Title :
Behavior learning to predict using neural networks (NN): Ttowards a fast, cooperative and adversarial robot team (RoboCup)
Author :
Chohra, A. ; Schöll, P. ; Kobialka, H.-U. ; Hermes, J. ; Bredenfeld, A.
Author_Institution :
Inst. for Autonomous Intelligent Syst. (AiS), GMD - German Nat. Res. Center for Inf. Technol., St. Augustin, Germany
Abstract :
To build a fast, cooperative and adversarial robot team (RoboCup), prediction behaviors became necessary. In the paper, a behavior learning method using neural networks (NN) is developed to enhance the behavior of GMD mobile robots. In fact, the suggested NN called NN-Prediction learns to predict successfulness of the elementary behavior "Kick" the ball towards the goal in order to act as consequence. The training is carried out by the supervised gradient back-propagation learning paradigm. This NN-Prediction has been specified on the Dual Dynamics Designer, to be thereafter implemented and tested on both the Dual Dynamics Simulator and GMD mobile robots, and analyzed on the Real-Time Trace Tool. NN-prediction demonstrated, during the 4th World Championships RoboCup 2000, cooperative and adversarial behaviors especially face to situations where the successfulness of "Kick" is not guaranteed. Then, a discussion is given dealing with the suggested prediction behavior and how it relates to some other works
Keywords :
backpropagation; cooperative systems; mobile robots; multi-robot systems; neural nets; Dual Dynamics Designer; Dual Dynamics Simulator; GMD mobile robots; NN-Prediction; Real-Time Trace Tool; RoboCup 2000; behavior learning; cooperative adversarial robot team; kick; neural networks; supervised gradient back-propagation learning; Artificial intelligence; Biological neural networks; Cognitive science; Intelligent robots; Intelligent sensors; Intelligent systems; Learning systems; Mobile robots; Neural networks; Psychology;
Conference_Titel :
Robot Motion and Control, 2001 Proceedings of the Second International Workshop on
Conference_Location :
Bukowy Dworek
Print_ISBN :
83-7143-515-0
DOI :
10.1109/ROMOCO.2001.973435