A VISION-BASED APPROACH FOR LEARNING A GOAL-REACHING BEHAVIOR
Grazia Cicirelli, Tiziana D’Orazio, Cosimo Distante, Giovanni Attolico · 1998
This paper describes a navigation approach that is based on a reinforcement learning algorithm, which allows a mobile robot to autonomously move in an unknown indoor environment and reach a particular location. The system is a control architecture merging visual information and sonar readings. Designed for learning the best movement for reaching a location, the robot is able to learn efficient behavioral sequences and improve its learning incrementally.