摘要
针对空战智能决策模型生成问题,首先,分析了空战智能决策构建的意义,并对基于博弈论的方法、基于规划理论的方法以及专家系统与机器学习方法等空战决策模型生成技术进行了对比分析;然后,阐述了博弈对抗规则建模、机器学习优化、智能算法模型验证评价、大样本并行仿真推演和柔性平台集成接口设计等关键技术;最后,对智能技术空战应用的发展趋势进行了总结。
Aimed at the generation problem of air combat intelligent decision model,firstly,the significance of air combat intelligent decision construction is analyzed,and the air combat decision model generation technologies are compared,such as the methods based on game theory,the methods based on planning theory,and the expert system and machine learning methods.Then,the key technologies are described,such as the game confrontation rule modeling,the machine learning optimization,the intelligent algorithm model validation and evaluation,the large sample parallel simulation deduction,and the flexible platform integration interface design.Finally,the development trend of intelligent technology in air combat is summarized.
作者
陈希亮
李清伟
孙彧
CHEN Xiliang;LI Qingwei;SUN Yu(College of Command Information System,Army Engineering University,Nanjing 210007,China;The 28th Research Institute of China Electronics Technology Group Corporation,Nanjing 210007,China;Unit 31102 of PLA,Nanjing 210000,China)
出处
《指挥信息系统与技术》
2021年第2期1-6,共6页
Command Information System and Technology
基金
国家自然科学基金(61806221)、装备发展部“十三五”预研课题(31505550302)
国防科技重点实验室基金(6142101180304)资助项目。
关键词
博弈对抗
深度强化学习
智能空战
game confrontation
deep reinforcement learning
intelligent air combat