstreams: 1 knu_thesis.pdf: 734340 bytes, checksum: f774e4f446b3597684590f384913e42f (MD5)
  • description.provenance : Approved for entry into archive by Linda Kathman(linda.kathman@oregonstate.edu) on 2008-04-07T21:54:37Z (GMT) No. of bitstreams: 1 knu_thesis.pdf: 734340 bytes, checksum: f774e4f446b3597684590f384913e42f (MD5)
  • description.provenance : Approved for entry into archive by Julie Kurtz(julie.kurtz@oregonstate.edu) on 2008-04-01T17:50:05Z (GMT) No. of bitstreams: 1 knu_thesis.pdf: 734340 bytes, checksum: f774e4f446b3597684590f384913e42f (MD5)
  • description.provenance : Submitted by Matthew Knudson (knudsonm@onid.orst.edu) on 2008-02-11T16:25:21Z No. of bitstreams: 1 knu_thesis.pdf: 734340 bytes, checksum: f774e4f446b3597684590f384913e42f (MD5)

Relationships

Parents:

This work has no parents.

Last modified

Downloadable Content

Download PDF

Items