Distributed Adaptive Observer-Based Fault-Tolerant Output Consensus for Heterogeneous MASs
研究了异构多智能体系统在领导者系统矩阵未知、通信链路和执行器故障下的输出一致性问题，设计了分布式自适应观测器和容错控制器，使每个跟随者的输出渐近收敛到领导者的输出。
The output consensus or tracking problem for heterogeneous multiagent systems (MASs) under jointly observable condition (JOC) is investigated in the presence of unknown system matrix of the leader and unknown faults on communication link and actuator. Unlike the recent works, where neighbors of the leader have access to the state information of the leader, only part output information of the leader can be measured by its neighbors under JOC in this article. In addition, the system matrix of the leader only is known to neighbors of the leader. The problem of unknown leader’s state and unknown leader’s system matrix, unknown communication link faults (CLFs) is transformed into constructing the fully distributed state and system matrix observers by virtue of adaptive control technique and JOC. In light of the developed distributed observer, a distributed adaptive fault-tolerant controller against actuator faults is designed. Here, compared with the existing ideas, actuator faults of an agent will not affect the dynamics of its neighbors over the communication graph in this article. It is shown that the output of each follower asymptotically converges to the output of the leader. Finally, the developed control scheme is illustrated by a simulation example.