Perception, reasoning and learning of multiple agent systems for robot soccer

Abstract

Presents a supervisory control strategy for coordination of soccer playing mobile robots. Within the framework of a hierarchical control structure, three layered components of supervisor, coordinator, and executor emulate the basic three concepts of human intelligence, perception, reasoning, and learning. A small size discrete event system model is derived and implemented in the supervisor and coordinator for state-action reasoning and coordination of multiple robotic agents for a successful soccer game. Experimental results of real soccer games are given to demonstrate the feasibility and effectiveness of the developed supervisory control strategy in terms of structural simplicity and computational speed for real-time control.