Research On Several Event-triggered Control Mechanisms For Networked Systems

Due to the issues of limited network resources and bandwidth in embedded microprocessors,the event-triggered mechanism(ETM)for networked system emerges at the historic moment.With the basic principle that the controlled system can maintain desired system performance,ETM will reduce the number of data transmission and controller update,which not only saves network resources but also reduces communication consumption.Thus,with ETM,the stabilization problems of network control systems(NCSs)and the cooperative control of networked multi-agent systems(MASs)have drawn wide attention,and have important applications in many aspects.Available ETMs include sampleddata-based ETM,combinational-measurements-based ETM,model-based ETM,integral-type ETM,dynamic ETM and so on.However,some issues such as frequent triggering,continuous communication among agents while event detection,continuous controller update may occur in available works.In consideration of network environment subject to abrupt changes,delays,uncertainties,disturbances and so on,this paper proposes new designs on ETMs for several kinds of networked systems.Moreover,in light of tools such as algebraic graph theory,Lyapunov stability theory,linear matrix inequality,and ordinary differential equation,together with event-triggered control,and sliding mode control method,stability problems for networked systems have been analyzed detailedly.The main contents of this paper are as follows:●Sampled-data-based event-triggered control problem for networked semiMarkovian jump systems(SMJSs)with randomly occurring factors is studied.At first,stabilization problem for networked SMJSs with transmission delay,randomly occurring uncertainties and nonlinearities under ETM is investigated.By using stochastic analysis and Lyapunov method,sufficient conditions are proposed to ensure the stochastic stability for closed-loop SMJSs in terms of linear matrix inequalities(LMIs).Then an improved sampled-data-based ETM is introduced and H∞ stabilization problem for networked SMJSs with randomly occurring uncertainties and transmission delay is investigated.Similarly,LMIs conditions are derived to guarantee the H∞ stabilization of closed-loop SMJSs with a prescribed performance index γ.Compared with traditional sampled-data-based ETM,data transmission and resources consumption will be further reduced.●Under the modified integral-type ETM,the consensus and leader-following consensus problem of networked MASs are investigated.With this ETM,the continuous communication among agents can be avoided during event detecting.Since the state of agent may not be measurable,state-based and observer-based event-triggered protocols are designed respectively.Lyapunov analysis method and Barbalat’s lemma are used to obtain the conditions for achieving consensus of MASs.Moreover,it is proved by theoretical analysis that no agent will exhibit Zeno phenomenon.●The problem of scaled consensus control based on ETM for first-order MASs under undirected topology is considered firstly,but the event detection process requires continuous communication among agents.Then,under directed topologies,practical scaled consensus problem for nonlinear MASs with input delay is discussed via a new integral-type ETM.With this ETM,the requirement of continuous communication among agents can be removed in event detection,and interevent time can be prolonged.With trajectory analysis method,sufficient conditions to realize practical scaled consensus are given and the upper bound of input delay is obtained.It is theoretically proved that each agent can exclude Zeno behavior.●The scaled consensus problem of general linear MASs with double ETMs is studied.Firstly,a novel dynamic ETM is designed for communication by using model-based estimation technique such that the continuous communication among agents is not required for both event detection process and event-triggered control protocol.In addition,a decentralized ETM is designed for controller update such that the controllers don’t need to be updated continuously.Then by constructing appropriate Lyapunov function,conditions for guaranteeing scaled consensus of MASs are obtained.●A new design for event-triggered sliding mode control mechanism is proposed,and the tracking consensus problem of second-order MASs subject to external disturbances is investigated.Influenced by sliding mode control method and ETM,the controlled system has the advantages of fast response,good transient performance and robustness when dealing with uncertainties.Meanwhile,resource consumption and network usage can be greatly reduced and Zeno behavior of event-triggered communication will not occur.Distributed event-triggering based sliding mode control protocols are proposed by using equivalent control method,and implementation of ETMs don’t require continuous communication among follower agents.In order to increase the freedom of parameter selection,heterogeneous event-triggered conditions are designed,and sufficient conditions with less conservativeness are obtained to ensure the accessibility of the sliding mode surfaces.