Abstract
In this article, a novel event-based adaptive dynamic programming algorithm is developed to solve the infinite-horizon (H∞) control problem for continuous-time input-constrained nonlinear systems. The communication complexity and computation load of the system can be obviously reduced through the state-dependent triggering condition of the state and control strategy. Moreover, an event-triggered single-network adaptive critic method is utilized, with the aim of obtaining the approximate optimal solution of the Hamilton-Jacobi-Isaacs (HJI) equation. In the process of algorithm learning, only weights of the critic network need to be adjusted, and the number of updated weights can be reduced by eliminating the action network. By means of Lyapunov theory, the stability of the designed closed-loop control system and the convergence of weights of the critic network are demonstrated. Finally, a real power system simulation example is presented to verify the stability, the capability to suppress disturbance and the effectiveness of the proposed method.
Original language | English |
---|---|
Title of host publication | Proceedings of 2023 International Annual Conference on Complex Systems and Intelligent Science (CSIS-IAC) |
Number of pages | 6 |
Publisher | IEEE |
Publication date | 2023 |
Pages | 709-714 |
ISBN (Electronic) | 979-8-3503-0900-3 |
DOIs | |
Publication status | Published - 2023 |
Event | 2023 International Annual Conference on Complex Systems and Intelligent Science - Shenzen, China Duration: 20 Oct 2023 → 22 Oct 2023 |
Conference
Conference | 2023 International Annual Conference on Complex Systems and Intelligent Science |
---|---|
Country/Territory | China |
City | Shenzen |
Period | 20/10/2023 → 22/10/2023 |
Keywords
- Adaptive dynamic programming (ADP)
- H∞ control
- Zero-sum game
- Event-based control
- Input constraints