Data-Driven Cooperative Output Regulation of Linear Discrete-Time Multiagent Systems With Unknown Dynamics
针对动力学未知的线性异构离散时间多智能体系统，提出一种数据驱动分布式控制协议，实现渐近跟踪领导者参考信号并抑制外部扰动，通过求解凸优化问题获得控制增益。
This technical article conducts the research on the cooperative output regulation problem (CORP) of linear heterogeneous discrete-time multiagent systems (MASs) with the constraints of unknown dynamics. On the basis of the data-driven scheme, a state-based distributed control protocol is constructed to achieve the asymptotic tracking for leader system’s reference signals while getting rid of external disturbances. Furthermore, the result is extended to the output-based case based on the least square approach. In order to cope with the unknown leader system, an observer is put forward to estimate the leader’s state after obtaining the system matrices for the exosystem by utilizing the persistently exciting data. To further overcome the difficulties of the unknown dynamics of each follower, input-output relationship is constructed to infer the system matrices. Moreover, the solution to the unknown regulator equations is provided by solving the data-dependent matrix equations. Ultimately, the control gains are derived by solving the convex optimization problem. The affirmation of the whole design scheme is demonstrated by two simulation cases.