Finite-time Control Of Networked Control Systems With Stochastic Time Delay

With the continuous advancement of technological reforms,the requirements of remote control,resource sharing and complex large systems have been put forward.Traditional control methods have been difficult to cope with.Strengthening the integration with communication,computer and other fields to realize the networked control systems is the current development trend.However,the intervention of network brings new problems to the control system,such as communication limitation,network induced delay,packet loss,etc.This thesis mainly studies the finite-time control problem of networked control systems with random time delays.For problem of time delay,a finite-time stability criterion based on linear matrix inequality is given by using Lyapunov equation.For problem of packet loss,the idea of time offset is used to convert packet loss into time delay,thus the problem of packet loss is solved by processing time delay.For problem of packet loss,the idea of time offset is used to convert packet loss into time delay,thus the problem of packet loss is solved by processing the time delay.The main research contents are as follows:(1)The problem of finite time delay-dependent state feedback stabilization for a class of networked control systems with a bounded external disturbance is investigated.The sensorto-controller time delay and the controller-to-actuator time delay are modeled as two Markov chains,and the time delay is introduced in the design of the controller.The time delay is converted into the system state delay,and the change rule of the time delay is used to describe the change rule of the system.An improved finite-time delay-dependent stabilization criterion is given in the form of linear matrix inequalities.(2)For a class of networked control systems with bilateral random time delay,the finitetime stabilizing controller based on observer is designed.The basic idea is to transform time delay into a random sequence by using Markov chains and establish time delay compensation state observer.Then a observer-based delay correlation controller is designed.Considering the complexity of the network environment,this thesis studies the situation where the partial transfer probability is unknown.By solving linear matrix inequalities,a time delay-dependent finite-time state feedback controller is obtained.(3)We consider the problem of finite-time state feedback stabilization for a class of networked control singular systems with bilateral time delay and bilateral packet loss.The basic idea is to transform packet loss into time delay by introducing a time offset and transform timedelay into state delay.It gives the probability formula of continuous packet loss based on the full probability formula,and uses the Markov chain to characterize the variation law of time-delay.The improved finite time delay correlation stability criterion is given in the form of linear matrix inequality,and the effectiveness of the proposed method is verified by numerical simulation.