论文部分内容阅读
利用自适应动态规划的在线迭代算法来研究线性多智能体系统的一致性问题.所研究的多智能体系统的状态矩阵和输入矩阵可以是已知的或未知的.首先,给出多智能体系统依赖初始时刻、终端时刻的性能指标;然后,将由初始时刻和终端时刻确定的时间段进行划分;接着,结合代数Riccati方程推导出迭代方程,并在划分后的时间段上重复地利用系统的状态信息和输入信息进行迭代计算,直至算法收敛为止;最后,利用仿真试验验证了该算法的有效性.
The online iterative algorithm of adaptive dynamic programming is used to study the consistency of linear multi-agent systems.The state matrices and input matrices of the studied multi-agent systems can be known or unknown.First, The system relies on the initial moment and the performance index of the moment of the terminal. Then, the time period determined by the initial moment and the moment of the terminal is divided. Then, the iterative equation is derived based on the algebraic Riccati equation, and the system is repeatedly used in the divided time period State information and input information are iteratively calculated until the algorithm converges. Finally, the effectiveness of the proposed algorithm is verified by simulation experiments.