Event-triggered constrained neural critic control of nonlinear continuous-time multiplayer nonzero-sum games

• Published in: Information sciences Vol. 631; pp. 412 - 428
• Main Authors: Li, Menghua, Wang, Ding, Zhao, Mingming, Qiao, Junfei
• Format: Journal Article
• Language: English
• Published: Elsevier Inc 01.06.2023
• Subjects: Adaptive dynamic programming; Asymmetric input constraints; Event-based optimal control; Multiplayer nonzero-sum games; Neural critic control; Asymmetric input constraints; Multiplayer nonzero-sum games; Adaptive dynamic programming; Neural critic control; Event-based optimal control
• ISSN: 0020-0255; 1872-6291
• DOI: 10.1016/j.ins.2023.02.081

In this article, by exploiting the event-triggered neural critic control scheme, the optimal control question of nonlinear continuous-time multiplayer nonzero-sum games with asymmetric control constraints is investigated. First, to satisfy the asymmetric constraints, we propose an appropriate nonquadratic function. Also, the time-based coupled Hamilton-Jacobi (HJ) equations and the optimal control policies are deduced. To save communication resources and improve control efficiency, the event-driven mechanism is innovated. Further, the event-triggered optimal control strategies and the event-triggered coupled HJ equations are given. Afterwards, a novel triggering condition is established, and the critic neural networks are built to evaluate the optimal cost functions, thereby obtaining the event-triggered near-optimal controls. During neural critic learning, a new weight adjustment rule is proposed. Moreover, by exploiting the Lyapunov approach, the stability of the system state and the weight estimation errors is guaranteed. Finally, two examples are given to verify the efficacy of the established event-triggered neural critic control scheme.