Research On Robust Filtering Method For Networked Time-Delay Systems

The research problem of networked control systems has always been a hot issue in the control disciplines.When information is transmitted through the network,data transmission delays and packet loss are inevitable due to network bearer capacity and bandwidth limitations.The existence of these problems makes the networked control systems essentially stochastic systems,so the analysis and design of the networked control systems are moving towards more complicated direction.Since the theory of filtering has been proposed,filtering has been the basis and focus of research in the field of automatic control and signal processing.It has been widely used in the fields of aircraft navigation and radar tracking.Therefore,the research on the filtering problem of networked time-delay systems is still a problem worthy of further investigation.Firstly,the filtering problem of a class of uncertain distributed time-delay systems is studied.Considering the time-varying delays,distributed delays and uncertainties of systems models,constructing a non-fragile filters.Through the Lyapunov stability theory and the norm-bounded matrix elimination method,the sufficient conditions for the existence of L1 performance and H? performance filters are obtained respectively.The numerical simulation is used to prove that the non-fragile filter has good state estimation ability.Secondly,the chapters have studied the L1 filtering and L1 dynamic output feedback control for networked control systems.On the one hand,in the networked control systems,two independent Bernoulli random variables are introduced to model the random delay and missing measurement,and construct non-fragile filter and non-fragile controller.Using Lyapunov function,integral inequality and linear matrix inequality technique establishes sufficient conditions for non-fragile L1 filters and dynamic output feedback controllers.On the other hand,considering the time-delay in the systems according to the probability distribution in different intervals,the network control system with time-varying delay and measurement data loss is modeled as a stochastic systems with random parameters,using the same analysis method,the solution method of the filter is obtained when the interference signals are sustained and peak-bounded.Finally,the chapter is concerned with the event-triggered H? fault detection problem for discrete-time networked control systems with unreliable communication links.First of all,an event-triggered mechanism is proposed to determine whether the sampled data is transmitted to the network in order to save network resources.What's more,a Bernoulli process is employed to model the phenomenon for the packet losses.Lastly,considering the above constraints,a networked control system model is established that can simultaneously describe event-triggered strategy,data losses,distributed delay and time-varying delay.The sufficient conditions for the fault detection filter are obtained by Lyapunov stability theory and integral inequality method.Finally,the numerical simulation is used to prove that the fault detection filter under the event-triggered mechanism not only saves networked resources but also effectively detects system fault information.