智能博弈对抗算法及其在情报领域中的应用  

Study of intelligent game adversarial algorithms and their applications in the intelligence field

作　　者：刘赓 刘星 LIU Geng;LIU Xing(National University of Defense Technology,Nanjing 210000,China;Naval Aviation University,Yantai 264000,China)

机构地区：[1]国防科技大学外国语学院,江苏南京210000 [2]海军航空大学岸防兵学院,山东烟台264000

出　　处：《指挥控制与仿真》2024年第6期49-54,共6页Command Control & Simulation

摘　　要：智能博弈对抗算法不仅充分利用了博弈模型的刻画精度,还通过神经网络的强大计算能力和强化学习的试错机制求解均衡解,使得智能博弈对抗算法在诸多领域都取得了不错的效果。通过多智能体博弈学习、多智能体博弈强化学习和多智能体博弈深度强化学习三个层面对智能博弈对抗算法进行了系统梳理,并结合情报领域的工作特点分析,论证了智能博弈对抗算法运用在情报领域的可行性和必要性,最后给出了智能博弈对抗算法在情报领域的具体应用以及后续提升质效的有效措施。Intelligent game adversarial algorithms not only make full use of the portrayal accuracy of the game model,but also solve the equilibrium solution through the powerful computational ability of neural network and the trial-and-error mechanism of reinforcement learning,which makes the intelligent game adversarial algorithms achieve good results in many fields.Through the multi-intelligence body game learning,multi-intelligence body game reinforcement learning and multi-intelligence body game deep reinforcement learning three levels of intelligent game confrontation algorithm is systematically sorted out,and the corresponding mapping with the intelligence field of work,demonstrates the feasibility and necessity of intelligent game confrontation algorithm in the field of intelligence,and finally gives the specific application of the intelligent game confrontation algorithm in the field of intelligence and the effective measures of the follow-up to improve the quality and efficiency.Finally,it gives the specific application of intelligent game confrontation algorithm in the field of intelligence,as well as the effective measures to improve the quality and efficiency.

关 键 词：智能对抗 博弈论 强化学习 情报处理 

分 类 号：E917[军事]