Behavior tree design of intelligent behavior of non-player character (NPC) based on Unity3D

摘要

At this stage, the rapid development of computer technology and information technology in China has provided favorable conditions for the development of the game. In order to pursue the game experience, how to use artificial intelligence in the game has become a new research hotspot. Therefore, the current situation of artificial intelligence used in games was investigated, and the principles of Unity3D game engine were studied; then the intelligent behavior model for NPC was established by using the behavior tree as the basic algorithm, and the AI architecture of the agent in the game was designed; moreover, combined with the above analysis, the behavior tree model based on Q learning algorithm was calculated, and the application of Unity3D in the game was completed; finally, a game model was developed in combination with Unity3D game engine. The results show that the behavior tree based on Unity3D game engine can realize NPC's intelligent behavior simply and efficiently, and the system can run at a good speed, which has theoretical guidance for the follow-up research of game artificial intelligence and simulation training.