Title: Growing RBF networks for learning reactive behaviours in mobile robotics

Authors: Jun Li, Tom Duckett

Addresses: Centre for Applied Autonomous Sensor Systems, Department of Technology, Orebro University, SE-701 82 Orebro, Sweden. ' Centre for Applied Autonomous Sensor Systems, Department of Technology, Orebro University, SE-701 82 Orebro, Sweden

Abstract: This paper investigates a learning system based on growing Radial Basis Function (RBF) networks for acquiring reactive behaviours in mobile robotics. The learning algorithm integrates unsupervised and supervised learning, directly mapping the sensor information to the required motor action. The learning system is evaluated through a number of experiments on a real robot. The experimental results show that our learning system can learn a wide range of robot behaviours from simple tasks to complex tasks and demonstrate that the task need not be known at the programming time. This means that many different behaviours could potentially be acquired by the same learning architecture, thus dramatically reducing the development cost of autonomous robotic systems.

Keywords: behaviour learning; growing RBF networks; mobile robots; offline learning; online learning; unsupervised learning; supervised learning; reactive behaviour; robot learning; robot behaviours; vehicle autonomous systems; autonomous vehicles; neural networks.

DOI: 10.1504/IJVAS.2006.012213

International Journal of Vehicle Autonomous Systems, 2006 Vol.4 No.2/3/4, pp.285 - 307

Available online: 28 Jan 2007 *

Full-text access for editors Access for subscribers Purchase this article Comment on this article