Community Research and Development Information Service - CORDIS

Abstract

This paper describes a self-organising controller for a mobile robot. Equipped with a controller based on the self-organising map, a mobile robot can combine its sensory information without any outside assistanceto obtain a representation that is directly applicable for navigation and obstacle avoidance. The feasibility of the proposed self-organising control system is tested with a Nomad mobile robot simulator. The experiments demonstrate that the control system learns quickly to select suitable actions for sensory situations, adapts easily to new environments, and collides less and less frequently while navigating.

Additional information

Authors: HEIKKONEN J, JRC Ispra (IT);SURAKKA M, Mechanical Engineering Laboratory, Machine Intelligence Division, Ibaraki-Ken (JP);RIEKKI J, Mechanical Engineering Laboratory, Machine Intelligence Division, Ibaraki-Ken (JP)
Bibliographic Reference: Paper presented: International Conference on Engineering Applications of Neural Networks, Espoo (FI), August 21-23, 1995
Availability: Available from (1) as Paper EN 39237 ORA
Record Number: 199511235 / Last updated on: 1995-10-13
Category: PUBLICATION
Original language: en
Available languages: en