Reinforcement Learning With Adjustable Convergence Rate for Data-Based Nonlinear Control - 文章详情页

作者：

Tang, Guohan (Tang, Guohan.) | Wang, Ding (Wang, Ding.) | Li, Xin (Li, Xin.) | Ren, Jin (Ren, Jin.) | Liu, Nan (Liu, Nan.)

收录：

EI Scopus

摘要：

In　this　paper,　a　value-iteration-based　off-policy　Q-learning　algorithm　is　developed.　The　proposed　algorithm　solves　the　optimal　regulation　problem　of　nonlinear　systems　with　unknown　dynamics.　Under　the　off-policy　mechanism,　the　algorithm　utilizes　the　behavioral　policy　for　full　exploration,　which　is　beneficial　to　avoid　the　target　policy　from　falling　into　the　local　optimal　solution.　In　addition,　a　relaxation　factor　is　introduced　to　adjust　the　convergence　rate　of　the　cost　function　sequence.　To　implement　the　algorithm,　the　critic　network　and　the　action　network　are　used　to　approximate　the　optimal　Q-function　and　the　optimal　control　policy,　respectively.　Finally,　a　simulation　example　is　presented　to　demonstrate　the　effectiveness　of　the　proposed　algorithm.　©　2024　IEEE.

关键词：

Dynamic programming Learning systems Reinforcement learning Learning algorithms Nonlinear systems Cost functions Adaptive control systems Iterative methods Discrete time control systems

作者机构：

[ 1 ] [Tang, Guohan]Beijing University Of Technology, Faculty Of Information Technology, Beijing, China
[ 2 ] [Wang, Ding]Beijing University Of Technology, Faculty Of Information Technology, Beijing, China
[ 3 ] [Li, Xin]Beijing University Of Technology, Faculty Of Information Technology, Beijing, China
[ 4 ] [Ren, Jin]Beijing University Of Technology, Faculty Of Information Technology, Beijing, China
[ 5 ] [Liu, Nan]Beijing University Of Technology, Faculty Of Information Technology, Beijing, China

通讯作者信息：

电子邮件地址：

查看成果更多字段

成果类型
所属机构

所有年份指定年份从至