指挥控制与仿真2026，Vol.48Issue(3)：41-48,8.DOI:10.3969/j.issn.1673-3819.2026.03.005
基于强化学习的直升机智能博弈方法研究
Research on helicopter intelligent game method based on reinforcement learning
于若颜 1吕增岁1
作者信息
- 1. 中国直升机设计研究所,江西 景德镇 333000
- 折叠
摘要
Abstract
Focusing on the problem of helicopter formation intelligent game,this paper uses the idea of combining rule rea-soning and reinforcement learning to propose a knowledge-data driven helicopter formation intelligent game decision-making method.In view of the simple situation,the distributed knowledge expression method is used to construct an expert rule base to quickly complete the analysis and decision-making.For complex or unknown situations,a reinforcement learning intelli-gent game model is constructed based on the Multi-Agent Proximal Policy Optimization(MAPPO)algorithm to make optimal decisions,and the coordination of helicopter formations is effectively improved through centralized training and distributed execution mechanism.Finally,the decision-making task of the red and blue game in the design scenario is completed in the simulation platform,and the efficiency is evaluated according to the deduction data,which verifies the effectiveness and practicability of the algorithm.关键词
智能博弈/强化学习/直升机编队协同/效能评估Key words
intelligent gaming/reinforcement learning/helicopter formation coordination/evaluation of effectiveness分类
军事科技引用本文复制引用
于若颜,吕增岁..基于强化学习的直升机智能博弈方法研究[J].指挥控制与仿真,2026,48(3):41-48,8.
