Distributed Dynamic Event-triggered Consensus Control For Linear Multi-agent Systems

Distributed event triggered control has been widely concerned by a large number of scholars in the unified control of multi-agent systems because of its advantages such as high resource sharing efficiency,easy implementation of modularization,strong robust-ness,fast speed and effective reduction of network transmission.However,how to design an appropriate distributed event triggering control mechanism to reduce a large amount of information transmission between agents and agents,as well as between each agent sensor and controller,is a big challenge.Therefore,the research on distributed event trig-gering consistency control of multi-agent system has important theoretical significance and scientific value.Despite the fact that numerous control algorithms have been developed for dis-tributed event-triggered consensus control of multi-agent system,there are the following issues to be noticed.(i)Event-triggered functions are static in a large number of existing distributed event-triggered mechanisms due to the invariance of triggering parameters in the implementation process,which greatly limits the implementation efficiency of event-triggered mechanisms and reduces the flexibility of event-triggered control.(ii)A large number of existing distributed event-triggered consensus control mechanisms are mainly applicable to first-order or second-order multi-agent system,and there is no clear con-clusion on the relationship between agents’communication topology and event-triggered architecture.(iii)Existing research on distributed event-triggered consensus control rarely involves situations such as channel fadings,input constraints,and unmeasurable continu-ous states.In order to solve the above issues and based on the idea that the single dynamic event-triggered control mechanism can further improve the utilization of network bandwidth resources and control flexibility,this thesis mainly focuses on the distributed dynamic event-triggered consensus control of linear multi-agent systems with fixed topology under different situations.The main contents are as follows:To address the distributed dynamic event-triggered consensus control problem of lin-ear continuous time-invariant multi-agent systems,a distributed dynamic event-triggered control protocol based on state estimators’information is designed.Then,the progres-siveness of the designed control mechanism in saving network bandwidth resources and the possibility of excluding Zeno phenomenon(the phenomenon of infinite triggering in a limited time)are analyzed.In addition,the quantitative relationship between the communication topology of multi-agent systems and the optimal convergence rate is es-tablished.Finally,the effectiveness and superiority of the designed distributed dynamic event-triggered control mechanism are verified through simulation.Aiming at the distributed dynamic event-triggered consensus control problem of lin-ear continuous time-invariant multi-agent systems with control input constraints and un-available continuous observation of system states,a distributed dynamic event-triggered control protocol is designed based on the research results in the previous part by using predictive state information.This protocol reduces the impact of input constraints on system control,removes the limitation in existing distributed event-triggered mechanisms that trigger functions require continuous observation of system states.At the same time,a mixed Lyapunov function is constructed based on the Lyapunov stability theory,and suf-ficient conditions for the consensus of the addressed system are analyzed.The possibility of Zeno phenomenon is excluded,the quantitative relationship among triggering param-eters,communication network topology,and system convergence rate is discussed,and a distributed dynamic event-triggered control algorithm is designed.Finally,simulation verifies the effectiveness and correctness of the proposed distributed dynamic predictive event-triggered control mechanism and its corresponding controller.For the distributed dynamic event-triggered consensus control problem of linear discrete time-invariant multi-agent systems with channel fadings,a distributed dynamic event-triggered control protocol in discrete time domain is designed on the premise that channel fading only occurs between each agent sensor and controller.The designed pro-tocol can not only offset the negative impact of channel fadings on event-triggered control of multi-agent systems,but also effectively improve the utilization of network bandwidth resources.Under the effect of the protocol,the whole closed-loop system model is ob-tained.Then,based on Lyapunov stability analysis,sufficient conditions for the existence of the system consistent controller in the form of linear matrix inequality are obtained.Subsequently,the design method of discrete distributed dynamic event-triggered control protocol is given.Finally,simulation verifies the effectiveness and progressiveness of the designed control protocol.For the finite-horizon distributed dynamic event-triggered H_∞consensus control problem of linear discrete time-varying multi-agent systems with energy-bounded exter-nal disturbances,firstly,the motivation for analyzing finite-horizon consensus problems is given.Subsequently,a discrete distributed dynamic event-triggered H_∞control pro-tocol is designed,and an augmented time-varying multi-agent consensus error system is established.Meanwhile,based on Lyapunov stability analysis and reverse Riccati differ-ence equation method,sufficient conditions for the system to achieve the finite-horizon H_∞consensus is obtained.As a result,a series of virtual error information is obtained to characterize the characteristics of time-varying systems.Then,the relationship between virtual errors and triggering errors is analyzed,and the original discrete time-varying system is transformed into a variable parameter system with time-varying interference information.Subsequently,the solution conditions for controller gain are given,and cor-responding control algorithms are designed.Finally,simulation verifies the effectiveness and superiority of the designed discrete distributed dynamic event-triggered control mech-anism.