Metody pro definování optimální strategie pro fotbal robotů

Abstract

The aim of this thesis is to propose the artificial intelligence for our robot soccer. The thesis focuses the design and implementation of methods for generating game strategy based on the principle of reproduction in nature. A game theory methods are analyzed for automatic evaluation of good strategies. There is applied the principle of evolution for a develop the optimal strategy. There is using an automatic learning applied to a reproduction process. Furthermore, the thesis describes a automatic learning during a game, using fuzzy logic for ambiguous strategy behavior and there is also designed and implemented an algorithm for recognizing of rival strategies.