Multilayer Decision-Making Framework for Adversarial Game Scenarios
Zikang Xie, Boyang Chen, Hongyu Wu, Yunhe Pan, Jueming Zhao, Wei Wang · 2024
This article proposes a hierarchical decision-making framework designed for robotic agents engaged in adversarial gaming scenarios, to improve adaptability and decision-making efficiency in dynamic and complex environments. The framework consists of functional, task, decision, and information perception layers, complemented by a physical layer that simulates real-world conditions. By encapsulating local tasks, the framework enhances the stability of application scenarios, allowing the decision-making process to focus on information processing within adversarial games. Experimental verification was conducted within a simulated environment, confirming the frame-work's effectiveness in achieving efficient robotic control and advantageous results. The research not only provides theoretical backing but also delivers practical evidence for the implementation of robots in adversarial games, possessing considerable academic and practical significance.