| With the development of science and technology, virtual reality (VR) technology are widely used. Now virtual crowds in virtual reality field has become a hot research topic. therefore, in a dynamic environment, how to improve cooperation among virtual humans is a problem that we must study and solve. On this issue , football match system of virtual human provides an experiment platform, which is concentrative embodiment of virtual crowds technology and artificial intelligence and is a typical multi-agent system. In football match system of virtual human , the decision-making system is a key sub-system, which will directly influence the score of a team. So it is the focus of attention in the dissertation.In the paper the decision-making system of football match of virtual human is researched from the virtual human intelligentized aspect. the following work has been accomplished. Firstly, on studying the robot soccer decision-making system and multi-agent decision-making model, based on the virtual humans have certain characteristics of the cognitive ability, decision-making model of football match of virtual human is established, which reflects a characteristic of multi-agent teamwork with the ability to play personality; the decision-making model are analyzed. Secondly, a hierarchy architecture model of arrays system is established,the model include: formation, role, strategy, action; flow charts of autonomous decision-making processes of roles are given; correlation degree matrix (CDM) concept are proposed, which establish the relationship for cooperation among the virtual players. Thirdly, no-ball running in football match of virtual human is studied, some basic tactics for local running strategy and a method for global running strategy are presented. Finally, arrays under 433 formation is studied and simulated. And the simulated result is good. At the end of this paper, all main conclusions in dissertation and further suggestions were given.
|
PDF Full Text Request