动态规划
汉密尔顿-雅各比-贝尔曼方程
跟踪误差
控制理论(社会学)
控制器(灌溉)
计算机科学
人工神经网络
最优控制
数学优化
有界函数
约束(计算机辅助设计)
数学
控制(管理)
人工智能
农学
生物
机器学习
数学分析
几何学
作者
Rui Tang,Biao Luo,Yuxin Liao
出处
期刊:Neurocomputing
[Elsevier]
日期:2023-11-01
卷期号:557: 126711-126711
被引量:1
标识
DOI:10.1016/j.neucom.2023.126711
摘要
A composite control method based on adaptive dynamic programming (ADP) is developed to solve the profile tracking problem of the hypersonic gliding vehicle (HGV) with multiple constraints (flight and input constraints). First, based on the framework of the standard profile guidance method, a standard drag acceleration–velocity profile is planned as the tracking target, which satisfies the flight constraints. Second, a composite controller based on ADP and dynamic surface control (DSC) is proposed to track the planned profile and satisfy both the terminal conditions and the input constraint. An ADP method is developed to realize the optimal control through a critic neural network (NN) approximation for the solution of Hamilton–Jacobi–Bellman (HJB) equation. By using DSC method, the steady control is realized. Finally, the states of the closed-loop system and the weight estimation error are proved to be uniformly ultimately bounded, and the effectiveness of the proposed composite control method is verified via simulations.
科研通智能强力驱动
Strongly Powered by AbleSci AI