Towards Autonomous Decision Making in Multi-agent Environments Using Fuzzy Logic

摘要

In this paper, a step-by-step procedure for designing a fuzzy decision making system for RoboCup agents is introduced and implemented. By using this mechanism agents can determine, which direction to look at. The development process and debugging of this approach are much less time consuming and very simple to follow, in comparison to other analytical hand coded implementations with complex conditions and many parameters in the code. The results show that, the methodology can be employed in many decision-making problems like the soccer agent's case and other potential fields, to decrease the development time and improve the efficiency of a decision making system. Up until now, fuzzy systems have been used rarely by any of the participating teams in the annual RoboCup competitions. This paper could serve as the first inception for the design of a RoboCup agent, which uses internal fuzzy systems for many of its decision-making tasks.