Abstract
This paper deals with the design of an integrated decision-making system for robot soccer. Three main tasks groups of decision-making are discussed based on an analysis of artificial intelligence means used by individual teams-choice of playing strategy, navigation, and kicking. The paper shows that crisp splitting of these decision-making groups is suspicious, and, instead, the use of a convenient implementation means to keep them together is recommended. In this case, fuzzy cognitive maps are introduced whose role is to integrate decision tasks, which is the main contribution of this paper. The paper describes a new design for kicking decisions. The system was successfully tested in the simulator Webots and some concluding remarks are made.
Original language | English |
---|---|
Pages (from-to) | 156-163 |
Number of pages | 8 |
Journal | Journal of Advanced Computational Intelligence and Intelligent Informatics |
Volume | 15 |
Issue number | 2 |
DOIs | |
Publication status | Published - Mar 2011 |
Externally published | Yes |
Keywords
- Fuzzy cognitive map
- Marketdriven approach
- Multiagent systems
- Nearness diagram navigation
- Robot soccer