Mixed Driven Iterative Adaptive Critic Control Design Towards Nonaffine Discrete-Time Plants. Issue 2 (2020)

• Record Type: Journal Article
• Authors: Wang, Ding; Ha, Mingming
• Is Part Of: IFAC-PapersOnLine. Volume 53:Issue 2(2020)
• Page Start: 3803
• Page End: 3808
• Publication Date: 2020
• Subjects: Discrete-time nonlinear dynamics -- iterative adaptive critic algorithm -- mixed driven design -- neural optimal control -- triggering condition; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2020.12.2071
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: In this paper, an effective mixed driven framework is constructed involving both data and event considerations. The primary purpose lies in that the mixed driven iterative adaptive critic method is established to address approximate optimal control towards discrete-time nonlinear dynamics. The neural dynamic programming technique is inventively integrated with the mixed driven architecture, such that the knowledge of the controlled plant is needless and the number for updating control inputs is prominently reduced. A triggering threshold is also designed with theoretical guarantee, which renders that the control signals can be updated conditionally. Through carrying out simulation studies with comparisons, the superiority of the present near-optimal regulation approach is confirmed at last.