Agent-based Control Architecture for Mobile Robot Navigation

Abstract

The problems of mobile robotics relate to the design and the study of functions of perception decision and action and the coherent integration of these functions in a physical machine. This article proposes to use the paradigm multi-agents and fuzzy logic in the design of the control architecture for the mobile robot navigation in an environment constrained. The control architecture is conceived to tackle various problems produced during navigation. It consists of four agents : agent perception, agent feasibility, agent of locomotion and agent fuzzycontroller. This last uses in its reasoning the fuzzy logic principle. The method used enabled us to have an intelligent system able to adapt to all the changes in the environment of navigation.