Title :
Reduction of the RoboCup state space
Author :
Lin, Liu ; Zhiqiang, Zheng
Author_Institution :
Coll. of Mechatro-Electron. & Autom., Nat. Univ. of Defense Technol., Changsha, China
Abstract :
As a standard research platform of artificial intelligence and robotics, RoboCup is gathering more and more attention. Because in the match, the environment is dynamically changing, the method of machine learning is widely used. An important problem is how to reduce the complex environment in learning, so that the number of states is decreased. The paper briefly introduces Soccer Server. Then several methods of constructing the state space are analyzed. A method of reduction of the state space is described. The method divides the environment around the client into some regions which are not equal in size. The state space is constructed with these regions.
Keywords :
client-server systems; mobile robots; multi-robot systems; RoboCup; Soccer Server; complex environment; machine learning; reinforcement learning; state space reduction; Artificial intelligence; Educational institutions; Intelligent robots; Learning; Noise reduction; Orbital robotics; Robotics and automation; Space technology; State-space methods; Working environment noise;
Conference_Titel :
Intelligent Control and Automation, 2002. Proceedings of the 4th World Congress on
Print_ISBN :
0-7803-7268-9
DOI :
10.1109/WCICA.2002.1022155
