Analysis And Synthesis Of Networked Control System Based On Markov Process

With the Big Data,Cloud Computing,Industrie4.0 and other new concepts putting forward,the control system is increasingly relying on the network and the traditional control has been unable to meet the needs such as Informatization,Intelligence and Webification.Therefore,the networked control system(NCS)is the future of control system.Only by strengthening the integration with communication technology and computer technology can the control technologies have a greater breakthrough.The intervention of network makes the NCS a random system.The completeness and certainty of the data,the regularity of the arrival time and the causality of the received data are no longer guaranteed.Besides,the current delay is only dependent on the previous delay in NCS,which is similar to the non-aftereffect property of Markov chain.Therefore,in this thesis,we study the phenomenon of delay and missing measurement in the networked control system mainly based on the continuous Markov jump system.Markov process is used to model the delay and measurement loss,and the stability analysis is carried out by the Lyapunov theory.The linear matrix inequality technique(LMI)is used to solve the controller.The main work of this research paper is as follows:(1)The feedback control problem for a class of uncertain networked control systems with random input time delay is concerned.The random delay is modeled as a Markov process while the uncertainties are time varying but norm-bounded.By employing a new stochastic lyapunov function and transferring the original random system into a singular form,which has the same mean square stability,sufficient stable and stabilizable conditions for the underlying systems are derived via a Lyapunov-Krasovskii functional approach.(2)The stabilization and H? control problem for a class of continuous-time networked control system(NCS)with random time-delay modeled as a Markov process with incomplete transition rates is concerned.The research motivation is to achieve numerical testable stabilization criteria.First,the stochastic stabilization of the underlying NCS with jumping parameters is investigated by transferring the original system into a singular one,which has equivalent mean square stability.Then,based on the results of stabilization,the H? disturbance attenuation is further considered.The sufficient conditions are presented in the form of standard linear matrix inequalities(LMIs),which is rarely reported in the existing references.(3)Both missing measurement and time-delay are considered for the analysis and synthesis of networked control system(NCS).Markovian jump systems characterized by a doubly stochastic process are applied to model the NCS when missing measurement and time-delay are modeled as Markovian jumping parameters.A mode-dependent control law is proposed to deny the influence caused by the missing measurement and time delay.Sufficient conditions of the stabilization for the underlying NCS is developed in the form of standard linear matrix inequalities(LMIs)can be obtained by constructing a Lyapunov-Razumikhin function,and further,a feedback controller can be obtained by solving the LMIs.