Real time reactive strategies based on potential fields for robot soccer
Biliang Zhong, Qi Zhang, Yimin Yang · 2004
Soccer robot games are rapidly becoming a major research tool in field of AI, machine vision and motion control. The dynamism of the environment in the size of state space makes such a design challenging. In this paper we develop optimal reactive strategies for the game using groundtruth data from experts, with potential fields as base of the strategies. Experiment shows that the strategy has good active characteristics and can satisfy the game demand.