摘要
自适应动态规划方法(ADP)是一种基于强化学习框架的智能控制方法,通过函数近似技术,最终得到动态规划问题的近似最优控制策略.本文对ADP方法在航空航天飞行器鲁棒控制的研究进行综述.首先,介绍了ADP方法基本结构框架与典型算法实现原理.进一步,对于ADP方法在高超声速飞行器系统,导航制导系统以及无人机系统在鲁棒控制中的相关研究进行介绍.最后,对未来航空航天飞行器领域ADP方法的发展前景进行了分析.
Adaptive dynamic programming(ADP)is an intelligent control method based on the reinforcement learning framework.Through the function approximation technique,the approximate optimal control strategy of dynamic programming is finally obtained.The researches on the robust control of aviation aircraft based on ADP are intensively reviewed in this paper.Firstly,the basic structural framework and typical implementation principle of ADP method are introduced.Further,related researches of ADP in the robust control of hypersonic vehicles,navigation,guidance and control,micro-miniature aircraft are studied.Finally,the development prospects of ADP methodology in the field of aviation and aerospace are analyzed.
作者
穆朝絮
张勇
余瑶
孙长银
MU Chaoxu;ZHANG Yong;YU Yao;SUN Changyin(School of Electrical and Information Engineering,Tianjin University,Tianjin 300072,China;School of Automation and Electrical Engineering,Universityof Science and Technology Beijing,Beijing 100083,China;School of Automation,Southeast University,Jiangsu 210096,China)
出处
《空间控制技术与应用》
CSCD
北大核心
2019年第4期71-79,共9页
Aerospace Control and Application
基金
国家自然科学基金资助项目(61773284
61533008
61520106009)~~
关键词
自适应动态规划
航空航天飞行器
鲁棒控制
强化学习
adaptive dynamic programing(ADP)
aviation and aerospace aircraft
robust control
reinforcement learning