Event-based Control And State Estimation For The Networked Discrete-time Singular Systems

The singular system is a more general dynamic system which can describe the actual system better than the traditional state space system.The networked control system is a new class of control system,which appears with the rapid development of network communication technology,computer technology and automatic control technology,and has been widely concerned because of its important practical application background.The networked singular system is a networked control system in which the controlled plant is a singular system,the study of networked singular system has important theoretical significance and application value.This dissertation will study the analysis and synthesis of networked singular systems based on event-driven scheme.The network control,state estimation and filter design of singular systems are discussed.According to the characteristics of networked control systems,several new event-triggered control strategies are designed and improved,which are applied to the control of various singular systems.New and more general system models are designed for singular systems with mixed delay,data packet loss or sensor saturation.Based on the techniques of Jensen’s inequality,finite-sum inequality and de-lay partitioning technology,the new admissible criteria for singular systems are derived and based on the technique of linear matrix inequality,a co-design method of the con-troller gain,state estimator gain,filter parameters,and event-driven control strategy is presented.The main content of this dissertation are summarized as follows:(1)The event-driven control problem of networked discrete-time singular systems is studied.A new criterion to guarantee the admissibility of networked discrete-time singular systems is proposed.The event-driven H_∞control problem for a class of discrete-time nonlinear singular systems is further investigated,a delay-distribution-dependent criterion is derived which guarantees the closed-loop networked discrete-time nonlinear singular system is admissible with a certain H_∞performance index.The event-driven control problem of singular systems with networked attacks is preliminarily analyzed.(2)The state estimation problem for a class of discrete singular systems with mixed delay and sensor saturation is considered.In order to save network bandwidth and improve the utilization efficiency of network resources,a distributed event triggering scheme is constructed.A comprehensive model of the state estimation error system is established,and by using Lyapunov theory and LMI technology,sufficient conditions are derived making the state estimation error system regular,causal and stable,with a given performance index of H_∞.A co-design method of event trigger strategy parameters and estimator gain matrix is presented.(3)The H_∞filtering design problem is investigated for a class of networked discrete-time nonlinear singular systems with the discrete and distributed delays and random packet dropouts.A more general and more precise discrete-time nonlinear singu-lar networked system model is proposed by using the independent Bernoulli distributed random variables.Sufficient conditions are presented to guarantee the closed-loop filter error system is admissible with a certain H_∞performance index,and the H_∞filtering design method is provided.(4)The network-based filtering problem for discrete-time singular Markovian jump systems with an event-triggered scheme is studied.A mode-dependent event-triggered scheme is proposed,and the singular Markovian jump system with time-delay is es-tablished.By using the delay partitioning technology,new criteria are derived which guarantee the filtering error system to be stochastically admissible with a certain H_∞performance index.The variable order H_∞filters with mode-independent characteri-zation are designed in a unified framework.The problem of event-triggered H_∞/passive filtering for singular Markovian jump systems is further studied.By introducing switching variables,mode-dependent and mode-independent filters are designed in a unified form.According to Lyapunov theory and finite-sum inequality technology,sufficient conditions are obtained for the system to be stochastically admissible and has a mixed of H_∞/passive performance index,and a collaborative design method for filter and event-driven control strategy parameters is given.The proposed method can deal with the mode-dependent or mode-independent filter design problems for SMJSs including H_∞,passive and mixed filter(i.e.,mixed H_∞and passive filter)in a unified form.(5)Based on the graph theory and singular systems theory,the event-triggered guaranteed-cost consensus problem of singular multi-agent systems with switching topolo-gies is investigated.To save the limited network bandwidth of systems,a novel event-triggered control scheme is proposed by considering the consensus protocol.Sufficient conditions of guaranteed-cost consensus control for the discrete-time singular multi-agent are derived and given in the form of the linear matrix inequalities.Finally,the cost func-tion is determined,and the explicit expression of consensus functions is given.Finally,the conclusion and the future research topic and direction are given.