Autonomous Game Character Based On Agent

With the development of computer software and hardware technologies, the industry of digital games has become more and more prosper. Compared with the graphic images and physical effect, though the game AI is better than before, it is a little backward and still has more space to develop. So, after the study of game character, we designed a game character architecture, which is called"autonomous character architecture". The main work is as follow:Designed autonomous character architecture. Focusing on the problems of no unified architecture, high coupling degree and difficult to implement when designing the intelligent game character, we designed a hierarchical architecture based on Agent technologies, which is called autonomous character architecture. This architecture divides the game character into geometric model, motor model, behavioral model, intention model, and perception system.Implemented the motor model by some algorithms, which are used to simulate the animals'behaviors. Implemented the behavior model based on the composite pattern, whitch makes the atomic behavior and composite behavior have identical interfaces, and endows the autonomous character with a memory.Implemented a fuzzy logic system and then used it to decide the autonomous character's intention based on predefined logic rules. Compared the differences between"point-detected"and"space-separated"perception algorithm and used the"point-detected"algorithm to implement the visual perception of soccer game characters.In short, the research of autonomous character architecture provides game developers a good method to program the intelligent character which could tell itself what to do and how to do. And, a soccer game based on this a architecture has proven its effectiveness.