An Application Of Artificial Intelligence In Computer Game Software

Recent advances in computer science at both the software and hardware levels,especially in 3D visualization techniques, have tremendously enhanced the graphicalexpressiveness of computer games. This makes artificial intelligence in computergames a bottleneck for improving gaming experience. Furthermore, the real timeattribute of computer games requires that AI calculation consume as less as possibleresourcestofertilizeothermodulesofthegame.After introducing general architecture of a game engine, this paper firstlysummarizes some AI techniques frequently used in computer games but rarelydocumented, and discusses their possible extension, optimization and integration withother modules of the game. The main focus has been on actor movement control andpathplanning.Secondly,itsurveyssomeAItechniquessuitableforhandlinghighlevelcontrol and decision making, such as finite state machines, rule-based systems andinfluence mapping, and discusses their possible implementation and improvement ingames. Moreover, by analyzing some case studies, it discusses the feasibility ofimplementing learning and adaptive AI in computer games. Finally, this paperproposes a methodology that implements a script-based AI engine in terms ofdata-driven,andpresentsafinite-state-machine-basedscriptsystem.