Optimal Scaled Average Consensus Control of Discrete-Time Multi-Agent Systems Based on Adaptive Dynamic Programming
Peiyu Zhai, Mengji Shi, Weihao Li, Jiangfeng Yue, Kaiyu Qin, Jinwei Bai, Ruhao Hua · 2023
This paper investigates the data-driven scaled average consensus control problem of discrete-time (DT) multi-agent systems. An optimal distributed control scheme is designed to solve this problem and enable model-free online training based on adaptive dynamic programming (ADP). Firstly, the optimal scaled average consensus problem is defined via Bellman optimality principle. Then, the ADP algorithm is designed to calculate a proper approximate solution for the discrete-time Hamilton-Jacobi-Bellman (DT-HJB) equation based on generalized policy iteration. In addition, critic-actor neural networks are well constructed to fit the iterative control law and performance indices to realize the model-free online training for the data-driven control. Finally, some simulation results are given to verify the validity of the proposed optimal distributed control scheme.