A Vision-Based Reinforcement Learning For Coordination Of Soccer Playing Behaviors

摘要

A method is proposed which acquires a purposive behavior of shooting a ball into the goal avoiding collisions with an enemy. In [Asada et al., 1994], we have presented the soccer robot which learned to shoot a ball into the goal without any enemy, using the Q-learning, one of the reinforcement learning methods. Since a simple extension of the method is not practical due to its huge state space, two different behaviors each of which are previously learned independently are combined into a coordinated behavior. One is a shooting behavior without any enemy, and the other is a collision avoiding behavior against a moving obstacle. Three kinds of coordinations are considered; simple sum of the two action value functions, switching the two functions according to the situation, and the Q-learning with the previously learned behaviors. The simulation results are shown and a discussion is given. 1