Design And Performance Analysis Of Closed-loop Systems In Network Communication Enviroment

Nowadays,the Internet has become an indispensable tool in daily life.With the increase of Internet penetration,the use of network to control industrial equipment has become an important trend.The networked control system is composed of the controlled object and network connection.Compared with the traditional control systems,networked control systems have many advantages such as: save the cost of laying lines,digital information transmission,hierarchical control structure,intelligent control systems,etc.In addition,the existence of limited network communication problems: network induction delay,packets dropout,packet disorder,multi-packet transmission,etc.For the problems caused by the networks,this paper analyzes and studies networked control systems under limited communication.The main research results of this paper are as follows:Firstly,the networked prediction of second-order nonlinear systems with quantization and an extended state observer is studied.The extended state observer is not only used to deal with nonlinear term,but also generates prediction states which are used to deal with induced-delay.Two arbitrary regional quantizers are used to effectively cope with forward and feedback channel signals.Through the "zoom" strategy,the system stability of the nonlinear networked control system with quantization is proved.Secondly,the stability of event-driven networked control system under the influence of quantization and Markov packets dropout is studied.A discrete event-monitor is used to monitor the specific event in the networked control system.Furthermore,an arbitrary regional quantizer and Markov packets dropout are considered in the networked control system.On the basis of "zoom" strategy and Lyapunov theory,the mean square stability of the closed-loop system is proved.Finally,the feedback controller is designed to ensure the stability of the networked control system.Thirdly,a nonlinear networked control system based on switching model predictive control strategy is studied.Switching cost functions are considered to improve the performance of continuous time nonlinear networked control system.At the same time,wireless networks exist in two communication channels,and predictive inputs are used to wireless networks exist in two communication channels,and predictive inputs are used to solve the problem of packets dropout.Therefore,the application of the self-triggered strategy reduces the amount of networks burden.Finally,the asymptotic stability of continuous time nonlinear networked control system is guaranteed under the average residence time condition.Finally,asynchronous sampling distributed stochastic model predictive control is proposed for a networked control system with multiple dynamic decoupling subsystems.Each subsystem is influenced by random disturbances and coupled probability constraints.In the networked control system,the delta operator is used to sample subsystems with asynchronous sampling periods.By considering packets dropout and the distribution law of disturbances,the coupled probability constraints are transformed into deterministic constraints of the decision variables.In addition,the iterative feasibility of local and coupled probability constraints and the quadratic stability of networked systems are guaranteed under the proposed algorithm,respectively.