Skip navigation

Multiple Mobile Robots Controlled by Artificial Neural Networks

Multiple Mobile Robots Controlled by Artificial Neural Networks

Seals, Richard and Seals, T. A. (2017) Multiple Mobile Robots Controlled by Artificial Neural Networks. [Working Paper] (Submitted)

[img]
Preview
PDF (Author Submitted Manuscript)
17144 SEALS_Multiple_Mobile_Robots_2017.pdf - Submitted Version

Download (345kB) | Preview

Abstract

Multiple small mobile robots have been created that were controlled by individual artificial neural networks. Each mobile robot was self-contained and capable of independent actions, as determined by the on-board artificial neural network. Information about the environment was collected from sensors mounted on each individual mobile robot chassis. Different sensors were available that were capable of providing information about different aspects of the environment. Currently there were sensors for detecting and following a black line as well as short range distance sensors for detecting and interacting with objects and other mobile robots. The artificial neural networks on the individual mobile robots were all provided with the same training data and a standard back-propagation training algorithm was used. However the randomised component of training the artificial neural networks did mean that there could have been subtle differences in the responses of individual mobile robots to the same sensor data. This effect was eliminated when needed by using an off-line training process and programming all the mobile robots with the same trained ANN. The small group of mobile robots was used to investigate two simple aspects of swarm behaviour; that of flocking and also of follow-my-leader, which are examples where the swarm appeared to operate with more intelligence than the individual members.

Item Type: Working Paper
Uncontrolled Keywords: Mobile robots, Artificial Neural Networks
Subjects: T Technology > TA Engineering (General). Civil engineering (General)
Faculty / Department / Research Group: Faculty of Engineering & Science
Faculty of Engineering & Science > Department of Engineering Science
Faculty of Engineering & Science > Future Technology and the Internet of Things
Last Modified: 31 May 2017 08:57
Selected for GREAT 2016: None
Selected for GREAT 2017: None
Selected for GREAT 2018: None
URI: http://gala.gre.ac.uk/id/eprint/17144

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year

View more statistics