Application of the Experience-shared Distributed Q-learning Model in RoboCup
Chen Zhe-pin · Microcomputer Information · 2010
Robocup soccer simulation provides a totally distributed control and real-time asynchronous multi-agent environment.It is an excellent platform for the research of multi-agent intelligent control and artificial intelligence theory.The paper presents a model based on the experience-shared distributed Q-learning,which is used in Robocup simulation experiments and gets a good result.