Research On Predictive Control Of Discrete Linear Networked Control System

In the last few years,with the great improvement of the control technique and the communication network,the strong concern among domestic scholars has been aroused by the networked control system which is the crystallization of the common development of control system and computer network technology.Compared with the traditional control system,networked control system possess the advantages of lower cost,maintenance more convenient,and higher flexibility,so it is widely used in related fields.With the introduction of the network,some new problems have arisen.Some new problems and challenges have been caused for the design and the analysis of networked control system due to the network-induced time delays,packet loss,and network resource constraint.Therefore,this thesis deals with the problem of predictive control for discrete linear networked system Based on the Lyapunov stability theorem.The main contents are listed as follows.Firstly,the problem of networked predictive control is addressed for discrete linear networked control system with parameter uncertainties.A novel networked predictive control strategy is designed to deal with the negative effects caused by network-induced time delays.Based on the Lyapunov-Krasovskii stability theorem,the condition is given to guarantee the finite-time stability of closed-loop system in the form of bilinear matrix inequality.The alternate algorithm is proposed to obtain the predictive control gain matrices.A numerical example is utilized to illustrate the effectiveness of the proposed control method.Secondly,under the transmission strategy of redundant channels and the event-trigger mechanism,the networked predictive control problem is studied for networked control system with input delay.The redundant channels transmission scheme is introduced to reduce the probability of packet loss between the sensor and the observer.In to deal with the negative effect of input delays and network-induced delays and save the network resources,a networked predictive control scheme is proposed based on the event-trigger mechanism.The sufficient criterion of mean-square exponentially stable is established by employing the LyapunovKrasovskii stability theorem.Subsequently,the gain matrices of observer and controller are obtained by solving nonlinear minimization problem with convex constraints.A numerical example is utilized to demonstrate the usefulness of the networked predictive control method based on event triggering mechanism.Thirdly,the networked predictive control problem is discussed for a class of multi-rate networked control system.In order to guarantee the finite-time stability,a novel networked predictive control scheme is designed based on multi-rate sampling mechanism.The closed-loop system is established by employing the techniques of lifting and augmenting.The sufficient conditions of finite-time stability are derived.The solving method of gain matrices is given for the observer and the controller.Furthermore,the results of the study are extended.The limited communication resources are saved by introducing the encoding and decoding protocol.A novel networked predictive control scheme is designed to overcome the negative effects of input delays and network-induced delays.After unifying the sampling rate,the sufficient criterion of input-state stability is acquired.Subsequently,the design means of gain matrices is provided for the observer and the controller.Finally,the feasibility and effectiveness is illustrated by a numerical example for the networked predictive control strategy based on multi-rate sampling mechanism.