The University of Queensland Homepage
School of ITEE ITEE Main Website

 Seminar: Cerebellar Modelling Techniques for Mobile Robot Control in a Delayed Sensory Environment.

ITEE seminar: David Collins, 03.00PM, Thu 23 Oct 2003

Cerebellar Modelling Techniques for Mobile Robot Control in a Delayed Sensory Environment.

Speaker: David Collins, (speaker organisation unavailable)

When: 03.00PM, Thursday 23 Oct 2003

Venue: 78-420

Host: A/Prof Janet Wiles

Abstract:

  Fast and accurate movement control for a system exhibiting
  significant feedback delay is traditionally a difficult problem to
  solve. In biological systems, it is thought that a part of the brain
  called the cerebellum overcomes such difficulties. This presentation
  outlines the use of three cerebellar learning techniques (CEL, TEL
  and SEL) used to overcome the effects of sensory delay in the
  control of a simulated mobile robot. All models use Albus's CMAC
  neural network as the central adaptive element.

  The three models are able to control the robot during a
  high-accuracy coordinated kicking manoeuvre, despite significant
  feedback delay. The Command Error Learning (CEL) technique achieves
  this by learning commands provided by a skilled teaching module. The
  Trajectory Error Learning (TEL) model evolves it's own commands
  given a desired trajectory profile and the State Error Learning
  (SEL) architecture learns to predict the actual (non-delayed) state
  of the robot, for use by a conventional feedback controller.

  All models produced trajectories equal to or better than the
  non-delayed, conventional feedback controller. When used in the
  presence of feedback delay, several interesting properties which
  equate well with biological evidence were observed.

Biography:

 
  David Collins received a B.E. (Hons I) in Computer Systems
  Engineering from the University of Queensland in 1996 and is in the
  final stages of his PhD at the University of Queensland. He expects
  to graduate in December 2003. His research interests include
  Robotics, Machine Learning, Control Systems and Neuroscience. His
  thesis investigates methods of overcoming sensory delay in control
  systems, by using predictive control techniques inspired by the
  biology of the human cerebellum. David is currently working as a
  consulting engineer.

Contact:

A/Prof Janet Wiles, seminar host (janet@itee.uq.edu.au)
or Guido Governatori (ITEE seminar co-ordinator) (guido@itee.uq.edu.au)

ITEE seminar web page: http://www.itee.uq.edu.au/~seminar


[All seminars]