终端(电信)
非线性系统
终端滑动模式
滑模控制
控制理论(社会学)
计算机科学
曲面(拓扑)
模式(计算机接口)
动态规划
自适应控制
控制(管理)
控制工程
工程类
数学
算法
人工智能
物理
电信
几何学
量子力学
操作系统
作者
Chao Jia,Xinyu Li,Xiaohua Liu,Xuanyue Shangguan
标识
DOI:10.1177/09596518241272750
摘要
For a nonlinear system, a novel complementary terminal sliding mode surface-based finite-time adaptive dynamic programming optimal control scheme (FT-NCTSMS-ADP) is proposed. To begin with, with the aim to improve convergence speed and tracking precision, a novel complementary terminal sliding mode surface (NCTSMS) is developed. Then, a specific cost function related to the NCTSMS is developed, and the initial control problem is then transformed toward a series of optimal control problems. The Hamilton-Jacobi-Bellman (HJB) problem is then solved by a single critic neural network, yielding the estimated optimal controller. Finally, using the finite-time lemma, we design a Lyapunov function and verify the system’s fast finite-time convergence. Both the simulation and the real-world experiment validate the practical significance of the proposed method.
科研通智能强力驱动
Strongly Powered by AbleSci AI