Event-based consensus with relative state measurements
Yilin Wang, Shuanghe Yu, Xiaojun Long, Lina Jin
Abstract
Yilin Wang, Shuanghe Yu, Xiaojun Long, Lina Jin
Abstract
In this paper, the event-triggered consensus problem of multi-agent systems (MAS) is considered when only the relative state measurements are available. Some beneficial spectral properties of edge Laplacian matrix are first investigated for a connected undirected graph topology, then the event-triggered consensus problems are respectively formulated for the double integrator MAS under the edge dynamics. In particular, the distributed events depending on the edge state measurement errors with respect to the last triggered values are properly defined to achieve global asymptotic consensus in the Lyapunov sense. Furthermore, the positive lower bounds for inter-event times are derived to guarantee the absence of Zeno behavior. Compared with the existing works mainly on agent dynamics, less communication efforts and controller updates are needed for similar performance and the estimation of inter-event time is less conservative. Finally, simulation results are provided to demonstrate the effectiveness of the proposed approach.
OpenAlex reports 1 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In this paper, the event-triggered consensus problem of multi-agent systems (MAS) is considered when only the relative state measurements are available. Some beneficial spectral properties of edge Laplacian matrix are first investigated for a connected undirected graph topology, then the event-triggered consensus problems are respectively formulated for the double integrator MAS under the edge dynamics. In particular, the distributed events depending on the edge state measurement errors with respect to the last triggered values are properly defined to achieve global asymptotic consensus in the Lyapunov sense. Furthermore, the positive lower bounds for inter-event times are derived to guarantee the absence of Zeno behavior. Compared with the existing works mainly on agent dynamics, less communication efforts and controller updates are needed for similar performance and the estimation of inter-event time is less conservative. Finally, simulation results are provided to demonstrate the effectiveness of the proposed approach.
Key concepts: Double integrator, Laplacian matrix, Consensus, Computer science, Multi-agent system, Control theory (sociology), State (computer science), Graph