Consensus of Lur'e Multi-Agent Systems With Directed Switching Topology

We study the consensus tracking problems of Lur'e multi-agent systems (MASs) with directed switching topology and a leader that is subject to unknown inputs. We firstly use the pure relative outputs to design a distributed unknown input observer (UIO) that can track the relative states exactly. Then we use this UIO to design a discontinuous controller which contains a consensus part and a discontinuous part, where the impacts of the leader's unknown inputs can be completely removed by the discontinuous part. Furthermore, by using the average dwell time (ADT) method, we show that asymptotical consensus can be achieved if the ADT is larger than a positive threshold. Secondly, we design a continuous controller under which the consensus tracking error is uniformly ultimately bounded (UUB). Finally, we perform a simulation on the multiple Chua's circuits to validate the theoretical result.

Automated summary

In this study, the consensus tracking problems of Lur'e multi-agent systems are investigated with directed switching topology and an unknown input leader. A distributed unknown input observer is designed to track the relative states exactly. A discontinuous controller is then proposed using this observer to remove the impacts of the leader's unknown inputs. It is shown that the system can achieve asymptotical consensus by using the average dwell time method. A continuous controller is also designed to ensure uniformly ultimately bounded consensus tracking error. The theoretical results are validated through simulations on multiple Chua's circuits.