Research On H_∞ Robust Control Strategies For Networked Systems With Access Constraints

With the growing automation and intelligent requirements in the engineering field,traditional point-to-point control methods cannot meet the needs of modern industrial control.In such a situation,Networked control Systems(NCS)came into being.NCS have been widely used in power systems,intelligent transportation,telemedicine and data acquisition due to their advantages of low cost,simple installation,high reliability and easy later maintenance.In addition,the networked control mode can effectively deal with some special environment brought by manual inconvenience,which is not realized by the traditional control method,but also lays the foundation for the industrial control system of safe production.However,while the communication network brings a lot of convenience,it also causes a lot of problems,such as limited network bandwidth,time-sharing multiplexing of channels,etc.,which brings great challenges to the information interaction and real-time control of networked systems.Therefore,how to schedule the limited network communication channels reasonably,while taking into account the quality of network service and the quality of system control,has become an important problem in the application of NCS.In order to solve these problems,this thesis conducts in-depth research on the control strategy of networked systems with access constraints.The main work is as follows:Firstly,the design of H_∞ robust control strategy is investigated for networked systems with both access constraints and data packet loss.Bernoulli stochastic process is used to describe the packet-loss process of the system,and a discrete-time Markov jump system model is established according to the random access mechanism of medium access constraints.At the same time,considering that the state information of most controlled plants cannot be obtained directly,an observer-based H_∞robust control strategy is proposed.On this basis,by means of Lyapunov stability theory and linear matrix inequality technology,sufficient conditions are derived for the stochastic stability of closed-loop system,and the design of H_∞robust control strategy is presented,which ensures the good control performance of the system.Secondly,the H_∞ state estimation problem is discussed for nonlinear networked control systems with access constraints.Considering the nonlinear characteristics of most systems in networked control practice,nonlinear variables satisfying the global Lipschitz condition are introduced to describe this characteristics,and networked system are established as error equivalent systems with nonlinear characteristics.By means of Lyapunov stability theory and linear matrix inequality technology,sufficient conditions for the stability of error equivalent system are derived.On this basis,the H_∞state estimation algorithm is designed and the effective state estimation of the system is implemented.Thirdly,the asynchronous control problem of dynamic output feedback is investigated for discrete time-delay networked systems with access constraints and data loss.By introducing a joint Markov chain,networked systems are modeled as stochastic systems subject to Markov jump characteristics,and an asynchronous control strategy based on dynamic output feedback is proposed.On this basis,by using the processing method of time-delay systems,the Lyapunov-Krasovskii functional is constructed,and sufficient conditions are derived for the stochastic stability of closed-loop systems with H_∞performance.At the same time,a more general matrix decoupling method is provided by using singular value decomposition technology,and gain matrices of dynamic output feedback controller are solved,which achieves the stabilization control of networked systems.Finally,the research work of this thesis is summarized,and the future work is prospected on the basis of the existing research results.