Abstract The aim of this paper is to introduce a multi-agent architecture for an autonomous robot control oriented to obtain high design modularity and high parallelism of decision processes. A multi-agent approach allows us to design several specialized agents and each of them is optimized in order to solve a single part of the whole task. Each agent is able to cooperate with other agent exchanging information, so the robot is able to make the appropriate policy of management to reach a prefixed goal. We tested the proposed architecture with a simple application: an autonomous mobile robot performs object retrieval tasks into unknown structured environments. The robot is not given a priori map of the environments and it can explore the unknown environments, equipped with a simple b/w camera sensor.
A Neural Multi-Agent Architecture for an Autonomous Mobile Robot Control
1999-01-01
6 pages
Article/Chapter (Book)
Electronic Resource
English
Real-time Control Architecture for an Autonomous Mobile Robot
British Library Conference Proceedings | 1994
|European Patent Office | 2022
|