Role Assignment Strategy for Heterogeneous Robots Based on Fuzzy Evaluation And Fuzzy Inference

Abstract

Role assignment according to robot features is a crucial step in the coordination of multiple heterogeneous robots. This paper presents a strategy to pick heterogeneous players and form a soccer team in the RoboCup simulation environment, which is a multi-robot coordination research platform. Using fuzzy evaluation and fuzzy inference, we identify the most suitable role in a soccer team for a given heterogeneous robot. In games, the team based on this strategy showed significantly better performance compared to a team based on the previous hand-tuning solution