Cai Yuliang, Zhang Huaguang, Wang Yingchun, Gao Zhiyun, He Qiang
IEEE Trans Neural Netw Learn Syst. 2022 Sep;33(9):4688-4698. doi: 10.1109/TNNLS.2021.3059763. Epub 2022 Aug 31.
This study investigates the bipartite fixed-time time-varying output formation-containment tracking issue for heterogeneous linear multiagent systems with multiple leaders. Both cooperative communication and antagonistic communication between neighbor agents are taken into account. First, the bipartite fixed-time compensator is put forward to estimate the convex hull of leaders' states. Different from the existing techniques, the proposed compensator has the following three highlights: 1) it is continuous without involving the sign function, and thus, the chattering phenomenon can be avoided; 2) its estimation can be achieved within a fixed time; and 3) the communication between neighbors can not only be cooperative but also be antagonistic. Note that the proposed compensator is dependent on the global information of network topology. To deal with this issue, the fully distributed adaptive bipartite fixed-time compensator is further proposed. It can estimate not only the convex hull of leaders' states but also the leaders' system matrices. Based on the proposed compensators, the distributed controllers are then developed such that the bipartite time-varying output formation-containment tracking can be achieved within a fixed time. Finally, two examples are given to illustrate the feasibility of the main theoretical findings.
本研究针对具有多个领导者的异构线性多智能体系统，研究了二分固定时间时变输出编队-包容跟踪问题。考虑了相邻智能体之间的协作通信和对抗通信。首先，提出了二分固定时间补偿器来估计领导者状态的凸包。与现有技术不同，所提出的补偿器具有以下三个亮点：1）它是连续的，不涉及符号函数，因此可以避免抖振现象；2）其估计可以在固定时间内完成；3）相邻智能体之间的通信不仅可以是协作的，也可以是对抗的。需要注意的是，所提出的补偿器依赖于网络拓扑的全局信息。为了解决这个问题，进一步提出了完全分布式自适应二分固定时间补偿器。它不仅可以估计领导者状态的凸包，还可以估计领导者的系统矩阵。基于所提出的补偿器，进而设计了分布式控制器，使得二分固定时间时变输出编队-包容跟踪能够在固定时间内实现。最后，给出了两个例子来说明主要理论结果的可行性。
