par Patel, Mukesh J.;Colombetti, Marco;Dorigo, Marco
Référence Intelligent Automation and Soft Computing, 1, 1, page (29-42)
Publication Publié, 1995
Article révisé par les pairs
Résumé : Industrial automation calls for behavioral intelligence, that is, a mixture of flexibility, robustness and adaptiveness of robot behavior. We argue that efficient machine learning techniques can be a valuable tool for achieving behavioral intelligence. As a case study we.1pply ALECSYS, an implementation of a learning classifier system on a net of transputers, to a gross-motion problem for an industrial manipulator (an IBM 7547 with a SCARA geometry). A simple simulation environment allows us to experiment with different sensor configurations, and to obtain an initial coarse approximation of the robot's controller through learning. The controller is subsequently refined through a learning session run on the physical robot. As a whole, our work demonstrates some interesting distinctive features of the evolutionary computation approach, viewed as a possible alternative to classical methods of software development. © 1995 Taylor & Francis Group, LLC.