Logo image
A biologically inspired quadruped: an educational platform to study walking machines
Thesis   Open access

A biologically inspired quadruped: an educational platform to study walking machines

Karl Bernet
Internship Report (Bachelor of Engineering), Murdoch University
2009
pdf
Bernet_2009.pdfDownloadView
Author’s Version Open Access

Abstract

This thesis addresses the main objective of providing a suitable platform to be used in the study of embedded systems and control of complicated process mechanisms through the development of a working quadruped walking machine that has many practical representations of a dog. The vision is that the in future courses that by providing a functioning basic walking machine platform for students, though coding in Forth, they will be able to ass or alter the control behaviour. This could be a simple change in the gate delay time or switching the gait pattern to a trot or a run. Yet the change could be extended to optimise for a certain goal such as maximum distance covered or minimum energy expended. Once the new code is added then the emergent behaviour can be observed. Shi Peng (Peng 2006) had hinted at such an inspiration in his thesis and later when interviewed expressed frustration at the complexity of the machine he had built. This projects sets to restore the quadrupedal robot built by Shiqi to provide Murdoch University a platform for continuing studies in control theory, robotics and sensor integration. The platform itself is remarkable because it makes intelligent use of low cost and low technology products (such as the 68HC11 processor) and yet was able to make possible the complex interaction required for locomotion which on the other robots could only be achieved by using off-board computers or multiple processors. In this thesis a new walking machine algorithm is presented, the behaviour of which more aligned to the biological movement for a quadruped and also validated that the robot built by Shiqi is a suitable platform to be used in the study of embedded systems and control of complicated process mechanisms.

Details

Metrics

439 File views/ downloads
115 Record Views
Logo image