Impulsive Consensus Of Multi-Agent Systems With Semi-Markov Switching Topologies

Cooperative control of multi-agent systems(MASs)has undoubtedly been a popular research problem in the study of systems and control in the last decade.Consensus has been considered as a fundamental concept in the study of distributed control and cooperative control in the MASs,referring to the design of a distributed control protocol that enables a group of autonomous intelligences to reach an agreement in a sense,widely used in such areas as distributed optimization,spacecraft coordination,and robotics cooperation.In many applications of MASs,each agent is usually equipped with limited energy storage and communication resources.Therefore,how to reduce the occupation of network resources is of significance in practice.Compared with continuous control method,impulsive control has received a great deal of attention at present.Because it does not require real-time communication of MASs and only needs to send information at certain specific instant,while it does not influence the system at other moments,and the control cost is low,which can achieve the stability and synchronization of MASs.In addition,various uncertainties,such as switching of communication networks and transmission delays,usually exist in actual MASs,leading to local information exchange among agents that may be affected by stochastic topology switching.To accurately characterize such stochastic topology switching,we assumes that the switching between topologies satisfies a semi-Markov process.Therefore,this thesis focuses on the pulse consistency problem of MASs under semi-Markovian switching topology,aiming to propose more flexible,effective and realistic impulsive control algorithms.The main innovations of this thesis include the following aspects.First,the thesis investigates the impulsive control problem of MASs under semi-Markovian switching topology.Compared with the consistent results of multi-agent systems with continuous control methods,the distributed impulsive control can greatly reduce the information resources among agents,and the control cost is lower,which can be better applied to practice.The thesis introduces the one-sided Lipschitz condition in the process of controller design,making the conclusions obtained in this thesis less conservative than those in the existing theses.Second,the thesis explores the asynchronous impulsive control for MASs in semi-Markov switching topology.Compared with synchronous impulsive protocols for MASs,asynchronous impulsive control does not necessarily subject all the intelligences in the MASs to be controlled simultaneously,which not only reduces the amount of information transmission but also improves the efficiency and reduces the control cost at the same time.Third,the thesis discusses the intermittent impulsive control of MASs under semi-Markov switching topology.Compared with the existing researches,the theorem in this thesis is more flexible in practical applications by remove the upper limit restriction on the impulsive interval in the adjacent pulse interval.