A Preliminary Research on Decision Model Based on Bayesian Techniques for an NPC in Computer Games

In modern computer games, a NPC (non-player character) can make choices. In the previous researches, a NPC can have the certain knowledge in a game world; this will reduce the believability of perception for a NPC. A decision model based on Bayesian network and fuzzy logic for a NPC is set up in this paper; a NPC can partially know domain knowledge. A behavior animation of a virtual dog by the model is realized on PC, the results show that the virtual dog can behave believable in watching a house.