Implement AI system of action game with XNA

• Main Authors: Li, Dian En, 李典恩
• Other Authors: Wang, Kuen Jie
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/41756347120206725093

碩士 === 嶺東科技大學 === 數位媒體設計研究所 === 98 === Artificial intelligence (Artificial Intelligence, AI) is one of the important sources that the player obtains fun in games, almost all the games consis of the AI. However, the game technology rapidly grow up today by software and hardware improve each other. The game image developed from early 2-Dimensional to 3-Dimensional and finally became High Definition recently. The game images develop so fast to make surprise for customer, but the AI development was much slower than game image’s. This research designed an AI of Action Game with XNA Game Studio, using module concept of the Object-oriented programming to separate the function of game from AI independently. We construct a reusable, flexible design and extensible AI system through object composition approach. Furthermore we added a Finite-state machine, Fuzzy-state machine and Artifical neural network of AI techniques, and tried to combine the traditional rules-based systems and learning-adapration systems to increase the learning ability of the traditional AI-character.