Reactive Control Using Behavior Modelling of a Mobile Robot

Abstract

This paper deals with the reactive control of an autonomous robot which should move safely in a crowded unknown environment to reach a goal. A behavior based approach is used to realize obstacle avoidance within a neural model conceived from a set of examples of perception/action relations; supervised learning is used for the aim; while goal-reaching task is realized using a fuzzy rule-based system. A task activation module is used to generate the overall command, resulting from the fuzzy controller and the neural model. Real time simulation examples of generated path with proposed techniques are presented.