Adaptive Multi-Step Evaluation Design With Stability Guarantee for Discrete-Time Optimal Learning Control

doi:10.1109/JAS.2023.123684

CORC > 自动化研究所 > 中国科学院自动化研究所 > 学术期刊 > IEEE/CAA Journal of Automatica Sinica

	Adaptive Multi-Step Evaluation Design With Stability Guarantee for Discrete-Time Optimal Learning Control
	Ding Wang; Jiangyu Wang; Mingming Zhao; Peng Xin; Junfei Qiao
刊名	IEEE/CAA Journal of Automatica Sinica
	2023
卷号	10 期号:9 页码:1797-1809
关键词	Adaptive critic artificial neural networks Hamilton-Jacobi-Bellman (HJB) equation multi-step heuristic dynamic programming multi-step reinforcement learning optimal control
ISSN号	2329-9266
DOI	10.1109/JAS.2023.123684
英文摘要	This paper is concerned with a novel integrated multi-step heuristic dynamic programming (MsHDP) algorithm for solving optimal control problems. It is shown that, initialized by the zero cost function, MsHDP can converge to the optimal solution of the Hamilton-Jacobi-Bellman (HJB) equation. Then, the stability of the system is analyzed using control policies generated by MsHDP. Also, a general stability criterion is designed to determine the admissibility of the current control policy. That is, the criterion is applicable not only to traditional value iteration and policy iteration but also to MsHDP. Further, based on the convergence and the stability criterion, the integrated MsHDP algorithm using immature control policies is developed to accelerate learning efficiency greatly. Besides, actor-critic is utilized to implement the integrated MsHDP scheme, where neural networks are used to evaluate and improve the iterative policy as the parameter architecture. Finally, two simulation examples are given to demonstrate that the learning effectiveness of the integrated MsHDP scheme surpasses those of other fixed or integrated methods.
内容类型	期刊论文
源URL	[http://ir.ia.ac.cn/handle/173211/52369]
专题	自动化研究所_学术期刊_IEEE/CAA Journal of Automatica Sinica
推荐引用方式 GB/T 7714	Ding Wang,Jiangyu Wang,Mingming Zhao,et al. Adaptive Multi-Step Evaluation Design With Stability Guarantee for Discrete-Time Optimal Learning Control[J]. IEEE/CAA Journal of Automatica Sinica,2023,10(9):1797-1809.
APA	Ding Wang,Jiangyu Wang,Mingming Zhao,Peng Xin,&Junfei Qiao.(2023).Adaptive Multi-Step Evaluation Design With Stability Guarantee for Discrete-Time Optimal Learning Control.IEEE/CAA Journal of Automatica Sinica,10(9),1797-1809.
MLA	Ding Wang,et al."Adaptive Multi-Step Evaluation Design With Stability Guarantee for Discrete-Time Optimal Learning Control".IEEE/CAA Journal of Automatica Sinica 10.9(2023):1797-1809.