Event-triggered State Estimation And Control Of The Singularly Perturbed Networked Systems

With the development of networked systems,the state estimation and control of systems with multi-time-scale property(known as the singularly perturbed system-s)in the networked environment,has become a new hotspot in recent years.With the purpose of dealing with multi-time-scale phenomenon,the limited communication resources and data transmission collision,as well as to cope with network induced phenomena such as distributed time-delays and sensor saturations,in this paper,we have investigated the analysis and synthesis problems of the singularly perturbed net-work systems,including the singularly perturbed networked control system,singularly perturbed wireless sensor networks and singularly perturbed complex networks,based on the event-triggered schemes.It can be summarized as follows:(1)The state estimation problem for a class of singularly perturbed networked systems is investigated with distributed time-delays.In order to reduce the signal transmission frequency to save the constrained communication resources,especially network bandwidth,we have designed both static and dynamic event-triggered scheme to arrange the datapacket transmission from sensors to the desired state estimators.By designing a novel?-dependent Lyapunov-Krasovskii functional,in the case that?does not exceed a predefined upper bound,a design algorithm is developed for the desired state estimator ensuring that the error dynamics is exponentially mean-square ultimately bounded.In this case,by solving certain matrix inequalities,the estimator gain is characterized without needing to know the exact?(as long as it stays below the given upper bound).Moreover,the ultimate bound of the error dynamics is estimated.Finally,simulation results are given to confirm the validity and advantages of the proposed design scheme of the state estimator.(2)The state estimation problem is discussed for a class of nonlinear singularly perturbed wireless sensor networks.In order to estimate the state of the nonlinear singularly perturbed systems in a more complicated wireless sensor network environ-ment,a set of distributed H_?state estimators are designed based on the dynamic event-triggered mechanism.By using the Kronecker product and elementary matrix transformation,the augmented state estimation error dynamics is transformed into a standard singular perturbation form,and a less-conservative?-dependent Lyapunov function is designed.Taking the sparsity of large-scale wireless sensor network n-ode connections into consideration,we have ensured the correspondence between the desired state estimator and the topology structure.(3)The H_? control problem is investigated for a class of discrete-time singu-larly perturbed systems with time delays and sensor saturations.In order to save the network bandwidth and improve the communication reliability during the sig-nal transmission,a dynamic event-triggered mechanism is put forward to arrange the data-packets transmission from the saturated sensors to the proposed controller.By constructing a novel Lyapunov-Krasovskii functional,the dynamics of the DTSPSs combined with the dynamic event-triggered mechanism are analyzed,and an output feedback controller design scheme is then developed.By resorting to the singular value decomposition lemma,the desired controller gain is then parameterized in light of the feasible solutions of a set of matrix inequalities.(4)The synchronization control problem is studied for the singularly perturbed complex networks.Considering the coupling structure of the exiting singularly per-turbed complex network systems,a new nonlinear discrete-time singularly perturbed complex network model is constructed via the slow sampling method.In order to re-duce the signal transmission between the network nodes,alleviate the pressure of sig-nal transmission and reduce the signal collision,a distributed dynamic event-triggering mechanism is designed.By using Kronecker product and the matrix elementary trans-formation,the augmented synchronization error system is transformed into a standard singular perturbation form,and a less conservative Lyapunov-Krasovskii functional is established.Finally,a set of synchronization controllers are obtained based on the dynamic event-triggered scheme.Finally,a simulation example is given to show the correctness of the design methods.