Analysis And Synthesis For Networked Systems Based On Event-Triggered Mechanism

With the gradual maturity of network communication technology,the application scope of networked systems is expanding and has been extended to various fields of industrial production.The networked system is a distributed control system which transmits information or signal through the shared digital network.It replaces the traditional point-to-point connection mode,which realizes remote operation and resource sharing.It has the advantages of high reliability,convenient installation and maintenance and flexible communication architecture.However,Networked systems are also facing challenges.Firstly,due to the limited network bandwidth,data transmission in the channel will produce problems such as network-induced delay,packet loss and quantization error,which will lead to poor control performance and even instability of the system;Secondly,considering the system components（sensors,actuators,etc.）aging or varying degrees of failure,the system performance will be affected;Thirdly,the modeling of some systems is more complex due to the unmeasurable states,the nonlinearity of the systems,the time-varying and uncertain parameters.The event-triggered mechanism（ETM）reduces unnecessary transmission by transmitting data on demand,which effectively reduces the impact of network bandwidth constraints while ensuring system performance,and is gradually replacing the traditional time-triggered mechanism.Therefore,integrating the above aspects,it is of great significance to introduce ETM into the modeling,analysis and design of networked systems.In this paper,aiming at the H_∞andL₂-L_∞performance index,the analysis and synthesis for networked systems based on ETM is gradually studied.The main research contents are as follows:Firstly,since the element components may fail in the actual system operation,considering the influence of sensor probabilistic faults and limited network resources on system performance,the H_∞filtering problem of distributed-delay systems under ETM is studied.Different fault states are represented by a set of random variables,and the ETM is introduced to remodel the system.The selected delay-dependent Lyapunov-Krasovskii functional is constructed by using the infinitesimal operator.According to the Lyapunov stability theory,the design method of robust H_∞filter with low conservatism is obtained.Secondly,in practical engineering,the communication resources are limited,and the controlled objects also have nonlinearity.Therefore,the T-S fuzzy model commonly used in nonlinear systems is adopted to comprehensively consider the distribution delays,time-varying delays and time-delay information caused by the ETM.By using the Lyapunov stability theory and the integral inequality method,the sufficient conditions for the existence of robust H_∞filter are obtained.Furthermore,the coupling parameters between the system and Lyapunov-Krasovskii functional are uncoupled by the method of full transformation and variable substitution,and the problem of solving the filter parameters is converted into a convex optimization problem.Finally,in the simulation process,a numerical example is given to verify the rationality and effectiveness of the proposed theoretical method.Thirdly,the ETM is used to improve the transmission efficiency of information and reduce the occupation of the network bandwidth.By introducing norm-bounded uncertainty to describe the uncertainty of the system model and the controller in the form of parameter perturbation,a class of distributed-delay systems non-fragileL₂-L_∞control problem under ETM is proposed.Based on the Lyapunov stability theory and related lemmas,the auxiliary analysis is carried out.Considering the combination of time-varying delay and ETM,the closed-loop system under ETM is deduced to be asymptotically stable and meet the L₂-L_∞performance,and the solution method of the non-fragile robust controller parameters and the ETM related parameters is obtained.The simulation results show that the non-fragile controller has better convergence effect than the conventional controller.Finally,aiming at the non-ideal network conditions with transmission delays and quantization errors,the quantization errors are described as uncertainties by means of the sector-bounded method,and two logarithmic quantizers are used to reduce the influence of the quantization error to improve the control accuracy.Using the Parallel Distributed Compensation（PDC）algorithm and combining with the T-S modeling idea,the global fuzzy model of the closed-loop system under actuator failure is obtained.On this basis,for the case that the output signal is energy bounded and peak bounded,respectively,a design method of H_∞andL₂-L_∞fault-tolerant controller based on LMIs is proposed to make the system asymptotically stable.