Fully Distributed Event-Triggered Coordinated Control For Multi-Agent Systems

A multi-agent system is composed of a group of autonomous agents that are connected through a communication network,perceive the state of the external environment,take actions,and adjust their decisions through feedback from the environment to achieve global or local goals.This system has been widely applied in areas such as smart grids,drone collaboration,and intelligent logistics,and is of great significance for promoting high-quality economic development and improving the national strategic layout in China.In multi-agent systems,traditional periodic control can waste computing and device energy resources due to the limited availability of energy and communication resources.Additionally,it can also cause network congestion.Furthermore,due to the spatial distribution of agents and limited sensing capabilities of sensors,coordinated control protocols in multi-agent systems can only rely on the local state or output information of each agent and its neighbors.Therefore,it is necessary to design a fully distributed coordinated control scheme based on local information and interaction with neighboring subsystems to achieve the coordination control objectives of the multi-agent system while effectively saving communication resources with event-triggered strategy.In this thesis,we propose some distributed control algorithms based on some event-triggered strategies for three types of multi-agent systems to address different coordinated control objectives,namely,consensus control problems,formation control problems and containment control problems.The main contents of this thesis are as follows:(1)The problem of event-triggered fully distributed consensus control for uncertain linear multi-agent systems is investigated.A fully distributed reduced-order observer is designed using relative output information.Moreover,an adaptive event-triggered strategy is proposed to eliminate the impact of asynchronous triggering time with different event-triggered mechanisms.Through above methods,based on the output regulation theory and the internal model principle,a distributed control protocol is designed,and the stability analysis of the closed-loop system is carried out using the Lyapunov direct method.Finally,numerical simulations and physical experiments are conducted to verify the effectiveness of the proposed control algorithm.(2)The problem of event-triggered fully distributed formation control is addressed for nonlinear heterogeneous multi-agent systems subject to actuator and communication faults.A fully distributed event-triggered state observer is proposed to reduce communication overhead and computational burden and effectively avoid the interference caused by communication faults.Then,an adaptive control law is designed based on the backstepping method,and a Nussbaum function is introduced to solve the problem of unknown control direction caused by actuator faults.The stability analysis of the closed-loop system is carried out using the Lyapunov direct method.Finally,numerical simulations are conducted to verify the effectiveness of the proposed control algorithm.(3)The problem of event-triggered distributed containment control is studied for networked Euler-Lagrange systems under denial-of-service(Do S)attacks.A periodic event-triggered state observer is proposed to reduce the number of triggering instances,save communication and computation resources,and effectively mitigate the impact of denial-of-service attacks on the system.Then,a fuzzy logic system is designed to approximate the strongly nonlinear term in the Euler-Lagrange system,and an adaptive control protocol is proposed by combining the designed state observer.The stability analysis of the closed-loop system is conducted using the Lyapunov direct method.Finally,the effectiveness of the designed control algorithm is verified through numerical simulations and visualization software.