Використання штучного інтелекту для моделювання поведінки NPC у середовищі Unreal Engine 5

Abstract

The modern gaming industry faces the challenge of creating realistic behav-ior for virtual characters that interact with players and the environment. Non-Play-able Characters (NPCs) are crucial in this context, as they operate based on pre-defined algorithms and make decisions according to programmed scenarios. Tra-ditional methods, such as scripted systems and Finite State Machines (FSM), have limitations in adaptability and interaction. This paper explores the use of Artificial Intelligence (AI) in Unreal Engine 5 to enhance NPC behavior through systems like Behavior Trees and Environmental Query System (EQS). Additionally, the integration of neural networks and reinforcement learning is discussed, offering potential for more adaptive and realistic NPCs. The paper concludes that AI-driven NPCs significantly improve player immersion and provide more dynamic and interactive gaming experiences.