Cooperative reinforcement learning based on zero-sum games

Abstract

The objective of this paper is to develop a strategy system in a robot soccer system with cooperative ability which is improved by self-learning. A reinforcement learning method based on the zero-sum game theory is developed in this paper. It enforces learning systems to choose an appropriate strategy complying with the opponent’s actions. In order to achieve the purpose of cooperation, the system consists of two sub systems, one is a role assignment system, and the other is a reinforcement learning system